Datasets:

jbrinkma

/

pile-500k

like 0

Chapter 6 He had made all kinds of plans to assassinate Li Mu. However, who knew that Li Mu hid in the office and did not come out? Although Bloody-moon faction was really powerful and had an official background, it was still risky to break into a county government and assassinate a ninth-ranked county magistrate. So he could only wait for the chance. But from where Zheng Longxing stood, Li Mu seemed to decide to hide in the county government office all the time, so when would he stop waiting for Li Mu and kill him? "I have to find a way to lure Li Mu to go out." Zheng Longxing thought that and he could hardly wait to do. Days passed by quickly. In a blink of an eye, half a month passed by. In the practice room of the county back office, Li Mu hit the granite which was more than one person tall with his bare hands. A noisy bang spread. The granite, which was hard to hurt by any sword, crumbled as if it was made of flour and became many little rubbles on the ground. Such power was almost comparable to Superman. "I don't know how powerful this punch is." Li Mu satisfactorily blew the stone scraps on his fists. These days, he practiced Zhenwu Boxing in the daytime, and finally, he could complete the first few steps of Zhenwu Boxing perfectly. Meantime, although Li Mu could roughly perform the first move, Rocking-sky Hammer, he could not grasp the deep meaning of it, and therefore after each performance, his muscles were like torn. If he forced himself to complete this entire move, then his muscles would be torn. Even his organ could be injured. Li Mu tried several times and then he quit. Till now, he had a more clear understanding of Zhenwu Boxing and Xiantian Skill. Zhenwu Boxing seemed to be a kind of body-forging technique. Each move and each step had a magical effect of strengthening the body. In this half-month period, Li Mu only practiced a first step and half of the first move of Rocking-sky Hammer. But that was enough for him to notice that his external skin became tough. If he scratched his skin with sharp gravel edges, nothing would be hurt except leaving a faint trace on his skin. The function of Xiantian Skill was opposite to that of Zhenwu Boxing. It could repair internal injuries and make one energetic. Every night, Li Mu was practicing Xiantian Skill. This kind of breathing method could make Li Mu vigorous. Even if he did not sleep all night long, he was still full of energy. Besides, it could also make Li Mu's five senses ability become stronger, which meant his hearing, visional and reaction ability had been greatly enhanced. And Xiantian Skill played a very strong role in the recovery of injuries. Several times, Li Mu used Xiantian Skill to cure those injured internal injuries and muscles caused by improper Zhenwu Boxing practicing. When practicing Xiantian Skill, Li Mu absorbed Spiritual Qi from sky and earth by the strange breathing rhythm and method, and those qis would be transported into the body and washed his organ, then dirty stuff would be excreted through breathing, which was similar to a legend of shaving off one's hair and washing all the bones. Those behaviors changed Li Mu little by little and he seemed to be reborn with a new body. It seemed that Li Mu could understand the intentions of the old faker. Xiantian Skill changed Li Mu's inner organs, while Zhenwu Boxing exercised his body. Those two were complementary to each other, which could completely change a person's body condition. Li Mu had lived on the earth with serious environmental pollution for fourteen years. He breathed in those harmful air and ate harmful substances, which left many wounds and impurities in his body. Now, by practicing these two skills, he could gradually restore the most primitive natural congenital state. Only in this way could he step on the path of interstellar martial arts and compete with those heroes in all the planets. The only thing that made Li Mu a little depressed was that it seemed neither Xiantian Skill nor Zhenwu Boxing could improve his actual combat power. "Hack. Hack..." Li Mu thought for a moment, and couldn't help coughing and spitting out a mouthful of sputum. There were some dark red blood and black dirt in the sputum. That was not the first time. The first time he spat out sputum with blood, he was shocked and thought he might be infected with deadly diseases. Soon, he realized it was because Xiantian Skill was washing his organs and expelling dirty stuff and inner wounds to come out from his body. So this scary bloody sputum spitting-out behaviour meant his lungs were strengthened and cleaned due to Xiantian Skill. "I have hidden in the government office for twenty days. Now it is time for me to come out and breathe some fresh air." Li Mu stretched his body while coughing. Li Mu was actually a lively and active teenager. If he was not afraid of being assassinated by warriors sent from Xue Yue Faction, he would already have hung out in the county streets. Since his personal strength was slightly improved, which gave him some confidence, Li Mu decided to go to the county to breathe some new air after many thought. Old faker once said to him that the taboo of practising martial art was to close oneself and not fight with others. Sometimes, it was more worthwhile to fight against others once rather than practice by yourself in a year. So, one time of a fierce and life-risking fight might equal to ten years of individual hard practice. Li Mu definitely did not want to fight others at the price of a likely death, but it was still good for him to get more experience. Now that he came to the world, he should try to integrate into it. Li Mu was thinking like this. Before he could get the time to ask for those two children when suddenly—— Boom! From the direction of the gate of county government came the thunderous drumming, and the whole county government was shaken by the thunder-like sound. He remembered the scenes that a county magistrate was dealing with lawsuits in court from earth movies and TV plays. Ha ha ha! Li Mu laughed wildly in his heart. Just take this great opportunity to experience the feeling to be a county magistrate and show off. Also, relax. Li Mu could not remember how many times he saw Famous Forensic Pathologist in Song Dynasty andWitness to a Prosecution. Finally, it was time for him to use some knowledge from those TV serials. It must be a serve lawsuit since they beat drums to show their innocence. Li Mu was very excited to think that just waited for me, an alien from earth, to give you inferior people of the martial arts world some lessons with our ancestors' wisdom. While practicing to enhance the fighting power, Li Mu made decisions to be a righteous official, who would serve people and be worshipped by the people. Even thinking about it was enough to make him so happy. "Come people. Hold court trial. Hold court trial!" Official Li couldn't wait to stride towards the court of the front office. "Hey? Childe. Wait a minute. You seem to have forgotten to change to the official clothes..." Qingfeng chased him up and said breathlessly. The little fellow was a little haggard. Since they came to Taibai County, Qingfeng had a feeling of being both his childe's father and mother. At the backyard garden, the little girl Mingyue was holding a net to catch cicadas. She was stunned when she heard the thunderous drums. But she immediately became really excited because she realized that they were going to deal with a lawsuit if someone hit the drum. Wow-ha, Didn't mean that there would be a good show? She opened her mouth, swallowed a cicada lying on the trunk and chewed it with relish. Her lighting-like behaviour was too fast to be observed by any human eyes. It seemed that she was not a normal human being. "Hold court trial... You Majesty..." Six guards stood loosely on both sides, holding the baton to the ground, weakly shouting "your majesty". The courtroom seemed to become a little solemn and formal. Li Mu was sitting on the courtyard in high spirits. Bang! Li Mu hardly patted the gavel. "Bring up the plaintiff." Li Mu began to play the role of a judge quickly. One of a government servant next to him hesitated for a moment, looking strange, then he walked towards Li Mu, coughed in a low voice and said, "Sir, the private adviser is not here, so nobody can record. We can not deal with the case now..." "Then, where is the private adviser? Why is he not coming? " "Ah...The private adviser is not feeling well, and he called for sickness a few days before." "Why did not I know that?" The servant's look was even more strange, and he said, "the private adviser had come to you and asked for sick leave for three or four times in person. But Your Excellency refused to meet him." Li Mu felt embarrassed and blushed. It turned out I should blame myself for this. Then what could he do? Just at that time, the little boy Qingfeng came here breathlessly to bring Li Mu his official outfit, which made Li Mu very excited, and Li Mu said, "Come here, little fellow, you will act as my private adviser for a while, and record the case over there..." Li Mu pointed out the position next to him. "Ah? My childe...Is this inappropriate?" Qingfeng was very confused. Li Muyi smiled and said: "Why is not appropriate? I am the only one who can decide whether it is suitable or not." "Oh." the little boy surely could not argue with his childe, but Qingfeng held his official clothes up and said, "My childe. You'd better put on those clothes first." Li Mu said, "But I am wearing something now." "You should wear an official suit when dealing with a lawsuit." "Those clothes are not comfortable. I do not want to wear. I can decide because I am the county magistrate." Qingfeng,"..." In a moment, the plaintiff was led into the room. Here came a ten-year-old girl wearing mourning dress with tears in her face, and she supported another seriously injured woman who also wore a mourning dress covered in blood stains. They stumbled into the lobby slowly, leaving bloody footprints behind them. My! What happened? Why the plaintiff looked so miserable? Was someone dead? Li Mu was quite anxious. "Please show mercy on us, my county magistrate..." The woman threw herself down on her knees, wailing, and some blood came from her mouth. Then she seemed she could not sit steadily, and spitting out more blood from her mouth. The little girl on the other side was frightened to pale: "Mother. Mother. Please don't frighten me. I don't have grandpa, grandma and daddy any more. Please don't...Mother. Qiner is afraid." Li Mu was also startled by this situation. One servant handed him over the appeal with blood on it. In the past twenty days, Li Mu had a rough understanding of the world's characters, which were quite similar to traditional Chinese characters in ancient official books. After reading their appeal, he could basically know what had happened. This bullying behaviour of gaining other's legal property by cheating or by force was very common in the earth TVs and novels. The badly-injured woman Zhang Li in front of the courtroom run a small pharmacy in Taibai County with her husband Zhang Sheng, her mother-in-law and father-in-law. Their business went really well because they did not cheat anyone and their products were inexpensive but with good quality, which was regarded as a fierce competitor by the No. one pharmaceutical company Shencao Tang, so they found a reason to purchase the small pharmacy by force but their given price was less than one-tenth its actual value. Therefore, her father-in-law refused their request and was beaten to death. Her husband and mother-in-law were so angered by their behaviour and also argued with them, but unfortunately, they were also beaten and seriously injured. Soon, both of them lost their lives. Then, the poor young woman Zhang Li and her daughter Xiao Qin had no one to rely on and were driven out of their drugstore...

Solid particles are known for use as benefit agents in a variety of formulations and personal care compositions. Solid particles can impart benefits both to the compositions comprising them or surfaces to which the compositions are applied. Solid particles can for example be used as pigments or coloring agents, opacifiers, pearlescent agents, feel modifiers, oil absorbers, skin protectants, matting agents, friction enhancers, slip agents, conditioning agents, exfoliants, odor absorbers, or cleaning enhancers. Additionally, many active ingredients useful as treatment agents for various disorders or socially embarrassing conditions are available and typically used in solid particulate form including antiperspirant agents, anti-dandruff agents, antimicrobials, antibiotics, and sunscreens. Typically when it is desired to modify the properties of a surface through application of particles, the particles are applied via leave-on preparations which are rubbed, sprayed, or otherwise applied directly onto the surface to be affected. Typical personal care preparations suitable for delivery of solid particles to hair or skin surfaces include as examples moisturizers, lotions, creams, loose or pressed powders, sticks, tonics, gels, and various sprays such as aerosol or pump sprays. These products are typically applied directly to the surface whereupon particles are deposited and retained by the composition itself or by residual non-volatile elements of the composition after evaporation and drying. It has also been known to formulate solid particle benefit agents into rinse-off or cleansing compositions such as hair rinses, shampoos, liquid and bar soaps, conditioners, or colorants. Frequently the solid particle benefit agent is used to affect the overall appearance, stability or aesthetics of the composition itself. As example, it is well known to add colorant particles, pigments, or pearlescent agents to compositions to improve the acceptability and attractiveness of the product to potential consumers. It is also well known to add particulate benefit agents to affect the in use performance, appearance or aesthetic properties of the composition or to provide a tactile signal to the user. As example, exfoliant particles are frequently used in cleansing compositions to improve abrasion and removal of oils and dirt from washed surfaces and to impart a perceptible “scrubbing” sensation to the user. Typically such particle agents are not intended or desired to be deposited onto the substrate and are removed during dilution and rinsing of the composition from the surface to which they are applied. Given the broad range of benefits which can be delivered through application and retention of solid particles on surfaces, however, it can be highly desirable to have rinse-off compositions capable of depositing an effective level of solid particles to the surface treated with compositions containing the desired solid particle benefit agent. Compositions intended to deposit solid particle benefit agents to hair or skin surfaces are known; however, the efficiency of deposition has heretofore been unacceptable, requiring either an excess of the solid particle agent in the composition to affect delivery or an imperceivable or unacceptable level of the benefit to be obtained. The efficient deposition and retention of solid particle benefit agents is particularly difficult from compositions intended for cleansing or washing of surfaces, such as shampoos or other personal cleansing products, which contain surfactants and other ingredients which are used to solubilize, suspend and remove particle and oily substances from the surfaces treated therewith. It remains, nonetheless, highly desirable to provide the benefits and convenience afforded through deposition of solid particle benefit agents via use of a simple washing composition. It is known to include solid particles in compositions containing cationic polymers; however, these particles are often added to modify the appearance or stability of the composition itself and are not deposited along with the conditioning oils or cationic polymers to the surface treated therewith. When deposition of solid particle benefit agents from washing compositions is intended, the compositions available heretofore have suffered from the drawbacks of inefficient deposition requiring use of excess amounts of the particle agent or ineffective benefit delivery. It has also been attempted to make specific modifications to solid particle benefit agents to improve their deposition efficiency or retention from rinse-off compositions; however this approach can negatively impact the inherent properties, availability, utility, and cost of the solid particle benefit agents to be used. It remains, therefore, highly desirable to have a rinse-off composition, preferably a cleansing composition, capable of containing and effectively depositing and retaining solid particle benefit agents on the surface treated therewith.

/* * Copyright (c) 2012-2018 Red Hat, Inc. * This program and the accompanying materials are made * available under the terms of the Eclipse Public License 2.0 * which is available at https://www.eclipse.org/legal/epl-2.0/ * * SPDX-License-Identifier: EPL-2.0 * * Contributors: * Red Hat, Inc. - initial API and implementation */ package org.eclipse.che.multiuser.organization.api.notification; import static java.util.Arrays.*; import static java.util.Collections.emptyList; import static org.eclipse.che.multiuser.organization.api.DtoConverter.asDto; import static org.mockito.ArgumentMatchers.anyInt; import static org.mockito.ArgumentMatchers.anyLong; import static org.mockito.ArgumentMatchers.anyString; import static org.mockito.Mockito.doReturn; import static org.mockito.Mockito.doThrow; import static org.mockito.Mockito.mock; import static org.mockito.Mockito.verify; import static org.mockito.Mockito.when; import com.google.common.collect.ImmutableList; import org.eclipse.che.api.core.NotFoundException; import org.eclipse.che.api.core.Page; import org.eclipse.che.api.core.notification.EventService; import org.eclipse.che.api.user.server.UserManager; import org.eclipse.che.api.user.server.model.impl.UserImpl; import org.eclipse.che.commons.test.mockito.answer.SelfReturningAnswer; import org.eclipse.che.mail.EmailBean; import org.eclipse.che.mail.MailSender; import org.eclipse.che.multiuser.organization.api.OrganizationManager; import org.eclipse.che.multiuser.organization.api.event.MemberAddedEvent; import org.eclipse.che.multiuser.organization.api.event.MemberRemovedEvent; import org.eclipse.che.multiuser.organization.api.event.OrganizationRemovedEvent; import org.eclipse.che.multiuser.organization.api.event.OrganizationRenamedEvent; import org.eclipse.che.multiuser.organization.shared.model.Member; import org.eclipse.che.multiuser.organization.spi.impl.MemberImpl; import org.eclipse.che.multiuser.organization.spi.impl.OrganizationImpl; import org.mockito.Mock; import org.mockito.testng.MockitoTestNGListener; import org.testng.annotations.BeforeMethod; import org.testng.annotations.Listeners; import org.testng.annotations.Test; /** * Test for {@link OrganizationNotificationEmailSender}. * * @author Sergii Leshchenko */ @Listeners(MockitoTestNGListener.class) public class OrganizationNotificationEmailSenderTest { public static final String API_ENDPOINT = "http://localhost/api"; public static final String DASHBOARD_ENDPOINT = API_ENDPOINT.replace("/api", "/dashboard"); @Mock private MailSender mailSender; @Mock private OrganizationManager organizationManager; @Mock private UserManager userManager; @Mock private OrganizationEmailNotifications emails; @Mock private EventService eventService; OrganizationNotificationEmailSender emailSender; @BeforeMethod public void setUp() throws Exception { emailSender = new OrganizationNotificationEmailSender( API_ENDPOINT, mailSender, organizationManager, userManager, emails); } @Test public void shouldSelfSubscribe() { // when emailSender.subscribe(eventService); // then verify(eventService).subscribe(emailSender); } @Test public void shouldSendNotificationAboutMembershipAdding() throws Exception { // given EmailBean email = mock(EmailBean.class, new SelfReturningAnswer()); when(emails.memberAdded(anyString(), anyString(), anyString(), anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new MemberAddedEvent( "admin", new UserImpl("id", "email", null), new OrganizationImpl("id", "/parent/name", "parent")))); // then verify(emails).memberAdded("name", DASHBOARD_ENDPOINT, "/parent/name", "admin"); verify(email).withTo("email"); verify(mailSender).sendAsync(email); } @Test public void shouldSendNotificationAboutMembershipRemoving() throws Exception { // given EmailBean email = mock(EmailBean.class, new SelfReturningAnswer()); when(emails.memberRemoved(anyString(), anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new MemberRemovedEvent( "admin", new UserImpl("id", "email", null), new OrganizationImpl("id", "/parent/name", "parent")))); // then verify(emails).memberRemoved("name", "admin"); verify(email).withTo("email"); verify(mailSender).sendAsync(email); } @Test public void shouldSendNotificationAboutOrganizationRenaming() throws Exception { // given MemberImpl member1 = new MemberImpl("user1", "org123", ImmutableList.of()); MemberImpl member2 = new MemberImpl("user2", "org123", ImmutableList.of()); doReturn(new Page<Member>(asList(member1, member2), 0, 2, 2)) .when(organizationManager) .getMembers(anyString(), anyInt(), anyLong()); doReturn(new UserImpl("user1", "email1", null, null, emptyList())) .when(userManager) .getById("user1"); doReturn(new UserImpl("user2", "email2", null, null, emptyList())) .when(userManager) .getById("user2"); EmailBean email = new EmailBean().withBody("Org Remaned Notification"); when(emails.organizationRenamed(anyString(), anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new OrganizationRenamedEvent( "admin", "oldName", "newName", new OrganizationImpl("org123", "/parent/newName", "parent")))); // then verify(emails).organizationRenamed("oldName", "newName"); verify(mailSender).sendAsync(new EmailBean(email).withTo("email1")); verify(mailSender).sendAsync(new EmailBean(email).withTo("email2")); } @Test public void shouldDoNotBreakSendingOfNotificationAboutOrganizationRenamingWhenUnableToRetrieveAUser() throws Exception { // given MemberImpl member1 = new MemberImpl("user1", "org123", emptyList()); MemberImpl member2 = new MemberImpl("user2", "org123", emptyList()); doReturn(new Page<Member>(asList(member1, member2), 0, 2, 2)) .when(organizationManager) .getMembers(anyString(), anyInt(), anyLong()); when(userManager.getById("user1")).thenThrow(new NotFoundException("")); doReturn(new UserImpl("user2", "email2", null, null, emptyList())) .when(userManager) .getById("user2"); EmailBean email = new EmailBean().withBody("Org Renamed Notification"); when(emails.organizationRenamed(anyString(), anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new OrganizationRenamedEvent( "admin", "oldName", "newName", new OrganizationImpl("org123", "/parent/newName", "parent")))); // then verify(emails).organizationRenamed("oldName", "newName"); verify(mailSender).sendAsync(new EmailBean(email).withTo("email2")); } @Test public void shouldSendNotificationAboutOrganizationRemoving() throws Exception { // given MemberImpl member1 = new MemberImpl("user1", "org123", emptyList()); MemberImpl member2 = new MemberImpl("user2", "org123", emptyList()); doReturn(new UserImpl("user1", "email1", null, null, emptyList())) .when(userManager) .getById("user1"); doReturn(new UserImpl("user2", "email2", null, null, emptyList())) .when(userManager) .getById("user2"); EmailBean email = new EmailBean().withBody("Org Removed Notification"); when(emails.organizationRemoved(anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new OrganizationRemovedEvent( "admin", new OrganizationImpl("id", "/parent/q", "parent"), asList(member1.getUserId(), member2.getUserId())))); // then verify(emails).organizationRemoved("q"); verify(mailSender).sendAsync(new EmailBean(email).withTo("email1")); verify(mailSender).sendAsync(new EmailBean(email).withTo("email2")); } @Test public void shouldDoNotBreakSendingOfNotificationAboutOrganizationRemovingWhenUnableToRetrieveAUser() throws Exception { // given MemberImpl member1 = new MemberImpl("user1", "org123", emptyList()); MemberImpl member2 = new MemberImpl("user2", "org123", emptyList()); doThrow(new NotFoundException("")).when(userManager).getById("user1"); doReturn(new UserImpl("user2", "email2", null, null, emptyList())) .when(userManager) .getById("user2"); EmailBean email = new EmailBean().withBody("Org Removed Notification"); when(emails.organizationRemoved(anyString())).thenReturn(email); // when emailSender.onEvent( asDto( new OrganizationRemovedEvent( "admin", new OrganizationImpl("id", "/parent/q", "parent"), asList(member1.getUserId(), member2.getUserId())))); // then verify(emails).organizationRemoved("q"); verify(mailSender).sendAsync(new EmailBean(email).withTo("email2")); } }

IIT JAM 2018 Online Registration Application Form Joint Entrance Exam (JAM) is the All India Online Entrance Examination organized by the Indian Institute of Technology (IIT) on rotation basis on behalf of the Human Resource Development Ministry (MHRD), Government of Continue reading →

Posts tagged discontinued Nintendo is about to halt production of its 8GB Wii U Basic in Japan, according to a notice on its product site. The white-clad product has been around since 2012, but was never super popular since the 32GB version is typically only $50 more. In fact, the rumor mill had it being killed off two yea... Microsoft will phase out sales for the original Kinect for Windows in 2015. The decision follows the arrival of the sensor's second version and PC adapter in October, as well as the latest hardware's SDK being made freely available. The company noted in an announcement blog that companies reliant o... Whether you think depth-sensing cameras will become the next big thing or fizzle out, one of the models that helped usher in the tech is set to be retired for good. The original Kinect for Windows will be phased out in 2015, nearly three years after Microsoft first introduced it. The sensor (which... "Certain sold-out Amiibo may return" to store shelves at some point, according to a new Nintendo of America statement provided to Joystiq last night. That may offer hope to those hunting down the rarest toys, but it doesn't necessarily run counter to earlier statements indicating "other [less popul... Nintendo of America confirmed this week some Amiibo toys "likely will not return to market," following a retailer's claim that the publisher had officially discontinued the Marth, Wii Fit Trainer and Villager figures. In a statement provided to several sites including Wired, Kotaku and Destructoid,... You know those cute Amiibo NFC-enabled toys/collectibles/game accessories that Nintendo debuted this year? They aren't going anywhere for a while. Well, some of them won't, at least -- and so far, that includes Mario and Link, according to Wired. The outfit says that it intends to have certain cha... The super-size 3DS XL is about to be downsized in Japan, at least according to new details spotted on Nintendo's website. As Siliconera reports, Nintendo's 3DS site lists "ending production soon" notices for the seven XL colors available in the country. Nintendo's not announced anything formally, b... The free music streaming feature for Xbox Music will be discontinued on December 1, Microsoft revealed on an Xbox Support page for the service. The company said it is "focusing Xbox Music to deliver the ultimate music purchase and subscription service experience for our customers" with Xbox Music P... Sony is closing off PlayStation Network access for its PlayStation Portable console in multiple regions starting this week, IGN reports. A Sony representative told IGN that the shutdown affects PSP owners in Europe, Asia, Oceania, and select regions in Africa. North American PSPs will still be abl... Every time Apple holds one of its keynotes, we think to ourselves, "Maybe this is the year they'll kill off the iPod classic." Finally, after a years-long stay of execution, Apple's oldest living media player is going the way of the dodo. The company just reopened its online store after announcing... Despite the game being discontinued on Xbox 360, existing owners of Ascend: Hand of Kul will remain able to play it until November 18, after which it will officially close. The news follows Microsoft's removal of the free-to-play action-RPG from the marketplace, a decision developer Signal Studios... Free-to-play action-RPG Ascend: Hand of Kul is no longer available on the Xbox 360 after Microsoft discontinued it this week, apparently without informing developer Signal Studios beforehand. In what's seemingly less a case of "So long, farewell" and more Jazzy Jeff leaving the Fresh Prince's house...

Danger Zone – Abella Danger Awesome Abella looks unbelievable as a blonde bombshell. After showing off her awesome curves as she teases you with affectionate glances, She retires to the hotel where her sexy suitor for the evening is eager to satisfy her every desire.

Ion mobility spectrometers are analytical instruments that are used to separate ions in time as a function of ion mobility. It is desirable to be able to control electric fields applied to such instruments in order to investigate properties of charged particles.

Is high-altitude environment a risk factor for childhood overweight and obesity in Saudi Arabia? To describe the prevalence of childhood overweight and obesity in rural high- and low-altitude populations of southwestern Saudi Arabia and to identify specific at-risk groups within these populations. A cross-sectional study was conducted on 912 school children and adolescents aged 6-15 years born and living permanently at high altitudes (2800-3150 m) and 972 children and adolescents of comparable ages born and living permanently at low altitudes (< or =500 m). Height and weight were measured. For children <10 years, the weight-to-height index according to World Health Organization (WHO) standards was used for assessing overweight and obesity. For adolescents 10-15 years, overweight and obesity were assessed by age and gender-specific percentiles for body mass index based on the WHO/National Centre for Health Statistics reference population. A questionnaire was used for measuring parents' socioeconomic status. The overall prevalence of overweight and obesity at high and low altitudes was 10%. The study showed that some school children and adolescents were at a significantly higher risk of developing overweight and obesity. Significant risk factors included moderate-to-high parental income, age > or =10 years, high-altitude birth and residence, and female sex (crude odds ratio 3.2 [95% CI, 1.8- 5.5], 2.3 [95% CI, 1.6-3.2], 2.1 [95% CI, 1.5-2.9], and 1.9 [95% CI, 1.4-2.6], respectively). A multivariate analysis using the direct binary logistic regression model revealed that moderate-to-high parental income, age > or =10 years, female sex, and high-altitude birth and residence were significant independent predictors of childhood overweight and obesity. (adjusted OR 3.2 [95% CI, 1.6-2.6], 2.6 [95% CI, 1.8-3.8], 2.0 [95% CI, 1.6-2.9], and 1.8 [95% CI, 1.3-2.6]), respectively. The present study identified risk factors for childhood overweight and obesity in Saudi Arabia. Among these, high altitude was a significant and independent factor. Future research is warranted to evaluate the exact mechanism by which a high-altitude environment may contribute to childhood overweight and obesity.

Q: Laravel edit children in parent view I try to get children's in their parents view with ability of editing them, Codes controller public function edit($id) { $option = Option::findOrFail($id); return view('admin.options.edit', compact('option')); } blade // getting parent info in edit blade {{ Form::model($option, array('route' => array('options.update', $option->id), 'method' => 'PUT')) }} {{ Form::label('title', 'Name') }} {{ Form::text('title', null, array('class' => 'form-control')) }} {{ Form::submit('Update', array('class' => 'btn btn-success')) }} {{ Form::close() }} // Geting this option subs (children's) in my second tab @foreach($option->suboptions as $sub) {{ Form::model($sub, array('route' => array('suboptions.update', $sub->id), 'method' => 'PUT')) }} {{ Form::label('title', 'Name') }} {{ Form::text('title', null, array('class' => 'form-control')) }} {{ Form::label('price', 'Price') }} {{ Form::text('price', null, array('class' => 'form-control')) }} {{ Form::submit('Update', array('class' => 'btn btn-success')) }} {{ Form::close() }} @endforeach Issue As my sub's form is inside the loop I get update button for each row (children) how can I only have 1 button? UPDATE check this video to see what I mean. UPDATE 2 More explain! my page is include 2 different forms (one of them edits my option eg. size that's fine we no have problem with that) my second form which is my issue will return subs of my option, eg. option = size, subs = 12 inch, 15 inch etc.) in order to edit this subs currently i have 1 button for each of them (as you see in video i shared), what i want is to have only 1 button for all of my second form. A: SOLVED I've made controller below and I can update my data as I wanted: public function subsupdate(Request $req) { for ($i=0; $i<count($req->id); $i++) { DB::table('suboptions') ->where('id',$req->id[$i]) ->update([ 'title' => $req->title[$i], 'price' => $req->price[$i], ]); } return redirect()->back()->with('success', 'Suboptions updated successfully.'); }

Today in Tribe History: April 2, 1995 The longest strike in major league history comes to an end. Having the first 23 days of this major league season canceled and 252 games of the last season lost, the owners accept the players’ March 31st unconditional offer to return to work. The players’ decision to return to work is made after a US District Court issued an injunction restoring terms and conditions of the expired agreement. Teams will play 144-game schedules. The strike had begun on August 12, 1994.

<?xml version="1.0" encoding="utf-8"?> <style xmlns="http://purl.org/net/xbiblio/csl" version="1.0" default-locale="de-DE"> <!-- Thieme Fachzeitschriften, generated from "thieme" metadata at https://github.com/citation-style-language/journals --> <info> <title>Endoscopy International Open</title> <id>http://www.zotero.org/styles/endoscopy-international-open</id> <link href="http://www.zotero.org/styles/endoscopy-international-open" rel="self"/> <link href="http://www.zotero.org/styles/thieme-german" rel="independent-parent"/> <link href="http://www.thieme.de/de/autorenlounge/fuer-zeitschriftenautoren-1789.htm" rel="documentation"/> <category citation-format="numeric"/> <category field="biology"/> <category field="medicine"/> <issn>2364-3722</issn> <eissn>2196-9736</eissn> <updated>2017-07-03T12:00:00+00:00</updated> <rights license="http://creativecommons.org/licenses/by-sa/3.0/">This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License</rights> </info> </style>

/* * Copyright (c) [2019-2020] Huawei Technologies Co.,Ltd.All rights reserved. * * OpenArkCompiler is licensed under the Mulan PSL v1. * You can use this software according to the terms and conditions of the Mulan PSL v1. * You may obtain a copy of Mulan PSL v1 at: * * http://license.coscl.org.cn/MulanPSL * * THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY OR * FIT FOR A PARTICULAR PURPOSE. * See the Mulan PSL v1 for more details. */ FUNCAPHASE(MeFuncPhase_DOMINANCE, MeDoDominance) FUNCAPHASE(MeFuncPhase_SSATAB, MeDoSSATab) FUNCAPHASE(MeFuncPhase_ALIASCLASS, MeDoAliasClass) FUNCAPHASE(MeFuncPhase_SSA, MeDoSSA) FUNCAPHASE(MeFuncPhase_IRMAP, MeDoIRMap) FUNCAPHASE(MeFuncPhase_BBLAYOUT, MeDoBBLayout) FUNCAPHASE(MeFuncPhase_MELOOP, MeDoMeLoop) FUNCTPHASE(MeFuncPhase_BYPATHEH, MeDoBypathEH) FUNCAPHASE(MeFuncPhase_MEANALYZECTOR, MeDoAnalyzeCtor) FUNCAPHASE(MeFuncPhase_MEABCOPT, MeDoABCOpt) FUNCAPHASE(MeFuncPhase_CONDBASEDNPC, MeDoCondBasedNPC) FUNCTPHASE(MeFuncPhase_MAY2DASSIGN, MeDoMay2Dassign) FUNCTPHASE(MeFuncPhase_LOOPCANON, MeDoLoopCanon) FUNCTPHASE(MeFuncPhase_SPLITCEDGE, MeDoSplitCEdge) FUNCTPHASE(MeFuncPhase_PROFGEN, MeDoProfGen) FUNCTPHASE(MeFuncPhase_PROFUSE, MeDoProfUse) FUNCTPHASE(MeFuncPhase_DSE, MeDoDSE) FUNCTPHASE(MeFuncPhase_HPROP, MeDoMeProp) FUNCTPHASE(MeFuncPhase_HDSE, MeDoHDSE) FUNCTPHASE(MeFuncPhase_SSADEVIRT, MeDoSSADevirtual) FUNCTPHASE(MeFuncPhase_SSAEPRE, MeDoSSAEPre) FUNCTPHASE(MeFuncPhase_SSALPRE, MeDoSSALPre) FUNCTPHASE(MeFuncPhase_STOREPRE, MeDoStorePre) FUNCTPHASE(MeFuncPhase_STMTPRE, MeDoStmtPre) FUNCTPHASE(MeFuncPhase_SSARENAME2PREG, MeDoSSARename2Preg) FUNCTPHASE(MeFuncPhase_PREGRENAMER, MeDoPregRename) FUNCTPHASE(MeFuncPhase_ANALYZERC, MeDoAnalyzeRC) FUNCAPHASE(MeFuncPhase_DELEGATERC, MeDoDelegateRC) FUNCAPHASE(MeFuncPhase_CONDBASEDRC, MeDoCondBasedRC) FUNCTPHASE(MeFuncPhase_RCLOWERING, MeDoRCLowering) FUNCTPHASE(MeFuncPhase_EMIT, MeDoEmit)

OFF-SITE QUICK REACTION SYNTHETIC CHEMISTRY

Various The Japanese youth, K. Yamamoto should act in accord with the Law of God—namely—he must first choose a companion (wife) for himself and then his father and mother must sanction. If their sanction and consent is not attained, that engagement is not completed. He must acquire the English language well, so as to enable him to translate the Divine Tablets into the Japanese language.

A novel approach to identify and map kitten clusters using geographic information systems (GIS): a case study from Tompkins County, NY. A retrospective study using a geographic information system (GIS) was conducted to capture, map, and analyze intake data of caregiver (owner)-surrendered kittens (aged 0-6 months) to the Society for the Prevention of Cruelty to Animals (SPCA) of Tompkins County, NY, from 2009 to 2011. Addresses of caregiver-surrendered kittens during the study period were mapped (n = 1,017). Mapping and analysis of the resultant data set revealed that the distribution of kittens was nonrandom. Seventeen statistically significant (p = .001) clusters were identified, 1 of which was the SPCA of Tompkins County (due to anonymously surrendered nonhuman animals). The remaining 16 clusters were composed of 52 homes; 27.5% (280/1,017) of the kittens in the data set originated from these 52 homes. The majority of kittens within clusters were surrendered from high-density residential and manufactured residential home parks. Analyzing such clusters using GIS is a novel approach for targeting spay/neuter and educational programs to areas contributing disproportionately to shelter populations. This method may prove useful to help shelters more effectively allocate their limited resources, but further evaluation of this and other targeted approaches is needed to assess the long-term efficacy of such programs.

A place where we could add stories and photos related to Kiribati! THANK YOUThanks to everyone who has signed up, used or visited SFK for whatever purpose. Your visits and connection indicate that our contents (stories & photographs) are being useful and therefore worth maintaining them. We’ll keep adding more items for you! UPGRADED SFKWe recently upgraded SFK to the next level. It’s now on a paid plan. Your activities (visits, views, messages, comments) as well as our interest have inspired us to go further and better! ii) The connection is now secure, safe and private (see lock in the url) - Great! iii) There is more web-space for our articles and pictures. iv) And more ... WHAT’S NEW?i) Noticeable changes were made to the skin, footer & navigation in the homepage ii) New stories & photographs have been added & updated iii) Donate button placed on the sidebar iv) More items/links may show up later on DONATE BUTTONThe button is only for those who wish to donate money. Your donation will be used for the costs and betterment of our website. Thank you. MOBILE PHONE USERSMany people are accessing the internet from their phones. With that in mind, we hope SFK is fast, clear and displayed well on your mobile phones. Yet, we can’t deny that SFK works and looks better on computers. PHOTOGRAPHSDon’t forget to visit the photo page as we keep adding more photographs. We’ve committed ourselves to take best shots for your use and joy. Equipped with the Nikon d3100 with kit and Tamron lens, we anticipate to bring up amazing photographs of the environment.

Exclusive Interview: George Galloway talks Media, Corbyn & Sexism George Galloway is a British politician, broadcaster, and writer. Aged 26, he became chairman of Labour in Scotland. Six years later, he became an MP in the House of Commons. He was expelled from the Labour Party in 2003 for his outspoken comments on the Iraq war. He then led the anti-war party Respect. In 2017, he produced the documentary “The Killing$ of Tony Blair”. M: One accusation I keep stumbling upon is that you’re sexist. I would like to clear this out for our readers. Lisa Nandy, the MP coordinating Labour’s campaign, said constituents in the seat left vacant after the death of Sir Gerald Kaufman last month, “deserve better than a man who has described the sexual assault of women as ‘bad sexual etiquette’”. Some Corbynistas buy into this story despite being in an ideal position to know how manipulative the media is. People might be disenchanted with the mainstream media in many regards, but many among them still believe the official narrative that Assange is a rapist and that you are a rapist apologist because of your defence of Julian Assange, by describing the allegations of sexual offence as bad etiquette. What do you say to such assertions based on your defence of Assange? George Galloway: Indeed, Corbynistas above anyone else should know MSM is an organised conspiracy against the truth. Why would you accept the traducing of me when they reject the traducing of Jeremy Corbyn so comprehensively? The people who attack me on these identity politics issues are the very people who attack Corbyn. It surprises me there are some people on the left who have not yet fully accepted that Julian Assange is victim of a gigantic conspiracy. I’m not myself a conspiracy theorist but that doesn’t mean there are no conspiracies. Assange was set up for the purpose of getting him to the US where he’d spend the rest of his days in a dungeon like Chelsea Manning would’ve, if she wasn’t released by presidential pardon. All the rest is detail. He raped no one. The alleged victims have stated that he raped no one. The Swedish authorities allowed him to leave the country because he was not wanted for any sexual offence. I will defend assange til my last breath. I believe in him, I know him well and I know what he’s done will stand the test of time, as it’s of historic importance. As to the issue of rape, quite clearly sex without consent is rape and a grave and serious crime. No means no, in popular parlance. But Julian’s defence of the “alleged allegations” (because these allegations are now denied by the women), was that he had a reasonable belief that they consented to consensual sex. Which is the agreed defence available to any accused in Britain and Sweden, by law, if someone has a reasonable belief then they are entitled to that defence. M: But in Sweden simply not wearing a condom is defined as rape by law, that’s what they accused him of.. George Galloway: Well, that’s what it was said they said — having to use this torturous phrasing because these allegations are now denied by the women themselves. M: Do you think the women tried to set him up? There are texts from one woman to another proposing the sale of the information to a local paper, now in the possession of the police. George Galloway: At this delicate stage — and it’s a very delicate stage, I don’t want to go further into the motives of allegations because this moved, long ago, into the realm of state and interstate politics. If the Swedish authorities truly wanted to question him on allegations sexual offences, they could have done so in London, but they refused to do so, even though with many other alleged criminals they have done it — even murder cases, or they could have said “come to Sweden to be questioned and we guarantee no extradition”. Their refusal to do either of these to me makes it abundantly obvious that this is an attempt to get Julian to the US where a grand jury has already prepared an indictment grand trial.. he’d go straight into an unfair trial which is not about the use of a condom but against the exposure of US war crimes. Update: Since this interview, taken on May 15th, Sweden has withdrawn the accusations against Assange, making it more obvious that what George Galloway said here about Assange’s case is true and fair. M: WHAT DO YOU SAY TO JEREMY CORBYN’S REMARKS THAT YOUR CAMPAIGN AGAINST NAZ SHAH WAS APPALLING? George Galloway: He never made it. M: But the press reported it. George Galloway: Lots of things Jeremy Corbyn never said were reported. I had 20 conversations with him and I can say to you he never said it. M: Now, in an article for Progress, Dawn Butler, who nominated Corbyn for the Labour leadership (later voting for Andy Burnham), has gone public with her communication with the Labour leader. “I have spoken to Jeremy Corbyn and he has told me he is not in favour of letting Galloway back in,” she writes. George Galloway: Also false. M: Do you think he’ll welcome you back to the party if Labour is in government after this election? George Galloway: If I’m elected, as soon as I’ve sworn in I’ll go and sit beside him in Parliament. No one can stop us. Not the NEC, not anyone. The NEC’s writ does not extend to MPs who wish to take the Labour party whip. A complex esoteric point, perhaps, one that needs to be made nevertheless. George Galloway: That would be an argument if this was a by-election and not a General Election. If this wasn’t now a General Election, the argument that the defeat of labour under Jeremy Corbyn by me in the context of a by-election would have had some way… which I carefully calculated. If Sam Wheeler, (the Corbyn candidate) were selected I wouldn’t have this conversation with you now. But when Watson and Vaz rigged the selection with the explicit purpose of further undermining Jeremy Corbyn, that argument went out the window. Now in a General Election, it doesn’t matter, because, he either wins or loses; but if he wins, that I defeated Afzal Khan doesn’t matter, he’ll have my vote… if he loses, my vote will be stronger. All of us know there will be a coup by Watson and co, they will seek to overthrow Jeremy and he’ll have an extra MP in me defending him. If it was Afzal it would be one extra MP to undermine him. M: You claimed that Afzal Khan would stab Jeremy Corbyn in the back. Afzal responded in a public meeting at the Klondyke that he would not stab him in the back. What evidence do you have that he would betray Jeremy? George Galloway: I know he’s Tom Watson and Vaz’ candidate, which is undisputable, if he wasn’t, they wouldn’t have put him up; they wouldn’t have rigged the short list if they were doing so for a Corbynite. M: You don’t think the pro-Corbyn faction of the NEC might have managed to put a Corbynite in? George Galloway: As soon as the five-person panel was surgically altered in one day, when Rebecca Longbailey was taken off the selection committee and replaced by Vaz, making a three Jeremy Corbyn majority into a three Corbyn minority, that dye was cast. What conceivable reason could there be for taking Longbailey, a Manchester MP, off the panel, if not to alter its political balance? And to replace her with the discredited figure of Keith Vaz, to me it’s screamingly obvious what was going on, and to Sam Wheeler also. M: So you think Afzal’s support of Palestine is just a facade? George Galloway: Not seen much of it. I understand he did have something on social media about that. We can no longer tell because he wiped it off. M: People’s understanding of politics tends to be slightly myopic, in that they vote for the party that is best at filling potholes or the candidate that gives them nice bike lanes, missing out the bigger picture, like war and a rigged economy. What do you have to say to those who criticise your career as MP on the grounds you don’t get too involved in local communities — never present at surgeries, and spend too much time on (alternative) TV? George Galloway: Pack of lies. I had a surgery in Bradford every Saturday for three years. Including holidays. Every week. Packed surgery, thousands of cases. Pack of lies. It was New Labour canard coined as soon as I defeated them in Bradford. Coming of the same stable is that I’ve lost my seat in Glasgow and Bethnal Green. I never defended my seat in Glasgow because it was abolished, wiped off the political map. Blair expelled me in 2003 and in 2004 abolished the constituency. That’s how I ended up in Bethnal Green. Before I stood in Bradford I announced I would stand there only once and I would make way at the end of my term for a Bengali MP. I didn’t lose my seats, I didn’t defend them. The only seat I ever lost was in 2015, Bradford West and that’s a whole different story. M: Well, this is another reason why people are left with the impression you are sexist, your accusations against Naz Shah’s accusations, what happened there? George Galloway: I don’t want to open that because my comrade Salma Yaqoob is standing against Naz Shah in Bradford West, I hope she wins and I don’t want to open any old wounds that might harm her campaign. M: What chances do you think Jeremy has in this election? Or what chances are there, of a Labour government? George Galloway: Not much chance of winning, if you mean by winning a Labour majority, that’s impossible for anyone but Jesus. The election result in 2015 was so poor under Ed Miliband that the mountain needed to be climbed is too great. The best that Jeremy Corbyn can do is head of a coalition of non-Tory parties to frustrate May’s ability to form a government. That too has been undermined by the complete failure of the Lib Dem soufflé to rise. At the start of the campaign everyone believed, including me, that Lib Dems would do well, recapture seats lost in 2015, but they’re going down in polls. Rather losing seats. SNP are — very untrustworthy coalition partners, let me tell you — also going down. Tory going up in Scotland. UKIP collapsing under our eyes: down 10%. So Corbyn’s chances of leading a non-Tory majority are also slim. Having said that Corbyn is doing much better than the MSM imagined he would do, significantly better than Ed, Labour is at 32%, it had 29.5% still two weeks to go; he has weaknesses but he’s strong at campaigning, he’s extremely good at it. This is campaigning weather, this is not “will the right honourable gentleman”, this is up close and personal on the streets and in communities and Jeremy is great at that and Theresa is dreadful at that, so in two weeks I expect him to continue to rise. I don’t believe he can rise to win enough seats to form government, to stop May forming government, but the idea that he’d lead Labour to a landslide defeat is dead, people still repeat it, including the labour traitors, but it’s clear to me he’d get a remarkably good result in the circumstances: 30% plus up to 35%. If the labour traitors had not spent two years stabbing him in the back, who knows where he’d be? Update: Labour is now at 40% according to Ipsos MORI. M: They did declare they’d rather Labour lost than win with Corbyn. George Galloway: Tony Blair said that… Peter Mandelson said he’s working every day to sabotage the leader… and they’re still Labour, while I am not. M: What do you think the Left can hope for if they lose this election? Do you think there will be tyranny under the Tories? George Galloway: There won’t be tyranny. Politics is a dynamic business. If Corbyn gets this result I predict, there will be a better platform to build on than any comparable left party in Europe. To compare with the French Socialist party’s 6%… Melenchon is comparable to Corbyn. If you had Labour in France with Corbyn’s score, he would have won. So it’s not discreditable to have 38% in a general election with a leader like Corbyn and given the extent of the fifth column activities in the party and the wall of sound media hostility. It’s by no means discrediting. Opposition will continue to be mounted, there won’t be tyranny. There’s only tyranny when there’s no democracy and we still have democracy. Tyranny is a strong word. You as a Romanian would know. M: Especially as a Romanian I would know! I don’t come with rosy glasses, it may seem on the surface that there’s democracy and freedom of opinion here but there isn’t such a fundamental difference, because in this society, also, an uncomfortable opinion can get you in pariah places, as you would know. George Galloway: You have to have perspective. If this was a tyranny I wouldn’t talk here to you, I wouldn’t stand in an election, I wouldn’t have two million followers on social media to communicate my message to. There are restrictions but this is balanced in civil society with a conscientious politically educated mass that is acting as counterweight to it. If the Tories win, they can do all kind of ghastly things, but they won’t be able to introduce a tyranny. They won’t be able to close down the new space available to opponents. M: But things like Snoopers Charter or their expert discrediting of dissidents shows they have depraved power, which Ceausescu would be jealous of! George Galloway: I’m not disputing they have depraved power and they are vile and dangerous, just we have to be careful about the word. The police would love to come and close us down right now, but they can’t. M: What media do you recommend? George Galloway: There’s an RT quote: “Whatever you’re watching, watch less”… read more widely. I read in the morning, online (never buy the paper!), in this order: Daily Mail, Independent, Guardian. Then Morning Star, which has limitations in web presence, and then BBC and watch RT. I, therefore, know what’s going on from different perspectives. Ignorance is a choice now. Young people adept at tech know how to find sources. The Canary is phenomenal, a must read every day. These new forces by the time you’ll be my age they’ll be giant. The MSM will have faded away. Not a brave world in the future, but 30 years in the future. It’s why I’m optimistic. The power of the ruling class is diminished and will diminish. M: Why trust the mainstream media? George Galloway: I don’t trust them. I don’t trust the Daily Mail to give me the date. But I need to see what they’re saying. If they’re in hysterical overdrive about Corbyn’s manifesto, that tells me that’s what they worry about. Don’t trust the Guardian. But I need to know what they’re saying. Russia Today… although I work for and watch it, I know they are biased on France and Scottish independence. M: Russia Today also lies? George Galloway: Not a question of lies, but bias. RT is favourable to Scottish independence. It would suit Russian state policy if Britain were weakened. It would suit if EU were disunited. M: So it seems Russian interests at this moment in time align with the survival interests of the rest of us! George Galloway: A good way of putting it. RT is right most of the time for its own reasons. M: How do you think it came to this plateau? I’m conditioned to have a vitriolic hatred for Russia, I grew up with a worldview where Russians were the bad guys, but shocked now that RT is flawless information and Americans are bad. George Galloway: I never thought Russians were as bad as you thought. Russia had a bad period under Yeltsin, the drunken fool. But as soon as Putin came to power I knew his preoccupation was to make Russia great again and he’s done that. Not being a Russian I have no particular interest in making Russia great again. I want that there isn’t one great power because the more great powers there are, the less monopoly one has. I grew up knowing American’s were the bad guys. Now that Russia is much stronger, united than it was before, this produces a coincidence of interests between them and the rest of us on the left. There will come a point where that will cease to be true, it did come in the French election last week, as Russia would have preferred Le Pen to have won. M: Wasn’t she preferable to the banker Macron? George Galloway: Neither Macron nor Le Pen. Melenchon. Which was his slogan. Melenchon was cheated by a place in the final ballot by a failed campaign. He would have won. I can’t offer support for a fascist, with a Vichy history. Russia Today did. Yes, she’s anti-globalist. So were the brown shirts and the Nazi party. A lot of fascism has a lot of anti-capitalism, anti-banker but that doesn’t disguise the fact they were Nazis, which would have led to deportations, the trains, of foreigners, aliens. That kind of socialism is the socialism of fools. M: Who do you think people should vote for in West Bromwich East? George Galloway: That’s the most difficult question I’ve been asked in this election. It’s an issue, like the camel, difficult to define but easy to recognise… Some Labour MPs I could hold my nose and vote for, some for whom I could skip into ballot station to vote for, but others I simply could not vote for nor could ask others to vote for. I suspect West Bromwich is one of those, Wallasey another. Tom Watson, John Woodcock, Wes Streeting, the worst offenders, the people who never stopped attacking Corbyn, Connor McGinn, he was a whip under Corbyn and actually used his time as whip against Corbyn, to organise the coup members to overthrow Corbyn. It would be impossible for me to recommend voting for them, even with the tightest hold on my noose. I think that because I was for 40 years a follower of Tony Benn, I would stop short of not voting for Hillary Benn. Comments Witnessing the Iron curtain collapse as a Romanian child forever colored my political leanings & understanding. Now a Labour and Momentum activist, I write for socialist paper the Word and tweet at @MaraLeverkuhn. Freedom is the ultimate goal, reclaiming democracy the strategy. The first step of which—to analytically undo sediments of mainstream misinformation.

// Generated by BUCKLESCRIPT, PLEASE EDIT WITH CARE 'use strict'; var a = 4; var b = 2; exports.a = a; exports.b = b; /* No side effect */

--- abstract: 'We consider the problem of modeling, estimating, and controlling the latent state of a spatiotemporally evolving continuous function using very few sensor measurements and actuator locations. Our solution to the problem consists of two parts: a predictive model of functional evolution, and feedback based estimator and controllers that can robustly recover the state of the model and drive it to a desired function. We show that layering a dynamical systems prior over temporal evolution of weights of a kernel model is a valid approach to spatiotemporal modeling that leads to systems theoretic, control-usable, predictive models. We provide sufficient conditions on the number of sensors and actuators required to guarantee observability and controllability. The approach is validated on a large real dataset, and in simulation for the control of spatiotemporally evolving function.' author: - 'Hassan A. Kingravi, Harshal Maske and Girish Chowdhary [^1]' title: ' Kernel Controllers: A Systems-Theoretic Approach for Data-Driven Modeling and Control of Spatiotemporally Evolving Processes ' --- =1 Introduction {#sec:intro} ============ Modeling, control, and estimation of spatiotemporally varying systems is a challenging area in controls research. These systems are characterized by dynamic evolution in both the spatial and temporal variables. Some examples of relevant problems include active wing-shaping based control of flexible aircraft, control of heat or particulate diffusion in manufacturing processes, control of rumor spreading across a social network, and tactical asset allocation and control problems in dynamically varying battlespaces. The traditional approach to modeling and control of spatiotemporal systems have relied on Partial Differential Equations (PDEs) [@Alabau2012controlpde], solutions to which are functions that evolve in both space and time. However, PDE models can be limited in situations where exact physics based models of the functional evolution are difficult to formulate, or are inherently limited due to the physical understanding of the process or unknown spatiotemporal interactions [@cressie2011statistics]. Furthermore, the control of PDEs is fundamentally more challenging than the control of finite-dimensional state-space systems because the evolution and control spaces are infinite dimensional Hilbert spaces, as opposed to $\mathbb{R}^n$ [@Alabau2012controlpde]. Accordingly, there has been significant work in approximate modeling of spatiotemporally evolving functions using data-driven or distributed parameter based approximations of PDEs [@cressie2011statistics; @wikle2001kernel]. One way to model spatiotemporally evolving functions is to approximate the function at several sampling locations and build an autoregressive model of the evolution of the function’s output over that grid [@baker2000finite]. The fidelity of these models heavily depends on the number of sampling (equivalently Euclidean grid locations in the independent variable space) locations employed, with a large number of grid locations leading to large-scale state-space models that are difficult to manage. An alternative approach to modeling spatiotemporal functional evolution relies on modeling the correlation between any two sampling locations through a smooth covariance kernel [@cressie2011statistics]. The model of the evolution is then formed through a linear, weighted combination of the kernels, and the hyperparameters of the spatiotemporal covariance kernel and the weights are learned by solving an optimization problem. The power and flexibility of this approach lies in the fact that kernels can be defined over abstract objects, and not just Euclidean grid locations, leading to a modeling technique that is domain agnostic. For example, kernel embeddings are available for graphical models studied in decentralized control [@johansson2014global], images [@ren2012coupled], and many other domains. However, formulating control-usable kernel-based models of spatiotemporal phenomena can be challenging due to the need to take into account the spatiotemporal dependence. Many recent techniques in spatiotemporal modeling have focused on covariance kernel design and associated hyperparameter learning algorithms [@garg2012AAAI; @ma2003nonstationary; @RasmussenWilliams2005; @plagemann2008nonstationary]. The main benefit of careful design of covariance kernels over approaches that simply include time in as an additional input variable [@perez:13:gaussian; @Chowdhary13_ACC2] is that they can account for intricate spatiotemopral couplings. However, there are two key challenges with these approaches: the first challenge is in ensuring the scalability of the model to large scale phenomena. This is difficult due to the fact that the hyperparameter optimization problem is not convex in general, and because when time is used as a kernel input, it is nontrivial to restrict the number of kernels used without losing modeling fidelity [@garg2012AAAI; @ma2003nonstationary; @plagemann2008nonstationary]. The second very important challenge is concerned with the formulation of feasible control strategies utilizing predictive kernel-based models of spatiotemporal phenomena. In particular, when the spatiotemporal evolution is embedded in the design of complex covariance kernel, the resulting model of functional evolution can be highly nonlinear and difficult to utilize in control design. In this paper, we pursue an alternative systems-theoretic approach to the modeling, control, and estimation of spatiotemporally varying functions that fuses the strengths of kernel methods with systems theory. Our main contribution is to provide a systems-theoretic formulation for approximating, with very high accuracy, spatiotemporal functional evolution by layering a linear dynamical systems prior over temporal evolution of weights of a kernel model. For a class of linearly evolving PDEs, such as the heat diffusion and the wave equation, our approach can lead to a very high-accuracy approximation. This modeling approach is also applicable to data-driven modeling of real-world phenomena, which we demonstrate on a challenging inference problem on satellite data of sea surface temperatures. One benefit of our model is that it can encode spatiotemporal evolution of complex nonlinear surfaces through an Ordinary Differential Equation (ODE) evolving in a Hilbert space induced by the specific kernel choice. Yet, the main benefit of our systems-theoretic approach is that it is highly conducive to control synthesis. To illustrate this fact, we demonstrate that feasible control strategies for a class of spatiotemporally evolving systems can be found using linear control synthesis. In particular, we derive sufficient conditions on the kernel selection to guarantee observability and controllability of the presented model. Furthermore, we demonstrate control synthesis for a diffusion PDE using simple Gaussian kernels distributed uniformly in the input domain. The outline of this paper is as follows, Section \[sec:observers\] focuses on the development of a systems-theoretic kernel-based model of spatiotemporal evolution, Section \[sec:theory\_results\] presents the main theoretical results, Section \[sec:experimental\_results\] presents modeling results on a real-world large dataset and control synthesis results for a diffusion PDE. Kernel Controllers {#sec:observers} ================== This section outlines our modeling framework and presents theoretical results associated with the number of sampling locations required for monitoring functional evolution. Problem Formulation {#sec:formulation} ------------------- We focus on predictive inference and control over a time-varying stochastic process, whose mean $f$ is temporally evolving: $$\begin{aligned} {\label{eq:e:bnp_model1}} & f_{k+1}& \sim \mathbb{F}(f_k,\eta_k) $$ where $\mathbb{F}$ is a distribution varying with time $t$ and exogenous inputs $\eta$. The theory of reproducing kernel Hilbert spaces (RKHSs) provides powerful tools for generating flexible classes of functions with relative ease, and is thus a natural choice for modeling complex spatial functions [@scholkopf:bk:2002]. Therefore, our focus will be on spatiotemporally evolving kernel-based models, such as Gaussian Processes (GPs). In a kernel-based model, ${k}:{\Omega}\times{\Omega}\to{\mathbb{R}}$ is a positive definite kernel on some compact domain ${\Omega}$ that models the covariance between any two points in the input space. A Mercer kernel [@scholkopf:bk:2002] implies the existence of a smooth map ${\psi}:{\Omega}\to{\mathcal{H}}$, where ${\mathcal{H}}$ is an RKHS with the property $$\begin{aligned} {k}(x,y) &= {\langle {\psi}(x),{\psi}(y) \rangle_{{\mathcal{H}}}} = {\langle {\psi}({k}(x,\cdot)),{\psi}({k}(y,\cdot)) \rangle_{{\mathcal{H}}}}. \end{aligned}$$ There is a large body of literature on modeling spatiotemporal evolution in ${\mathcal{H}}$ [@wikle2002kernel; @cressie2011statistics]. A simple approach for spatiotemporal modeling is to utilize both spatial and temporal variables as inputs to the kernel [@perez:13:gaussian; @Chowdhary13_ACC2]. However, this technique leads to an ever-growing kernel dictionary, which is computationally taxing. Furthermore, constraining the dictionary size or utilizing a moving window will occlude the learning of long-term patterns. Periodic or nonstationary covariance functions and nonlinear transformations have been proposed to address this issue [@ma2003nonstationary; @RasmussenWilliams2005]. Furthermore, work in the design of nonseparable and nonstationary covariance kernels seeks to design kernels optimized to environment-specific dynamics, and optimize their hyperparameters in local regions of the input space [@garg2012AAAI; @das2014nonstationary; @plagemann2008nonstationary]. The model of spatiotemporal functional evolution proposed in this paper builds on the idea that modeling the temporal evolution of mixing weights of a kernel model is a valid approach to spatiotemporal modeling. The key idea behind our approach is that the spatiotemporal evolution of a kernel-based model can be directly modeled by tracing the evolution of the mean embedded in a RKHS using switched ordinary differential equations (ODE) when the evolution is continuous, or switched difference equations when it is discrete (Figure \[fig:hilbert\_evolution\]). The advantage of this approach is that it allows us to utilize powerful ideas from systems theory for knowing necessary conditions for functional convergence; furthermore, it offers a natural framework for designing control mechanisms as well. ![Two types of Hilbert space evolutions. Left: the model, represented by the functions $m_i$, switches discretely in the Hilbert space ${\mathcal{H}}$; Right: the evolution of the function $m_t$ is smooth, represented by a solution to an ordinary differential equation in ${\mathcal{H}}$.[]{data-label="fig:hilbert_evolution"}](model.png){width="0.8\columnwidth"} In this paper, we restrict our attention to the class of functional evolutions $\mathbb{F}$ defined by linear Markovian transitions in an RKHS. While extension to the nonlinear case is possible (and non-trivial), it is not pursued in this paper to help ease the exposition of key ideas. Let $y\in{\mathbb{R}}^{{N}}$ be the measurements of the function available from ${N}$ sensors, ${\mathcal{A}}:{\mathcal{H}}\to{\mathcal{H}}$ be a linear transition operator in the RKHS ${\mathcal{H}}$, and ${\mathcal{K}}:{\mathcal{H}}\to{\mathbb{R}}^{{N}}$ be a linear measurement operator, the model for the infinite-dimensional functional evolution and measurement studied in this paper is: $$\begin{aligned} {\label{eq:ideal_lin_evol}} f_{k+1} &= {\mathcal{A}}f_k + \eta_k\\ y_k &= {\mathcal{K}}f_k + \zeta_k,\end{aligned}$$ where $\eta_k$ is a zero-mean stochastic process in ${\mathcal{H}}$, and $\zeta_k$ is a Wiener process in ${\mathbb{R}}^{{N}}$. For many kernels, the feature map ${\psi}$ is unknown, and therefore it is necessary to work in the dual space of ${\mathcal{H}}$. For concreteness, we work with an approximate space as follows: given points ${\mathcal{C}}= {\{c_1,\dots,c_{{M}}\}}$, $c_i\in{\Omega}$, we have a dictionary of atoms ${\mathcal{F}_{{C}}}= \begin{bmatrix}{\psi}(c_1) &\cdots & {\psi}(c_{{M}}) \end{bmatrix}$, ${\psi}(c_i)\in{\mathcal{H}}$, the span of which is a strict subspace of the RKHS generated by the kernel. Formally, we have $$\begin{aligned} {\mathcal{C}}\mapsto {\ensuremath{{\mathcal{H}}_{{C}}}}:= \operatorname{span}\begin{bmatrix}{\psi}(c_1) &\cdots & {\psi}(c_{{M}}) \end{bmatrix} \subset {\mathcal{H}}.\end{aligned}$$ This regime, which trades off the flexibility of a truly nonparametric approach for computational realizability, still allows for the representation of rich phenomena. Let ${N}$ represent the number of sampling locations, and ${M}$ be the number of bases generating ${\ensuremath{{\mathcal{H}}_{{C}}}}$. Note that every function $f\in{\ensuremath{{\mathcal{H}}_{{C}}}}$ has an expansion of the form $$\begin{aligned} {\label{eq:expansion}} f(x) = \sum_{i=1}^{{M}} w_i k(c_i,x).\end{aligned}$$ This expansion allows us to write the $w_i$ coordinates in the dual space as vectors $w\in{\mathbb{R}}^{{M}}$. We can show the relation of the function spaces to their Euclidean counterparts via commutative diagrams. Define ${\ensuremath{\mathcal{W}}}:{\ensuremath{{\mathcal{H}}_{{C}}}}\to{\mathbb{R}}^{{M}}$ as the operator that maps the coordinates $w_i$ in [(\[eq:expansion\])]{} to vectors ${w}\in{\mathbb{R}}^{{M}}$, and let ${\ensuremath{\mathcal{W}^{-1}}}:{\mathbb{R}}^{{M}}\to{\ensuremath{{\mathcal{H}}_{{C}}}}$. Note that for finite-dimensional spaces, this inverse map always exists. These definitions allow us to outline the relations between the dynamics operators ${\mathcal{A}}$ and $A$, and the measurement operators ${\mathcal{K}}$ and ${\ensuremath{K}}$ using the commutative diagrams in Figure \[fig:commute\_sysop\_dyn\] and Figure \[fig:commute\_sysop\_meas\] respectively. The finite-dimensional evolution equations equivalent to [(\[eq:ideal\_lin\_evol\])]{} in the dual space can be formulated as $$\begin{aligned} {w}_{k+1} &= A{w}_k + \eta_k {\label{eq:k_measure}}\\ y_{k} &= {\ensuremath{K}}_{k} w_{k} +\zeta_k, {\label{eq:k_measure1}}\end{aligned}$$ where we have matrices $A\in {\mathbb{R}}^{{M}\times{M}}, \ {\ensuremath{K}}_{k}\in {\mathbb{R}}^{{N}\times{M}}$, the vectors ${w}_k, {w}\in{\mathbb{R}}^{{M}}$, and we have slightly abused notation to let $\eta_k$ and $\zeta_k$ denote their ${\ensuremath{{\mathcal{H}}_{{C}}}}$ counterparts. Note that the measurement operator ${\mathcal{K}}$ is simply a sampling of the function $f$ at an arbitrary set of sensing locations ${\mathcal{X}}= {\{x_1, \dots, x_{{N}}\}}$, where $x_i\in{\Omega}$: we will see how this affects the structure of ${\ensuremath{K}}_k$ momentarily. The equations [(\[eq:ideal\_lin\_evol\])]{} suggest an immediate extension to functional control problems. Pick another dictionary of atoms ${\mathcal{F}_{{D}}}= \begin{bmatrix}{\psi}(d_1) &\cdots & {\psi}(d_{{\ell}}) \end{bmatrix}$, ${\psi}(d_j)\in{\mathcal{H}}$, $d_j\in{\Omega}$, the span of which, denoted by ${\ensuremath{{\mathcal{H}}_{{D}}}}$, is a strict subspace of the RKHS ${\mathcal{H}}$ generated by the kernel. The functional evolution equation is then as follows: $$\begin{aligned} {\label{eq:ideal_lin_evol_control}} f_{k+1} &= {\mathcal{A}}f_k + {\mathcal{B}}{\delta}_k + \eta_k\\ y_k &= {\mathcal{K}}_k f_k + \zeta_k,\end{aligned}$$ where the control functions ${\delta}_k$ evolve in ${\ensuremath{{\mathcal{H}}_{{D}}}}$, and ${\mathcal{B}}:{\ensuremath{{\mathcal{H}}_{{D}}}}\to{\ensuremath{{\mathcal{H}}_{{C}}}}$. To derive the finite-dimensional equivalent of ${\mathcal{B}}$, we have to work out the structure of the matrix $B$: since ${\ensuremath{{\mathcal{H}}_{{C}}}}$ is not, in general, isomorphic to ${\ensuremath{{\mathcal{H}}_{{D}}}}$, this imposes strict restrictions on $B$. We derive $B$ using least squares using the inner product of ${\mathcal{H}}$. Let ${\delta}= \sum_{j=1}^{{\ell}}{\acute{{w}}}_j{k}(d_j,x)$, and let ${\mathcal{F}_{{C}}}= \begin{bmatrix}{\psi}(c_1) &\cdots & {\psi}(c_{{M}}) \end{bmatrix}$ be the basis for ${\ensuremath{{\mathcal{H}}_{{C}}}}$. Then the projection of $\delta$ onto ${\ensuremath{{\mathcal{H}}_{{C}}}}$ can be derived as $$\begin{aligned} \begin{bmatrix} \l \delta, {\psi}(c_1) \r_{{\mathcal{H}}}\\ \vdots\\ \l \delta, {\psi}(c_{{M}}) \r_{{\mathcal{H}}} \end{bmatrix} &= \underbrace{ \begin{bmatrix} {k}(d_1,c_1) & \cdots & {k}(d_{{\ell}},c_1)\\ \vdots &\ddots &\vdots\\ {k}(d_1,c_{{M}}) & \cdots & {k}(d_{{\ell}},c_{{M}}) \end{bmatrix}}_{{{\ensuremath{K}}_{{C}{D}}}} \begin{bmatrix} {\acute{{w}}}_1\\ \vdots\\ {\acute{{w}}}_{{\ell}} \end{bmatrix},\end{aligned}$$ using the reproducing property. This derivation shows that the operator $B = {{\ensuremath{K}}_{{C}{D}}}\in{\mathbb{R}}^{{M}\times{\ell}}$, the kernel matrix between the data ${C}$ generating the atoms ${\mathcal{F}_{{C}}}$ of ${\ensuremath{{\mathcal{H}}_{{C}}}}$ and the data ${D}$ generating the atoms ${\mathcal{F}_{{D}}}$ of ${\ensuremath{{\mathcal{H}}_{{D}}}}$. Using similar arguments, it can be shown that, given sensing locations ${X}= \{x_1,x_2,\dots, x_{{N}}\}$, ${{\ensuremath{K}}_{{D}}}\in{\mathbb{R}}^{{N}\times{\ell}}$ is the kernel matrix between ${X}$ and ${D}$. Thus the finite-dimensional evolution equations equivalent to [(\[eq:ideal\_lin\_evol\_control\])]{} are $$\begin{aligned} {w}_{k} &= A{w}_k + {{\ensuremath{K}}_{{C}{D}}}{\acute{{w}}}_k{\label{eq:k_measure_c}}\\ y_{k} &= {\ensuremath{K}}_{k} w_{k} {\label{eq:k_measure_c1}}.\end{aligned}$$ We pause here to point out just how flexible the kernel-based framework is. First of all, the choice of kernel completely determines the space ${\mathcal{H}}$, which may allow wildly different functional outputs for the same dynamics matrix, as shown in Figure \[fig:kernel\_variation\]. Note also that the dynamical equations [(\[eq:k\_measure\_c\])]{} and [(\[eq:k\_measure\_c1\])]{} are *independent of the choice of domain ${\Omega}$*: different domains with different kernels may result in the same sequence of matrices ${\ensuremath{K}}_k$. This allows our results to hold for any domain over which a kernel can be defined, including examples like graphs, hidden Markov models, and strings, which are not typically studied in the controls literature, at virtually no extra complexity in implementation beyond the design of the actual sensors and actuators. This remarkable fact is why we denote our method to be *domain agnostic*. \ Since ${\ensuremath{K}}_{k+1}$ is the kernel matrix between the data points and basis vectors, its rows are of the form ${\ensuremath{K}}_{(i)} = \begin{bmatrix} {k}(x_i, c_1) & {k}(x_i, c_2) & \cdots & {k}(x_i, c_{{M}}) \end{bmatrix}$. In systems-theoretic language, each row of the kernel matrix corresponds to a *measurement* at a particular location, and the matrix itself acts as a measurement operator. We define the *generalized observability matrix* [@zhou:bk:96] as $$\begin{aligned} {\label{eq:obs_mat}} {\mathcal{O}}_{{\Upsilon}} = \begin{bmatrix} {\ensuremath{K}}_{t_1} A^{t_1}\\ \cdots\\ {\ensuremath{K}}_{t_{L}} A^{t_{L}} \end{bmatrix},\end{aligned}$$ where ${\Upsilon}= \{t_1, t_2, \dots, t_{{L}}\}$ are the set of instances $t_i$ when we apply the measurement operators ${\ensuremath{K}}_{t_i}$. Note that ${\mathcal{O}}_{{\Upsilon}}\in{\mathbb{R}}^{{N}{L}\times{M}}$. Similarly, we can define the *generalized controllability matrix* as $$\begin{aligned} {\label{eq:control_mat}} {\Psi}_{{\Upsilon}} = \begin{bmatrix} {A^{t_1}}^T {{{\ensuremath{K}}_{{D}}}}_{t_1} & {A^{t_2}}^T {{{\ensuremath{K}}_{{D}}}}_{t_2} & \cdots {A^{t_{L}}}^T {{{\ensuremath{K}}_{{D}}}}_{t_{L}} \end{bmatrix},\end{aligned}$$ ${\Psi}_{{\Upsilon}}\in{\mathbb{R}}^{{M}\times{L}{\ell}}$ A linear system is said to be observable if ${\mathcal{O}}_{{\Upsilon}}$ has full column rank (i.e. $\mathrm{Rank}~{\mathcal{O}}_{{\Upsilon}}={M}$) and is controllable if ${\Psi}_{{\Upsilon}}$ has full row rank, for ${\Upsilon}= \{0, 1, \dots, {M}-1\}$ [@zhou:bk:96]. Observability guarantees that a feedback-based observer can be designed such that the estimate of $w$ denoted by $\hat {w_k}$ converges exponentially fast to the true state $w_k$. In particular, observability is the necessary condition for the existence of a unique solution to the Riccatti equation required in designing a Kalman filter. Therefore, when $\eta,\zeta$ have a zero mean Gaussian distribution, a Bayes optimal filter can be designed for estimating $w$ if and only if $\mathrm{Rank}~{\mathcal{O}}_{{\Upsilon}}={M}$. Similarly, controllability guarantees that a feedback-based controller can drive the current functional state of the system $f_{k}$ to a reference function $f_{\text{ref}}$, as long as $f_{\text{ref}}\in{\ensuremath{{\mathcal{H}}_{{C}}}}$. We are now in a position to formally state the spatiotemporal monitoring and control problem considered: Given a spatiotemporally evolving system modeled using [(\[eq:ideal\_lin\_evol\_control\])]{}, choose a set of ${N}$ sensing locations ${\mathcal{X}}= {\{x_1, \dots, x_{{N}}\}}$ and ${\ell}$ actuating locations ${\mathcal{D}}= {\{d_1,\dots,d_{{\ell}}\}}$ such that even with ${N}\ll {M}$ and ${\ell}\ll {M}$, the functional evolution of the spatiotemporal model can be estimated robustly, and driven (controlled) to a reference function $f_{\text{ref}}$. Our approach to solve this problem relies on the design of the measurement operator ${\ensuremath{K}}$ such that the pair $(A,{\ensuremath{K}})$ is observable, and the control operator ${{\ensuremath{K}}_{{D}}}$ such that the pair $(A,{{\ensuremath{K}}_{{D}}})$ is controllable. Theoretical Results {#sec:theory_results} ------------------- In this section, we prove results concerning the observability of spatiotemporally varying functions modeled by the functional evolution and measurement equations [(\[eq:k\_measure\])]{} and [(\[eq:k\_measure1\])]{} formulated in Section \[sec:formulation\]. In particular, observability of the system states implies that we can recover the current state of the spatiotemporally varying function using a small number of sampling locations ${N}$, which allows us to 1) track the function, and 2) predict its evolution forward in time. It should be noted that the results are also applicable to controllability of the system in [(\[eq:k\_measure\_c1\])]{} since the structure of the control matrix $K_{CD}$ is also that of a Kernel matrix. We first show in Proposition \[prop:1\] that if $A$ has a full-rank Jordan decomposition, the kernel matrix meeting a condition called *shadedness* (to be defined below) is sufficient for the system to be observable. In Proposition \[prop:2\], we prove a lower bound on the number of sampling locations required for observability which holds for more general $A$. Finally, in Proposition \[prop:3\], we outline a method that achieves this lower bound for certain kernels. Since both ${\ensuremath{K}}$ and ${{\ensuremath{K}}_{{C}{D}}}$ are kernel matrices generated from a shared kernel, these observability results translate directly into controllability results. To prove our results, we will leverage the spectral decomposition of $A$. Specifically, recall that any matrix $A\in{\mathbb{R}}^{{M}\times{M}}$ is similar to a unique block diagonal matrix ${\Lambda}$ (i.e. $\exists P\in{\mathbb{R}}^{{M}\times{M}}$ invertible such that $A = P{\Lambda}P^{-1}$) whose diagonal blocks are matrices of the form $$\begin{aligned} {\Lambda}_k(\lambda_i, \lambda_i^*) := \begin{bmatrix} {M}& {I_2}& \cdots & {0}\\ \vdots & \vdots & \ddots & {I_2}\\ {0}& {0}& \cdots & {M}{\label{eq:jor_com}} \end{bmatrix}.\end{aligned}$$ where $(\lambda_i, \lambda_i^*)$ is a complex conjugate eigenvalue of $A$, and $ {M}= \begin{bmatrix} \mu_1 & \mu_2\\ -\mu_2 & \mu_1 \end{bmatrix}$ and $ {I_2}= \begin{bmatrix} 1 & 0\\ 0 & 1 \end{bmatrix}$. Real eigenvalues ${\lambda}_i$ correspond to the case ${M}= {\lambda}_i$ and ${I_2}= 1$. Thus the complete real Jordan form of $A$ will be the appropriate diagonal array of these blocks. If all the eigenvalues $\lambda_i$ are nonzero and real, we say the matrix has a *full-rank Jordan decomposition*. \[def:shaded\] **(Shaded Kernel Matrix)** Let ${k}:{\Omega}\times{\Omega}\to{\mathbb{R}}$ be a positive-definite kernel on a compact domain ${\Omega}$. Let $C = [c_1, c_2, \cdots , c_{{M}}\}$, $c_j\in{\Omega}$ be the points generating a finite-dimensional covering of the reproducing kernel Hilbert space ${\mathcal{H}}$ associated to ${k}(x,y)$, and let ${\mathcal{X}}= {\{x_1, \dots, x_{{N}}\}}$, $x_i\in{\Omega}$ Let ${\ensuremath{K}}\in{\mathbb{R}}^{{N}\times{M}}$ be the kernel matrix, where $ {\ensuremath{K}}_{ij} := {k}(x_i,c_j)$. For each row ${\ensuremath{K}}_{(i)} := [k(x_i,c_1), k(x_i,c_2),\dots, k(x_i,c_{{M}})]$, define the set ${\mathcal{I}}_{(i)} := \{\iota_1^{(i)},\iota_2^{(i)},\dots, \iota_{{M}_i}^{(i)}\}$ to be the indices in the kernel matrix row $i$ which are nonzero. Then if $$\begin{aligned} {\label{eq:shaded_cond}} \bigcup_{1 \leq i \leq {N}} {\mathcal{I}}^{(i)} = \{1,2,\dots, {M}\}, \end{aligned}$$ we denote ${\ensuremath{K}}$ as a *shaded kernel matrix* (see figure \[shaded\_matrix\]). This condition implies that the null space of the adjoint of ${\ensuremath{K}}$ as a linear operator between Euclidean spaces, i.e. ${\ensuremath{K}}^T:{\mathbb{R}}^{{N}}\to{\mathbb{R}}^{{M}}$ is trivial. Note that, in principle, for the Gaussian kernel, a single row generates a shaded kernel matrix, although this matrix can have many entries that are extremely close to zero. With this definition in place, we can prove the following proposition, which shows that if $A$ has a full-rank Jordan decomposition, a shaded kernel matrix is sufficient to prove observability. \[prop:1\] Let ${k}:{\Omega}\times{\Omega}\to{\mathbb{R}}$ be a positive definite kernel on a domain ${\Omega}$. Let $C = [c_1, c_2, \cdots , c_{{M}}\}$, $c_j\in{\Omega}$ be the points generating a finite-dimensional covering of the reproducing kernel Hilbert space ${\mathcal{H}}$ associated to ${k}(x,y)$, and consider the discrete linear system on ${\mathcal{H}}$ given by the evolution and measurement equations [(\[eq:k\_measure\])]{} and [(\[eq:k\_measure1\])]{}. Let $A\in{\mathbb{R}}^{{M}\times{M}}$ be a full-rank Jordan decomposition of the form $A = {P}{\Lambda}{P}^{-1}$, where ${\Lambda}= \operatorname{diag}(\begin{bmatrix}{\Lambda}_1 & {\Lambda}_2 &\cdots & {\Lambda}_{{O}}\end{bmatrix})$, and there are no repeated eigenvalues. Given a set of time instances ${\Upsilon}= \{t_1,t_2,\dots,t_{{L}}\}$, and a set of sampling locations ${\mathcal{X}}={\{x_1, \dots, x_{{N}}\}}$, the system [(\[eq:k\_measure\])]{} is observable if the kernel matrix ${\ensuremath{K}}_{ij} := {k}(x_i,c_j)$ is shaded, ${\ensuremath{K}}^D$, the row vector generated by summing the rows of ${\ensuremath{K}}$, has all nonzero entries, ${\Upsilon}$ has distinct values, and $|{\Upsilon}| \geq {M}$. To begin, consider a system where $A = {\Lambda}$, with Jordan blocks $\{{\Lambda}_1, {\Lambda}_2, \dots, {\Lambda}_{{O}}\}$ along the diagonal. Then $A^{t_i} = \operatorname{diag}(\begin{bmatrix}{\Lambda}_1^{t_i} & {\Lambda}_2^{t_i} & \cdots & {\Lambda}_{{O}}^{t_i}\end{bmatrix})$. We have that $${\mathcal{O}}_{{\Upsilon}} = \begin{bmatrix} {\ensuremath{K}}A^{t_1}\\ \cdots\\ {\ensuremath{K}}A^{t_{L}} \end{bmatrix} = \underbrace{ \begin{bmatrix} {\ensuremath{K}}& \cdots & {\ensuremath{K}}\end{bmatrix}}_{{\widehat{\mathbf{K}}}\in{\mathbb{R}}^{{N}\times{M}{L}}} \underbrace{ \begin{bmatrix} {\Lambda}_1^{t_1} & \cdots & 0\\ \vdots & \ddots & \vdots\\ 0 & \cdots & {\Lambda}_{{O}}^{t_1}\\ \hline \vdots & \ddots & \vdots\\ \hline {\Lambda}_1^{t_{{L}}} & \cdots & 0\\ \vdots & \ddots & \vdots\\ 0 & \cdots & {\Lambda}_{{O}}^{t_{L}} \end{bmatrix}}_{{\widehat{\mathbf{A}}}\in{\mathbb{R}}^{{M}{L}\times{M}}}$$ Recall that a matrix’s rank is preserved under a product with an invertible matrix. Design a matrix $U\in{\mathbb{R}}^{{N}\times{N}}$ s.t. ${\widetilde{{\ensuremath{K}}}}:= U{\ensuremath{K}}$ is a matrix with one row vector of nonzeros, and all of the remaining rows as zeros. Then $\operatorname{rank}({\widehat{\mathbf{K}}}{\widehat{\mathbf{A}}}) = \operatorname{rank}(U{\widehat{\mathbf{K}}}{\widehat{\mathbf{A}}})$. Therefore, we have that $${\widetilde{{\ensuremath{K}}}}A^{t_j} = \begin{bmatrix} {\widetilde{{\ensuremath{K}}}}_{(1)}\\ 0\\ \vdots\\ 0 \end{bmatrix} A^{t_j}\\ = \begin{bmatrix} k_{11}{\lambda}_1^{t_j} & \binom{t_j}{1}{\lambda}_1^{t_j-1} + k_{12}{\lambda}_1^{t_j} & \cdots & k_{1{M}}{\lambda}_{{O}}^{t_j}\\ 0 & 0 & \cdots & 0\\ \vdots & \vdots & \ddots & 0\\ 0 & 0 & \cdots & 0 \end{bmatrix}$$ Therefore, following some more elementary row operations encoded by $V\in{\mathbb{R}}^{{M}{L}\times{M}{L}}$, we get that $$V \begin{bmatrix} {\widetilde{{\ensuremath{K}}}}& \cdots & {\widetilde{{\ensuremath{K}}}}\end{bmatrix} \begin{bmatrix} A^{t_1}\\ \vdots\\ A^{t_{{L}}} \end{bmatrix} = \begin{bmatrix} \tilde{k}_{11}{\lambda}_1^{t_1} & \cdots & \tilde{k}_{1{M}}{\lambda}_{{O}}^{t_1}\\ \tilde{k}_{11}{\lambda}_1^{t_2} & \cdots & \tilde{k}_{1{M}}{\lambda}_{{O}}^{t_2}\\ \vdots & \ddots & 0\\ \tilde{k}_{11}{\lambda}_1^{t_{{L}}} & \cdots & \tilde{k}_{1{M}}{\lambda}_{{O}}^{t_{{L}}}\\ \mathbf{0} & \cdots & \mathbf{0} \end{bmatrix}\\ = \begin{bmatrix} \boldsymbol{\Phi}\\ \widehat{\mathbf{0}} \end{bmatrix}.$$ If the individual entries $\tilde{k}_{1i}$ are nonzero, and the Jordan block diagonals have nonzero eigenvalues, the columns of $\boldsymbol\Phi$ become linearly independent. Therefore, if ${L}\geq {M}$, the column rank of ${\mathcal{O}}_{{\Upsilon}}$ is ${M}$, which results in an observable system. To extend this proof to matrices $A = {P}{\Lambda}{P}^{-1}$, note that $${\mathcal{O}}_{{\Upsilon}} = \begin{bmatrix} {\ensuremath{K}}A^{t_1}\\ \cdots\\ {\ensuremath{K}}A^{t_{L}} \end{bmatrix} = \begin{bmatrix} {\ensuremath{K}}{P}{\Lambda}^{t_1}{P}^{-1}\\ \cdots\\ {\ensuremath{K}}{P}{\Lambda}^{t_{L}}{P}^{-1}. \end{bmatrix} = \begin{bmatrix} {\ensuremath{K}}& \cdots & {\ensuremath{K}}\end{bmatrix} \boldsymbol{{P}} \boldsymbol{{\Lambda}}^t \boldsymbol{{P}^{-1}},$$ where $\boldsymbol{{P}}\in{\mathbb{R}}^{{M}{L}\times{M}{L}}$, $\boldsymbol{{\Lambda}}^t\in{\mathbb{R}}^{{M}{L}\times{M}{L}}$, and $\boldsymbol{{P}^{-1}}\in{\mathbb{R}}^{{M}{L}\times{M}{L}}$ are the block diagonal matrices associated with the system. Since $\boldsymbol{{P}}$ is an invertible matrix, the conclusions about the column rank drawn before still hold, and the system is observable. When the eigenvalues of the system matrix are repeated, it is not enough for ${\ensuremath{K}}$ to be shaded. The next proposition proves a lower bound on the number of observations required. \[prop:2\] Suppose that the conditions in Proposition \[prop:1\] hold, with the relaxation that the Jordan blocks $\begin{bmatrix}{\Lambda}_1 & {\Lambda}_2 &\cdots & {\Lambda}_{{O}}\end{bmatrix}$ may have repeated eigenvalues. Let ${r}$ be the number of unique eigenvalues of $A$, and let ${\gamma}({{\lambda}}_i)$ denote the geometric multiplicity of eigenvalue ${{\lambda}}_i$. Then there exist kernels ${k}(x,y)$ such that the lower bound ${l}$ on the number of sampling locations ${N}$ is given by the cyclic index of $A$, which can be computed as $$\begin{aligned} {\label{eq:geom_mult}} {l}= \max_{1\leq i\leq{r}}{\gamma}({{\lambda}}_i). \end{aligned}$$ We first prove the lower bound. WLOG, let ${\mathbf{{\ensuremath{K}}}}$ have ${l}-1$ fully shaded, linearly independent rows, and write it as $$\begin{aligned} {\mathbf{{\ensuremath{K}}}}&= \begin{bmatrix} k_{11} & k_{12} & \cdots & k_{1{M}} \\ \vdots & \vdots & \cdots & \vdots \\ k_{({l}-1)1} & k_{({l}-1)2} & \cdots & k_{({l}-1){M}} \end{bmatrix}. \end{aligned}$$ Since the cyclic index is ${l}$, this implies that at least one eigenvalue, say ${{\lambda}}$, has ${l}$ Jordan blocks. Define indices $j_1, j_2, \dots, j_{{l}} \in \{1,2,\dots,{M}\}$ as the columns corresponding to the leading entries of the ${l}$ Jordan blocks corresponding to ${{\lambda}}$. WLOG, let $j_1 = 1$. Using ideas similar to the last proof, we can write the observability matrix as $$\begin{aligned} {\mathcal{O}}_{{\Upsilon}} &:= \begin{bmatrix} {k}_{11}{{\lambda}}^{t_1} & \cdots & {k}_{1j_{{l}}}{{\lambda}}^{t_1} & \cdots\\ \vdots & \ddots &\vdots & \ddots\\ {k}_{11}{{\lambda}}^{t_{{L}}} & {k}_{1j_{{l}}}{{\lambda}}^{t_{{L}}} & \cdots\\ \vdots & \ddots & \vdots & \ddots\\ {k}_{({l}-1)1}{{\lambda}}^{t_1} \cdots & {k}_{({l}-1)j_{{l}}}{{\lambda}}^{t_1} & \cdots\\ \vdots & \ddots &\vdots & \ddots\\ {k}_{({l}-1)1}{{\lambda}}^{t_{{L}}} & \cdots & {k}_{({l}-1)j_{{l}}}{{\lambda}}^{t_{{L}}} & \cdots \end{bmatrix}. \end{aligned}$$ Define ${\boldsymbol{{{\lambda}}}}:= \begin{bmatrix}{{\lambda}}^{t_1} & {{\lambda}}^{t_2} & \cdots {{\lambda}}^{t_{{L}}}\end{bmatrix}^T$. Then the above matrix becomes $$\begin{aligned} {\mathcal{O}}_{{\Upsilon}} &:= \begin{bmatrix} {k}_{11}{\boldsymbol{{{\lambda}}}}& \cdots & {k}_{1j_2}{\boldsymbol{{{\lambda}}}}& \cdots & {k}_{1j_{{l}}}{\boldsymbol{{{\lambda}}}}& \cdots\\ \vdots & \ddots & \vdots & \ddots &\vdots & \ddots\\ {k}_{({l}-1)1}{\boldsymbol{{{\lambda}}}}& \cdots & {k}_{({l}-1)j_2}{\boldsymbol{{{\lambda}}}}& \cdots & {k}_{({l}-1)j_{{l}}}{\boldsymbol{{{\lambda}}}}& \cdots \end{bmatrix}. \end{aligned}$$ We need to show that one of the columns above can be written in terms of the others. This is equivalent to solving the linear system $$\begin{aligned} \begin{bmatrix} {k}_{1j_1}\\ {k}_{2j_1}\\ \vdots\\ {k}_{({l}-1)j_1} \end{bmatrix} &= \begin{bmatrix} {k}_{1j_2} & \cdots & {k}_{1j_{{l}}}\\ {k}_{2j_2} & \cdots & {k}_{2j_{{l}}}\\ \vdots & \ddots & \vdots\\ {k}_{({l}-1)j_2} & \cdots & {k}_{({l}-1)j_{{l}}}\\ \end{bmatrix} \begin{bmatrix} c_1\\ c_2\\ \vdots\\ c_{({l}-1)} \end{bmatrix}. \end{aligned}$$ Suppose the kernel matrix on the RHS is generated from the Gaussian kernel. From [@micchelli1984interpolation], it’s known that every principal minor of a Gaussian kernel matrix is invertible, which implies that ${\mathcal{O}}_{{\Upsilon}}$ cannot be observable. We now prove a sufficient condition for the observability of a system with repeated eigenvalues, but with the condition that the Jordan blocks are trivial. \[prop:3\] Suppose that the conditions in Proposition \[prop:1\] hold, with the relaxation that the Jordan blocks $\begin{bmatrix}{\Lambda}_1 & {\Lambda}_2 &\cdots & {\Lambda}_{{O}}\end{bmatrix}$ may have repeated eigenvalues, and where ${\Lambda}_i$ are single-dimensional. Let ${l}$ be the cyclic index of $A$. We define $$\begin{aligned} {\label{eq:empKShadFull}} {\mathbf{{\ensuremath{K}}}}= \begin{bmatrix} {\ensuremath{K}}^{(1)}\\ \vdots\\ {\ensuremath{K}}^{({l})} \end{bmatrix} \end{aligned}$$ as the *${l}$-shaded matrix* which consists of ${l}$ shaded matrices with the property that any subset of ${l}$ columns in the matrix are linearly independent from each other. Then system [(\[eq:k\_measure\])]{} is observable if ${\Upsilon}$ has distinct values, and $|{\Upsilon}| \geq {M}$. A cyclic index of ${l}$ for this system implies that there exists an eigenvalue ${{\lambda}}$ that’s repeated ${l}$ times. WLOG, let ${\mathbf{{\ensuremath{K}}}}$ have ${l}$ fully shaded, linearly independent rows, and, assume that the column indices corresponding to this eigenvalue are $\{1,2,\dots,{l}\}$. Define ${\boldsymbol{{{\lambda}}}}_i:= \begin{bmatrix}{{\lambda}}_i^{t_1} & {{\lambda}}_i^{t_2} & \cdots {{\lambda}}_i^{t_{{L}}}\end{bmatrix}^T$. Then $$\begin{aligned} {\mathcal{O}}_{{\Upsilon}} &:= \begin{bmatrix} k_{11} {\boldsymbol{{{\lambda}}}}_1 & k_{12} {\boldsymbol{{{\lambda}}}}_2 & \cdots & k_{1{M}} {\boldsymbol{{{\lambda}}}}_{{M}}\\ \vdots & \vdots & \ddots & \vdots\\ k_{{l}1} {\boldsymbol{{{\lambda}}}}_1 & k_{{l}2} {\boldsymbol{{{\lambda}}}}_2 & \cdots & k_{{l}{M}} {\boldsymbol{{{\lambda}}}}_{{M}}. \end{bmatrix} \end{aligned}$$ Let ${\boldsymbol{{{\lambda}}}}_1 = {\boldsymbol{{{\lambda}}}}_2 = \cdots {\boldsymbol{{{\lambda}}}}_{{l}} := {\boldsymbol{{{\lambda}}}}$. Focusing on these first ${l}$ columns of this matrix, this implies that we need to find constants $c_1,c_2,\dots, c_{{l}-1}$ s.t. $$\begin{aligned} \begin{bmatrix} k_{11}\\ \vdots\\ k_{{l}1} \end{bmatrix} &= c_1 \begin{bmatrix} k_{12}\\ \vdots\\ k_{{l}2} \end{bmatrix} + \cdots + c_{{l}-1} \begin{bmatrix} k_{1{l}}\\ \vdots\\ k_{{l}{l}}. \end{bmatrix} \end{aligned}$$ However, these columns are linearly independent by assumption, and thus no such constants exist, implying that ${\mathcal{O}}_{{\Upsilon}}$ is observable. An example of a kernel such that any subset of ${l}$ columns in ${\mathbf{{\ensuremath{K}}}}$ are linearly independent of each other is the Gaussian kernel evaluated on sampling locations ${\{x_1, \dots, x_{{N}}\}}$, where $x_i\in{\Omega}\subset{\mathbb{R}}^d$, and $x_i\neq x_j$. We can reuse Propositions \[prop:1\], \[prop:2\], and \[prop:3\] to prove kernel controllability results, because the structure of the control matrix ${{\ensuremath{K}}_{{C}{D}}}$ in [(\[eq:k\_measure\_c\])]{} is also that of a kernel matrix. Kernel ${k}$, basis points ${\mathcal{C}}$, final time step $T_f$. $1)$ Sample data $\{y^i_k\}_{i=1}^{{N}}$ from $f(x,k)$. $2)$ Estimate ${\widehat{w}}_k$ via standard kernel inference procedure. $3)$ Store weights ${\widehat{w}}_k$ in matrix ${\mathcal{W}}\in{\mathbb{R}}^{{M}\times T_f}$. Infer ${\widehat{A}}$ using method of choice (e.g. matrix least squares). Compute the covariance matrix ${\widehat{B}}$ of the observed weights ${\mathcal{W}}$. estimated transition matrix ${\widehat{A}}$, predictive covariance matrix ${\widehat{B}}$. Kernel ${k}$, basis points ${\mathcal{C}}$, estimated system matrix ${\widehat{A}}$, estimated covariance matrix ${\widehat{B}}$. Compute the cyclic index ${l}$ of ${\widehat{A}}$, and compute [(\[eq:empKShadFull\])]{}, by possibly iterating over ${\mathcal{X}}= {\{x_1, \dots, x_{{N}}\}}$. Use ${\widehat{A}}$, ${\widehat{B}}$, and ${\mathbf{{\ensuremath{K}}}}$ to initialize a state-observer (e.g. Kalman filter (KF)) on ${\ensuremath{{\mathcal{H}}_{{C}}}}$. 1) Sample data $\{y^i_k\}_{i=1}^{{N}}$ from $f(x,k)$. 2) Propagate KF estimate ${\widehat{w}}_{k+1}$ forward to time $t_f$, correct using measurement feedback with $\{y^i_k\}_{i=1}^{{N}}$. 3) Output predicted function $\widehat{f}(x,k+1)$ and predictive covariance of KF. Kernel ${k}$, basis points ${\mathcal{C}}$, estimated system matrix ${\widehat{A}}$, estimated covariance matrix ${\widehat{B}}$, and function $f_{\text{ref}}$ to drive initial function to. (see Algorithm \[alg:egp\_inf\]). Use Jordan decomposition of ${\widehat{A}}$ to obtain control locations ${\mathcal{D}}$, compute kernel matrix ${{\ensuremath{K}}_{{C}{D}}}\in{\mathbb{R}}^{{\ell}\times{M}}$ between ${\mathcal{D}}$ and ${\mathcal{C}}$, and initialize controller (e.g. LQR) utilizing $({\widehat{A}}, {\widehat{B}})$. 1) Sample data $\{y^i_k\}_{i=1}^{{N}}$ from $f(x,k)$. 2) Utilize observer to estimate ${\widehat{w}}_{k+1}$. 3) Use ${\widehat{w}}_{k+1}$ and $f_{\text{ref}}$ as input to controller to get feedback. Experimental Results {#sec:experimental_results} ==================== We report experimental results on controlling synthetic and modeling real-world data. All experiments were performed using MATLAB on a laptop running Ubuntu 14.04 with $8$ GB of RAM, and an Intel core i7 processor. Prediction of global ocean surface temperature ---------------------------------------------- We first analyzed the feasibility of this modeling approach on a large dataset: the $4$ km AVHRR Pathfinder project, which is a satellite monitoring global ocean surface temperature. This data was obtained from the National Oceanographic Data Center. The data consists of longitude-latitude measurements on a 2D domain ${\Omega}\subset[-180,180]\times[-90,90]$; this dataset is challenging, with measurements at over $37$ million coordinates, and several missing pieces of data. The goal was to learn the day and night temperature models $f_k(x,y)\in{\ensuremath{{\mathcal{H}}_{{C}}}}$, where ${\ensuremath{{\mathcal{H}}_{{C}}}}$ was generated using the Gaussian kernel ${k}(x,y) = e^{-(\|x-y\|^2/2{\sigma}^2)}$. We first did a search for the ideal bandwidth ${\sigma}$ for a $304$-dimensional sparse Gaussian process model with a Gaussian kernel. The set of atoms ${\mathcal{F}_{{C}}}$ was determined through a linear independence test based sparsification algorithm [@csato2002sparse]. Once the parameters were chosen, a budgeted GP was learned for each date, resulting in weight vectors ${w}_i, \ i\in\{1,2,\dots,365\}$. We used Algorithm \[alg:egp\_trans\] to infer ${\widehat{A}}$, and applied Algorithm \[alg:egp\_inf\] with ${N}\in \{280,500,1000,2000\}$ chosen randomly in the ${\Omega}$ to track the system state given a random initial condition ${w}_0$. Figures \[fig:pathfinder\_errors\_boxplots\_day\] and \[fig:pathfinder\_errors\_boxplots\] show a comparison of the deviation in percentage of the estimated values from the real data, averaged over all the days. As can be seen, the observer enables the prediction of functional evolution *without needing all the measurements (37 million)*, and performance comparable to sampling over all locations is obtained with sampling only over $2,000$ locations. Note that here, even though the system model is observable at ${N}=280$, since the dynamics are not truly linear in ${\ensuremath{{\mathcal{H}}_{{C}}}}$, we get better performance with more sampling locations. Finally, \[fig:pathfinder\_tr\_times\_boxplots\_day\] and \[fig:pathfinder\_tr\_times\_boxplots\] show that the time required to estimate the state during function tracking with kernel observer are an order of magnitude better than retraining the model every time step (“original” in the figure), with comparable performance. \ Control of a linear PDE ----------------------- We then employed kernel controllers for controlling an approximation to the scalar diffusion equation $u_t = bu_{xx}$ on the domain ${\Omega}=[0,1]$, with $b=0.25$. The solution to this equation is infinite-dimensional, so we chose a kernel ${k}(x,y) = e^{-(\|x-y\|^2/2{\sigma}^2)}$, and a set of atoms ${\mathcal{F}_{{C}}}={\{c_1,\dots,c_{{M}}\}}$, $c_i\in{\Omega}$, with ${M}= 25$ generating ${\ensuremath{{\mathcal{H}}_{{C}}}}$, the space approximating ${\mathcal{H}}$, and another set of atoms ${\mathcal{F}_{{D}}}=\{{\psi}(d_1),\dots,{\psi}(d_{{\ell}})\}$, $d_j\in{\Omega}$, ${\ell}=13$, generating the control space ${\ensuremath{{\mathcal{H}}_{{D}}}}$. The number of, and the location of the observations was chosen to be the same as that of the actuation locations $d_j$. First, tests (not reported here) were conducted to ensure that the solution to the diffusion equation is well approximated in ${\ensuremath{{\mathcal{H}}_{{C}}}}$. Algorithm \[alg:egp\_trans\] was then used to infer ${\widehat{A}}$. Figure \[fig:uncontrolled\_pde\] shows an example of an initial function $f_{\text{init}}$ evolving according to the PDE. A reference function $f_{\text{ref}}\in{\ensuremath{{\mathcal{H}}_{{C}}}}$ was chosen to drive $f_{\text{init}}$ to $f_{\text{ref}}$ under the action of the PDE. Finally, Algorithm \[alg:egp\_control\] was used to control the PDE. Figure \[fig:controlled\_pde\] shows $f_{\text{init}}$ being driven to $f_{\text{ref}}$, while Figure \[fig:controlled\_pde\_error\] shows the absolute value of the error between $f_k $ and $f_{\text{ref}}$ as a function of time. Conclusions =========== In this paper we presented a systems theoretic approach to the problem of modeling, estimating, and controlling complex spatiotemporally evolving phenomena. Our approach focused on developing a predictive model of spatiotemporal evolution by layering a dynamical systems prior over temporal evolution of weights of a kernel model. The resulting model can approximate PDE evolution, while it has the form of a finite state linear dynamical system. The lower bounds on the number of sampling and actuation locations provided in this paper are non-conservative, as such they provide direct guidance in ensuring robust real-world sensor network and actuation matrix design that must also account for fault-tolerance and reliability considerations. [^1]: This work was supported in parts by DOE Award Number DE-FE0012173 and AFOSR Award Number FA9550-14-1-0399. Hassan Kingravi is with Pindrop Security, Harshal Maske, and Girish Chowdhary are with the Distributed Autonomous Systems (DAS) laboratory Oklahoma State University,[{hkingravi@pindropsecurity.com, maske@okstate.edu, girish.chowdhary@okstate.edu}]{}

// // Close icons // -------------------------------------------------- .close { float: right; font-size: ($font-size-base * 1.5); font-weight: $close-font-weight; line-height: 1; color: $close-color; text-shadow: $close-text-shadow; @include opacity(.2); &:hover, &:focus { color: $close-color; text-decoration: none; cursor: pointer; @include opacity(.5); } // [converter] extracted button& to button.close } // Additional properties for button version // iOS requires the button element instead of an anchor tag. // If you want the anchor version, it requires `href="#"`. button.close { padding: 0; cursor: pointer; background: transparent; border: 0; -webkit-appearance: none; }

5th Battalion, Royal Sussex Regiment The 5th (Cinque Ports) Battalion, Royal Sussex Regiment, was an infantry battalion of the British Army. The battalion was part of the Royal Sussex Regiment and existed from 1908 until 1966 when it was disbanded. History The 5th (Cinque Ports) Battalion, Royal Sussex Regiment was raised at Middle Street in Hastings on 1 April 1908 upon the creation of the Territorial Force, formed by the amalgamation of the Volunteer Force and the Yeomanry. First World War The 5th (Cinque Ports) Battalion was mobilised in August 1914, immediately upon the outbreak of World War I. The 1/5th Battalion was assigned to the 2nd Brigade, which included the Regular 2nd Royal Sussex, 1st Division from February to August 1915. On 20 August it transferred to the 48th (1st South Midland) Division as the pioneer battalion of the division, serving on the Western Front. 2/5th Battalion The 2/5th Battalion was raised at Hastings in November 1914 from the 'home service' men of the 1/5th Battalion. Second World War The battalion was mobilised on 1 September 1939, and was initially employed guarding vulnerable points in Sussex. It then moved to Dorset to carry out heavy training for deployment overseas. The HQ Wing seems to have been sent to Cattistock, with the other companies at Toller, Melbury and Evershot. Here they joined with the 2nd and 4th battalions to form the 133rd Infantry Brigade of the 44th (Home Counties) Infantry Division. Battle of France On 3 April 1940, the battalion left Cattistock for Southampton and was then shipped to Cherbourg on the night of the 8/9 April on board SS Amsterdam. The battalion was made up of 29 Officers and 690 other ranks. During the night of the 9/10 April the Bn was then moved on to Vivoin, and did some final training. The Bn then moved to Belleuse, and then after two days marched to St. Pol, with the HQ being set up at Conteville, Somme, with the other companies at Eps, Pas-de-Calais and Hestrus. On 11 May they then moved on to Lillers to perform L or C Duties (Light or Casual Duties) also listed as Guard duties. The unit then moved on to Vichtrat Peteghen (Belgium) via Motor transport to take up defensive duties. After arrival they were told to march back 17 miles into France, then after marching all night were ordered to turn round again and head straight back to where they had come from, so marched 35 miles in 24 hours. On 20 May they moved on to Wortegem to take up defensive duties covering the river Scheldt (Escaut), with trenches on the forward slopes. The battalion started combat on 20 May when they came under shell & mortar fire. On the 22nd the Bn retreated to Coutrai, and then on the 23rd to hospital for incurables at St. Andre (near Lille), 160 casualties were reported here. On 24 May the battalion was sent to Viller Berquin, and on the 25th to Strazeele where they encountered German tanks and took heavy fire. On the 25th/26th they moved on to Rouge Croix, taking a defensive position facing Hazebrouck, then pulled back to Mont des Cats. On the 27th they come under heavy fire from dive bombers and shells. Finally, on the 28th they retreated via Poperinghe, to Bray Dunes from where they were eventually evacuated. The battalion, along with the brigade and 44th Division, were evacuated to England. They spent the next two years on home defence and preparing for the German invasion of England, Operation Sea Lion, which never arrived. 7th (Cinque Ports) Battalion In the spring and summer of 1939 the Territorial Army was ordered to be doubled in size by the duplication of existing units to form a 2nd Line, on a similar basis as in the First World War. As such, the 5th Royal Sussex formed a duplicate unit, to be known as the 7th (Cinque Ports) Battalion, Royal Sussex Regiment which came into existence in June 1939, when the 5th Battalion was almost double its normal strength. The 7th Battalion was commanded by Lieutenant Colonel R. Gethen, with headquarters at Brighton. The 7th (Cinque Ports) Battalion was assigned to 37th Infantry Brigade, alongside 5th Buffs (Royal East Kent Regiment) 6th Royal Sussex, part of 12th (Eastern) Infantry Division. References Sources Further reading Cinque Ports Battalion, The story of the 5th (Cinque Ports) Battalion of the Royal Sussex Regiment, Author Colonel E.A.C. Fazan, ASIN: B0007AMWNG Category:Military units and formations established in 1908 Category:Royal Sussex Regiment Royal Sussex 005 Category:1908 establishments in the United Kingdom

Introduction to bifactor polytomous item response theory analysis. A bifactor item response theory model can be used to aid in the interpretation of the dimensionality of a multifaceted questionnaire that assumes continuous latent variables underlying the propensity to respond to items. This model can be used to describe the locations of people on a general continuous latent variable as well as on continuous orthogonal specific traits that characterize responses to groups of items. The bifactor graded response (bifac-GR) model is presented in contrast to a correlated traits (or multidimensional GR model) and unidimensional GR model. Bifac-GR model specification, assumptions, estimation, and interpretation are demonstrated with a reanalysis of data (Campbell, 2008) on the Shared Activities Questionnaire. We also show the importance of marginalizing the slopes for interpretation purposes and we extend the concept to the interpretation of the information function. To go along with the illustrative example analyses, we have made available supplementary files that include command file (syntax) examples and outputs from flexMIRT, IRTPRO, R, Mplus, and STATA. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jsp.2016.11.001. Data needed to reproduce analyses in this article are available as supplemental materials (online only) in the Appendix of this article.

Q: Printing nested JSON dictionary I have an API and I have problem accessing the contents using PHP. Below is the API structure: "data": { "1076056102": { "achievements": { "armorPiercer": 25, "rare1070": 1, "aimer": 4, "invader": 12, ...// Below is the code I have to try and access it. Any help would be nice thank you. $url = '...'; JSON PATH $data = file_get_contents($url); // put contents in variable $stats = json_decode($data); // decode the JSON feed ?> <html> <table> <tbody> <tr> <th>Achievement</th> </tr> <?php foreach ($stats as $stat) : ?> <tr> <td> <?php echo $stat->armorPiercer; ?> </td> <td> <?php echo $stat->rare1070; ?> </td> <td> <?php echo $stat->aimer; ?> </td> <td> <?php echo $stat->invader; ?> </td> </tr> <?php endforeach; ?> </tbody> </table> A: echo $object->data->{1076056102}->achievements->armorPiercer; will work. However, I'm guessing you won't know those ID's specifically. So, turn it into an associative array rather than a stdClass by passing true to json_decode(): $array = json_decode($json, true); foreach ($array['data'] as $row) { echo $row['achievements']['armorPiercer']; // echoes 25 } Check it out here https://3v4l.org/4W3mm

<?xml version="1.0" encoding="UTF-8"?> <kml xmlns="http://www.opengis.net/kml/2.2" xmlns:gx="http://www.google.com/kml/ext/2.2" xmlns:kml="http://www.opengis.net/kml/2.2" xmlns:atom="http://www.w3.org/2005/Atom"> <Document> <name>Style Map Tests</name> <description>Placemarks using Styles defined in relative KML file</description> <Placemark> <name>Africa 1</name> <styleUrl>KML_styles.kml#sm_square</styleUrl> <Point> <gx:drawOrder>1</gx:drawOrder> <coordinates>14.56324920473027,8.823766947866202,0</coordinates> </Point> </Placemark> <Placemark> <name>Africa 2</name> <styleUrl>KML_styles.kml#sm_circle</styleUrl> <Point> <gx:drawOrder>1</gx:drawOrder> <coordinates>17.40955322880677,9.619815130815919,0</coordinates> </Point> </Placemark> <Placemark> <name>Line 1 with StyleUrl</name> <styleUrl>KML_styles.kml#linesymbology</styleUrl> <LineString> <coordinates> 14.7322444643474,8.741122422265486,0 14.98189150254889,8.876022509822219,0 15.1636208882102,8.614709363012477,0 </coordinates> </LineString> </Placemark> <Placemark> <name>Line 2 with StyleUrl and overrides</name> <styleUrl>KML_styles.kml#linesymbology</styleUrl> <Style> <LineStyle> <width>1</width> </LineStyle> </Style> <LineString> <tessellate>1</tessellate> <coordinates> 14.76679054044666,8.647733316197623,0 14.99402575335166,8.762810855977225,0 15.13946004407134,8.539054532876513,0 </coordinates> </LineString> </Placemark> <Placemark> <name>Polygon 1 with StyleUrl</name> <styleUrl>KML_styles.kml#polygonsymbology</styleUrl> <Polygon> <tessellate>1</tessellate> <outerBoundaryIs> <LinearRing> <coordinates> 15.45930089777413,8.649774762626734,0 15.78812145713398,8.826902630484089,0 15.56553132250632,9.117659207419846,0 15.28843281990799,8.916977130845003,0 15.45930089777413,8.649774762626734,0 </coordinates> </LinearRing> </outerBoundaryIs> </Polygon> </Placemark> <Placemark> <name>Polygon 2 with Style overrides</name> <styleUrl>KML_styles.kml#polygonsymbology</styleUrl> <Style> <LineStyle> <width>1</width> </LineStyle> <PolyStyle> <color>ffffaa7f</color> </PolyStyle> </Style> <Polygon> <tessellate>1</tessellate> <outerBoundaryIs> <LinearRing> <coordinates> 15.99240773337751,8.791408965716004,0 16.2407273758592,8.921982810633205,0 15.97838555087566,9.197063754094813,0 15.75799832823521,9.060468835708274,0 15.99240773337751,8.791408965716004,0 </coordinates> </LinearRing> </outerBoundaryIs> </Polygon> </Placemark> <Placemark> <name>Polygon 3 inline style fill only</name> <Style> <PolyStyle> <color>ff7fffaa</color> <outline>0</outline> </PolyStyle> </Style> <Polygon> <tessellate>1</tessellate> <outerBoundaryIs> <LinearRing> <coordinates> 16.45141663527584,8.99085343604397,0 16.62822777773094,9.105369181051785,0 16.3836231863226,9.367059726191357,0 16.22294327719263,9.238531882204896,0 16.45141663527584,8.99085343604397,0 </coordinates> </LinearRing> </outerBoundaryIs> </Polygon> </Placemark> <Placemark> <name>Polygon 4 inline style outline only</name> <Style> <LineStyle> <color>ffffff00</color> <width>4</width> </LineStyle> <PolyStyle> <color>ff7fffaa</color> <fill>0</fill> </PolyStyle> </Style> <Polygon> <tessellate>1</tessellate> <outerBoundaryIs> <LinearRing> <coordinates> 16.75311822571322,9.148653889350184,0 16.91888588059678,9.219599002061271,0 16.70384488661618,9.49955818101844,0 16.55253261598985,9.408508348830811,0 16.75311822571322,9.148653889350184,0 </coordinates> </LinearRing> </outerBoundaryIs> </Polygon> </Placemark> <Placemark> <name>Polygon 5 with island</name> <styleUrl>KML_styles.kml#polygonsymbology</styleUrl> <Polygon> <tessellate>1</tessellate> <outerBoundaryIs> <LinearRing> <coordinates> 17.156109304262,9.251432883064757,0 17.32967734249761,9.340659686674064,0 17.15439083960273,9.617539674318666,0 16.97897194415396,9.494094316822185,0 17.156109304262,9.251432883064757,0 </coordinates> </LinearRing> </outerBoundaryIs> <innerBoundaryIs> <LinearRing> <coordinates> 17.18499558552173,9.330908287399483,0 17.24428094114781,9.374604608036631,0 17.17390026104265,9.5020304960033,0 17.10654405342015,9.460253752017769,0 17.18499558552173,9.330908287399483,0 </coordinates> </LinearRing> </innerBoundaryIs> </Polygon> </Placemark> </Document> </kml>

Q: Replace ' ' Spaces in a URL with a '%20' - Coding Challenge This was a small coding challenge proposed at my school. The problem is: Given a string (or URL), replace the white-spaces inside of the string with %20, but any spaces on the outside should not be included. For example, input " Mr John Smith " would return "Mr%20John%20Smith". I have completed this challenge successfully using the code below. My question is, is there any way to improve the efficiency? I believe the complexity is currently \$O(2n) = O(n)\$ given the 2 for loops. I do not want to use any libraries or functions like str.replace(). But I'm assuming there is a better way than trimming and then counting the whitespace. public class URL { /** * @description URLify ~ A small method that makes a string with spaces URL Friendly! * @param str * @param length * @return String */ public static String URLify(String str) { str = str.trim(); int length = str.length(); int trueL = length; if(str.contains(" ")) { for(int i = 0; i < length; i++) { if(str.charAt(i) == ' ') { trueL = trueL + 2; } } char[] oldArr = str.toCharArray(); char[] newArr = new char[trueL]; int x = 0; for(int i = 0; i < length; i++) { if(oldArr[i] == ' ') { newArr[x] = '%'; newArr[x+1] = '2'; newArr[x+2] = '0'; x += 3; } else { newArr[x] = oldArr[i]; x++; } } str = new String(newArr, 0, trueL); } return str; } public static void main(String[] args) { String str = " https://google.com/ testing .pdf "; str = URLify(str); System.out.print(str); } } Note: I am looking for any criticism. I would really like to improve my skills in general. A: From a short review: If you are going to write your own replace, then you should create a method for it, so that you can call myReplace( string, target, replacement); Personally I would have checked if args got a String, and only if nothing is provided would I use the hard coded string, that would make it easier to show and test your design No reason to do this in 2 lines: str = URLify(str); System.out.print(str); you can just write System.out.print(URLify(str)); Javadoc commenting looks good (but is not actually true?), otherwise there are zero comments I am pretty you can merge those 2 loops into one if you used a stringbuilder. On the whole the replace algorithm could use work A: Let me give some comments to your code as well, with a refactoring coming out at finally: Building a char array, seems to be wasteful – Why don't you use StringBuilder? I know when reinventing the wheel, there could be some debate as to what you are allowed to use or not. But you're already using str.trim() and str.toCharArray(), so why not StringBuilder? Try to avoid multiple repeated loops – In your code you loop it once to get the new length into the strangly named trueL, this should (if kept) be named newLength or similar. Repeating loops is considered expensive, and should be avoided, if possible. oldArr and newArr? – In general I don't think variable names including array, arr, table, and so on are good names. In most cases you know it's an array due to the presence of the array indexing braces. Maybe originalChars or strChars, could be better names. Good use of length = str.length() – I like that you calculate then length just once, and use that in the for loop. Choose appropriate loops – Albeit the charAt() is slightly faster, I think that in this case I would opt for the for (char currentCharacter : text.toCharArray() ) loop variant. Simply because it clearer conveys what you are working with. In addition it removes the need for some extra temporary variables like x, length and trueL Do early returns, if possible – As Janos mentions, in your code you have most of your code within the if (str.contains(" ")) block. This can be reversed, and you can do an early return without almost doing any work. Include a little more vertical space – A personal nitpick of mine is to add a little more vertical space in the code. Adding an extra newline in front of for loops, and if blocks can greatly enhance the reading and understanding of code. Refactored solution Following all of this advice, gives code like this: public class URL { public static String URLify(String text) { if (!text.contains(" ")) { return text; } // Use urlifiedText for building the result text StringBuilder urlifiedText = new StringBuilder(); // Replace spaces with %20, after trimming leading and trailing spaces for (char currentChar : text.trim().toCharArray()) { if (currentChar == ' ') { urlifiedText.append("%20"); } else { urlifiedText.append(currentChar); } } return urlifiedText.toString(); } public static void main(String[] args) { System.out.println(URLify(" my example text ")); } } Even with some added comments, it still is slightly shorter than your original code, and hopefully understandable. Update: Alternate implementation What if you were not allowed to use any of the str.functions()? No str.toCharArray(), str.trim(), str.contains(), and so on? Here is an implementation using the basic String and StringBuilder functionality. If you wanted to you could also replace the StringBuilder method by concatenating into a new String. (Leaving that part to the reader). A possible implementation could then look like: /** * @description URLifyII ~ Remove leading and trailing spaces, * and replace remaining spaces with %20. * NB! Without using str.trim(), str.contains(), and so on... */ public static String URLifyII(String text) { int startIndex = 0; int endIndex = text.length() - 1; StringBuilder urlifiedText = new StringBuilder(); // Find first non-space character while (text.charAt(startIndex) == ' ' && startIndex < endIndex) { startIndex++; } // Find last non-space character while (text.charAt(endIndex) == ' ' && endIndex >= startIndex) { endIndex--; } // Repeat text, and replace spaces with %20 for (int i=startIndex; i <= endIndex; i++) { if (text.charAt(i) != ' ') { urlifiedText.append(text.charAt(i)); } else { urlifiedText.append("%20"); } } return urlifiedText.toString(); } This solution still only loops through the original text just once, but divided into three different parts. A: The next simplest solution not counting str.replace(...) would have been to use a StringBuilder: for each space append %20, otherwise append the character itself. But I guess you didn't want to do that. You took extra care to avoid unnecessary processing (by checking if the string contains space), and unnecessary allocations (counting the exact storage size needed). But then the trimming kinda defeats the purpose of the rest of the code: in case anything is trimmed, that will involve the allocation of a new string, and a full iteration over the source. To be consistent with the rest of the code, you could have done the trimming yourself. Some minor technical improvements are possible: Since you only use length and trueL when you know the string contains a space, it would be better to declare those variables in the scope of the if block Instead of the long piece of code inside the if (str.contains(" ")) {, you could negate the condition for an early return, and make the code flatter, which is a bit easier to read Instead of a counting loop, a for-each loop is more idiomatic when possible. You could create oldArr before the counting loop, and use it in the loop that counts spaces Instead of computing the strangely named trueL, you could count the number of spaces, and store it in the naturally named spaces Since newArr has the exact size for the result string, you can use the single-parameter constructor of String You could declare x in the initializer of the second for loop, to limit its scope to the loop's body Something like this: public static String urlencode(String str) { str = str.trim(); if (!str.contains(" ")) { return str; } char[] chars = str.toCharArray(); int spaces = 0; for (char c : chars) { if (c == ' ') { spaces++; } } char[] newArr = new char[chars.length + 2 * spaces]; for (int i = 0, x = 0; i < chars.length; i++) { char c = chars[i]; if (c == ' ') { newArr[x] = '%'; newArr[x + 1] = '2'; newArr[x + 2] = '0'; x += 3; } else { newArr[x] = c; x++; } } return new String(newArr); }

Correction added on March 15, 2017, after first online publication: Initials were removed from authors\' names and Timothy Steep was corrected to Timothy Stepp. 1. Introduction {#acm212057-sec-0005} =============== Trigeminal neuralgia (TN) is a neurologic syndrome that presents with spontaneous episodes sever, electric shock‐like pain along the trigeminal nerve dermatome(s). Typical primary treatment strategies consist of medical management with antiseizure medication, surgical intervention such as microvascular decompression, and stereotactic radiosurgery.[1](#acm212057-bib-0001){ref-type="ref"}, [2](#acm212057-bib-0002){ref-type="ref"}, [3](#acm212057-bib-0003){ref-type="ref"}, [4](#acm212057-bib-0004){ref-type="ref"} Historically, gamma knife‐based stereotactic radiosurgery (SRS) has been considered an effective and noninvasive alternative treatment modality associated with minimal toxicity --- particularly in patients with medically and surgically refractory TN or those who are not ideal surgical candidates.[5](#acm212057-bib-0005){ref-type="ref"}, [6](#acm212057-bib-0006){ref-type="ref"}, [7](#acm212057-bib-0007){ref-type="ref"}, [8](#acm212057-bib-0008){ref-type="ref"}, [9](#acm212057-bib-0009){ref-type="ref"} For example, in a multi‐intuitional review of 503 patients with TN who had been treated with gamma knife‐based SRS, 58% of patients achieved complete pain relief and 36% of patients achieved partial pain relief.[8](#acm212057-bib-0008){ref-type="ref"} Linac‐based SRS has become an increasingly popular treatment modality for TN due to technological advancements which have allowed for precise radiation delivery in a fast and effective manner.[10](#acm212057-bib-0010){ref-type="ref"}, [11](#acm212057-bib-0011){ref-type="ref"}, [12](#acm212057-bib-0012){ref-type="ref"}, [13](#acm212057-bib-0013){ref-type="ref"}Recently, many researchers have presented linac‐based SRS treatment outcomes for TN patients which are comparable with gamma knife data.[14](#acm212057-bib-0014){ref-type="ref"}, [15](#acm212057-bib-0015){ref-type="ref"}, [16](#acm212057-bib-0016){ref-type="ref"}, [17](#acm212057-bib-0017){ref-type="ref"}, [18](#acm212057-bib-0018){ref-type="ref"}, [19](#acm212057-bib-0019){ref-type="ref"} Due to the effectiveness of linac‐based SRS for treatment of smaller target such as TN, we sought to present a detailed description of our linac‐based SRS technique as well as report our long‐term clinical outcomes in patients with medically and/or surgical refractory TN. 2. Materials and methods {#acm212057-sec-0006} ======================== 2.A. Patient imaging and frame placement {#acm212057-sec-0007} ---------------------------------------- After obtaining approval from our institutional review board, a retrospective review was conducted consisting of a total of 27 TN patients who had been treated at our institution from 2009 to 2016 using frame‐based, linac‐based SRS. All patients underwent a high‐resolution magnetic resonance imaging (MRI) scan consisting of 1 mm thin slices with T1‐weighted, T2‐weighted, and 3D‐fast imaging employing steady state acquisition (FIESTA) sequences prior to treatment. On the day of radiosurgery treatment, an experienced neurosurgeon placed a BrainLAB stereotactic frame on the patient\'s head after application of a local anesthetic. Depth Helmet bobble[20](#acm212057-bib-0020){ref-type="ref"} measurement was performed for quality assurance of the frame placement and, immediately thereafter, the patient was set up for the planning computerized tomography (CT) simulation which was performed on a 16 slice Phillips Brilliance Big Bore CT Scanner (Phillips, Cleveland, OH) and BrainLAB CT localizer (BrainLab Head&Neck Localization Inc., Heimstetten, Germany). CT simulation images were acquired with 512 × 512 pixels at 0.75 mm slice thickness and 0.75 mm slice spacing following departmental SRS scanning protocol. 2.B. Target delineation and SRS treatment planning {#acm212057-sec-0008} -------------------------------------------------- The MRI was co‐registered with the planning CT image set and an experienced neurosurgeon and radiation oncologist delineated the trigeminal nerve root (TNR), for isocenter placement, using the 3D‐FIESTA MRI sequence. The target was localized to the base of the trigeminal nerve at the junction of nerve entry into Meckel\'s Cave and exit from the brainstem. Organs at risk (OAR) were delineated on the co‐registered MRI and consisted of the following structures: brainstem, optic apparatus (optic chiasm and bilateral optic nerves), eyes and lenses, and temporal lobe of the brain. For each treatment, a seven‐arc plan was devised in iPlan BrainLAB to deliver the single‐fraction prescription dose to the 100% isodose line (IDL), using six MV‐SRS beams (1000 MU/min), and a 4 mm diameter cone size. The treatment plans were optimized in order to minimize brainstem dose as well as avoided beam entry through the eyes. All treatment plans were performed using heterogeneity corrected pencil‐beam algorithm with 1.0 × 1.0 × 1.0 mm^3^ grid sizes for dose calculations. All plans employed a single‐fraction point dose of 80 Gy to the TNR and were forward‐optimized to maintain a maximum TNR point dose of 80 Gy, 40 Gy (50%, IDL) encompassing the TNR diameter, and maximum brainstem point dose less than 16 Gy. One example patient case (right trigeminal patient) of seven‐arc arrangement with associated digitally reconstructed radiograph (DRR) is shown in Fig. [1](#acm212057-fig-0001){ref-type="fig"}. In general, the total average arcing length of 130° (e.g., for right trigeminal nerve, 200 to 330°, clockwise rotation for each arc) was used and the couch separation was chosen from 15 to 35°. Due to the use of orbital avoidance vertex‐arc arrangement, the elliptical dose distribution along the longitudinal direction of TNR (optimized for target coverage) was devised that also reduced dose to brainstem and optic apparatus. {#acm212057-fig-0001} 2.C. Evaluation of dose distribution {#acm212057-sec-0009} ------------------------------------ For all TN SRS treatment plans, a dose--volume histogram (DVH) was generated in the iPlan BrainLAB TPS and subsequently evaluated by an experienced radiation oncologist, neurosurgeon, and medical physicist to ensure acceptable OAR doses were achieved. In addition to maximum dose to brainstem, the dose to 0.5 cc of brainstem was also documented. Dose distributions for an example patient are shown in Fig. [2](#acm212057-fig-0002){ref-type="fig"} and the corresponding DVH is shown in Fig. [3](#acm212057-fig-0003){ref-type="fig"}. {#acm212057-fig-0002} {#acm212057-fig-0003} 2.D. Independent second MU check {#acm212057-sec-0010} -------------------------------- A most commonly used TMR‐based spreadsheet independent MU calculation was devised and clinically implemented for second MU check. An independent MU verification is mandatory for safe and effective delivery of such a complex treatment plan. For the given SRS beam, the TMR‐based spreadsheet calculation takes into account of the 4 mm cone size output factors and independently computes MU on the arc‐by‐arc basis for the approved TN SRS treatment plan. For all patients, on a per‐arc basis, our computed BrainLab iPlan MU matched with TMR‐based spreadsheet calculation within ± 3.0%. 2.E. Machine quality assurance and patient setup {#acm212057-sec-0011} ------------------------------------------------ For the given collimator, couch, and gantry rotations, daily Winston--Lutz (WL) QA tests[10](#acm212057-bib-0010){ref-type="ref"} were performed using a 7.5 mm circular cone and a couch mount with a 5 mm diameter mechanical bearing ball (BB). In our clinic, due to the integration of WL QA procedure with ExacTrac system, ExacTrac system was calibrated before the WL QA with a pair of oblique kilo‐voltage x‐ray images of the BB was acquired and automatic 2D‐to‐3D image registration was performed. The WL QA results were considered acceptable if the 5 mm diameter mechanical BB was conformally encompassed by the 7.5 mm radiation field for every gantry, couch, and collimator angle. On a single strip of Gafchromic film, eight static fields (5 mm BB, with 7.5 mm cone size) with following gantry and couch angles were shot for daily WL QA test (G0, G90, G180, G270; with Couch 0) and (C270, C315, C45, C90; with Gantry 0), respectively. A total of 700 MU/beam was used for WL QA. On each shot, submillimeter coincidence of radiation and mechanical isocenters was maintained at all the times with the use of daily WL QA. In addition to the WL QA, a daily QA check of kilovoltage to megavoltage imaging isocenter coincidence was performed prior to patient setup for TN SRS. All QA procedures were in compliance for radiosurgery treatment delivery including QA for frame placement verification using the depth Helmet bobble measurement.[20](#acm212057-bib-0020){ref-type="ref"} It was ensured that the originality of the frame placement before CT simulation and prior to treatment was within ± 1 mm of reproducibility. For each treatment delivery, patient repositioning was achieved using the Target positioner (TaPo) prints out for isocenter localization with the help of gantry cross‐hair. Microadjustments to the couch mount were made following each change in table angle under the supervision of an experienced medical physicist in order to ensure precise isocentricity of each gantry arc. These microadjustment screws on the couch mount allow us to obtain fine adjustment on the TaPo localization in all three directions (anterior, patient left, and right lateral) as well as rotations. Prior to treatment, onboard cone beam CT imaging was performed to verify stereotactic frame placement, head position, and final isocenter location. From the verification cone beam CT, the isocenter localization errors in the left/right, posterior/anterior, and superior/inferior directions were, on average, 0.1 ± 0.7 mm (ranged, −1.0--1.0 mm), 0.3 ± 0.6 mm (ranged, −1.0--1.0 mm), and 0.3 ± 0.8 mm (ranged, −1.0--1.0 mm), respectively. The mean value of angular couch correction discrepancy was 0.1 ± 0.4° (ranged, −0.7--0.6°). These couch correction discrepancies, however, were not applied for the actual treatment considering that these errors were within the range of uncertainty for CBCT image reconstruction and OBI gantry rotation (within ± 1 mm for translational and ± 0.7° for rotational shifts). Overall purpose of verification CBCT was to conform that if there was any unanticipated huge shifts (≥± 2 mm/2°) have been observed, therefore, patient setup could be reconsidered. 2.F. Patient inclusion, clinical outcome, and toxicity evaluation {#acm212057-sec-0012} ----------------------------------------------------------------- For this review, we included a total of 27 refractory TN (*typical* and *atypical*) patients treated at our single institution between 2009 and 2016. All patients reported here were treated by one radiation oncologist and one neurosurgeon. Clinical response to treatment for all patients was retrospectively evaluated and characterized using the Barrow Neurological Institute (BNI) pain intensity score of I--V (see Table [1](#acm212057-tbl-0001){ref-type="table-wrap"} for detailed description). At each follow‐up visit, patient‐reported clinical outcomes including use of medical therapy, pain relief, and pain frequency was assessed and incorporated to generate patient respective BNI pain intensity scores. Treatment‐related brainstem or temporal lobe toxicity was evaluated by assessing any clinical symptoms of headache, new cranial nerve deficit, new focal neurological deficit, or presence of seizure activity. If available, temporal lobe necrosis was assessed radiographically by follow‐up MRI brain. ###### BNI pain intensity score Score description ------------------- ----------------------------------------------------------------------- I No trigeminal pain, no medications II Occasional trigeminal pain that is well tolerated, no medications III (A--B) Occasional trigeminal pain that requires medications to be controlled IV Some pain that is not adequately controlled with medications V Severe pain/no relief John Wiley & Sons, Ltd 3. Results {#acm212057-sec-0013} ========== 3.A. Patient characteristics {#acm212057-sec-0014} ---------------------------- The detailed descriptions of patient characteristics are listed in Table [2](#acm212057-tbl-0002){ref-type="table-wrap"}. Of the 27 refractory TN patients, 22 (81%) suffered from idiopathic/*typical* TN, while 5 (19%) suffered from secondary/*atypical* TN. Median age was 77 yr (ranged, 46--93 yr). Right to left TN ratio was 18/9. Male to female ratio was 14/13. ###### Characteristics of 27 clinically followed patients who underwent Linac‐based SRS for refractory trigeminal neuralgia Characteristics No. of patients (%) ---------------------------------- --------------------- No. of patients 27 Age (years) Median 77 Range 46--93 Gender Male 14 (52) Female 13 (48) Pain type Idiopathic/*Typical* TN (type 1) 22 (81) Secondary/*Atypical* TN (type 2) 5 (19) Side Right 18 (67) Left 9 (33) John Wiley & Sons, Ltd 3.B. Dosimetric and treatment delivery parameters {#acm212057-sec-0015} ------------------------------------------------- On a per‐patient basis, the total number of delivered MU for all 27 patients who underwent TN SRS is shown in Fig. [4](#acm212057-fig-0004){ref-type="fig"}. In our experience, the mean MU was 19,500 and was fairly standard for all TN patients treated with 80 Gy prescription doses. Knowledge of the average total number of MU is advantageous in that it allows for quick identification of some major errors related to dose calculation --- as would be suggested by a calculated total MU which is well above or below the average value. {#acm212057-fig-0004} In Fig. [5](#acm212057-fig-0005){ref-type="fig"}, we show total beam‐on‐time for all 27 patients included in the study. Our average beam‐on‐time was less than 20 min. Understandably, shorter beam‐on‐time helped for patient comfort and faster delivery. {#acm212057-fig-0005} Figure [6](#acm212057-fig-0006){ref-type="fig"} demonstrates the ability of linac‐based TN SRS to generate optimal clinical treatment plans that minimize dose to the brainstem. The plot on the left illustrates the 0.5 cc brainstem dose distribution (mean value = 3.6 ± 0.4 Gy, ranged, 1.2--4.8 Gy), and the plot on the right illustrates the maximum brainstem dose distribution (mean value = 13.4 ± 2.1 Gy, ranged, 9.4--15.9 Gy). None of the patients in this study demonstrated evidence of cranial nerve deficit or radio‐necrosis of temporal lobe. In addition, due to the use of orbital avoidance arc arrangement, the maximum dose to optic apparatus was effectively minimized (average \<1.2 Gy). Average max‐dose to eyes and lens was 0.3 Gy and 0.2 Gy, respectively. {#acm212057-fig-0006} 3.C. Clinical follow‐up outcomes {#acm212057-sec-0016} -------------------------------- Median overall follow‐up time was 12.5 months (range 1--53 months). Figure [7](#acm212057-fig-0007){ref-type="fig"} depicts patient‐reported pain relief (in terms of change in the BNI pain intensity score) following SRS for the 22 patients treated for *typical* TN. With a median follow‐up of 12.5 months (ranged, 1--45 months) in this subpopulation, 82% of patients responded to treatment (BNI score I--IIIB). Nine patients (41%) achieved complete pain relief with a BNI score of I--II. Another nine patients (41%) showed partial pain reduction with a BNI score of IIIA--IIIB. Four patients (18%) had no response to radiosurgery treatment --- all four having baseline pain of IV--V on BNI scale. For the patients who achieved a response to treatment, the average time to response was 5.5 months (range, immediate to 12 months), and the average duration of response was 13 months (range, 1--53 months). Of the patients who responded to treatment, actuarial pain recurrence free survival rates were approximately 100%, 75%, and 50% at 12 months, 15 months, and 24 months, respectively (see Fig. [8](#acm212057-fig-0008){ref-type="fig"}). These results were generated using the Kaplan--Meier product limit method using SPSS 13.0 statistical software (SPSS Inc., Chicago, IL, USA). {#acm212057-fig-0007} {#acm212057-fig-0008} While excellent response rates were observed in the patients treated for *typical* TN, none of the patients treated for *atypical* TN responded to treatment. Subset analysis of these five *atypical* patients suggested that 3 (60%) patients had some pain that was not adequately controlled with medications (BNI score IV), while 2 (40%) patients had severe pain without any relief at all (BNI score V). Clinically, the major reasons why there was no response to those five patients who underwent for *atypical* TN needs further investigations. 4. Discussion {#acm212057-sec-0017} ============= Using a seven‐arc orbital avoidance arrangement with a 4 mm circular cone size, the maximum dose to the point target was delivered 80 Gy and maximum dose to brainstem never exceeded 16 Gy (20% IDL). At a median follow‐up of 12.5 months, 82% of patients treated for *typical* TN had responded to treatment. Nine patients (41%) had complete pain relief with a BNI score of I--II, while another nine patients (41%) had partial pain relief with a BNI score of IIIA--IIIB. Four patients had no response to radiosurgery treatment, showing that all four patients having baseline pain of IV--V on BNI scale. Actuarial pain recurrence free survival rates for the 22 *typical* TN patients were approximately 100%, 75%, and 50% at 12 months, 15 months and 24 months, respectively. On the other hand, none of the five *atypical* TN patients who underwent linac‐based SRS treatment responded to treatment (BNI score of IV--V). The safety, efficacy, and localization accuracy of linac‐based TN SRS has been studied by several researchers.[10](#acm212057-bib-0010){ref-type="ref"}, [11](#acm212057-bib-0011){ref-type="ref"}, [12](#acm212057-bib-0012){ref-type="ref"}, [13](#acm212057-bib-0013){ref-type="ref"} In our clinical implementation of linac‐based TN SRS, we adhered with those standard clinical protocols and guidelines. Treatment planning procedures and patient outcomes for linac‐based TN SRS has also been reported by many investigators.[14](#acm212057-bib-0014){ref-type="ref"}, [15](#acm212057-bib-0015){ref-type="ref"}, [16](#acm212057-bib-0016){ref-type="ref"}, [17](#acm212057-bib-0017){ref-type="ref"}, [18](#acm212057-bib-0018){ref-type="ref"}, [19](#acm212057-bib-0019){ref-type="ref"} For instance, using a seven‐arc geometry with a 4 mm circular cone size, Richards et al.[17](#acm212057-bib-0017){ref-type="ref"} have shown that overall 75% patient achieved complete pain relief. In their study, 26 patients with medication refractory idiopathic trigeminal neuralgia were treated using an 80 Gy prescription dose and the median follow‐up time was 12 months. In another study with 179 patients, Smith et al.[19](#acm212057-bib-0019){ref-type="ref"} reported dose‐dependent pain control rates using a 90 Gy prescription dose and a 5 mm cone size. In that study, almost 80% of patients had experienced significant pain relief at a median follow‐up of 28.8 months. Despite the excellent/good pain relief results, the 90 Gy prescription dose and 5 mm cone size resulted in more toxicity with almost 49% of patients developing numbness. Utilization of CyberKnife SRS for the treatment of TN has also been reported by many investigators.[21](#acm212057-bib-0021){ref-type="ref"}, [22](#acm212057-bib-0022){ref-type="ref"}, [23](#acm212057-bib-0023){ref-type="ref"} Although the treatment outcomes were similar to linac‐based SRS treatment, the reported beam‐on‐times for CyberKnife treatment were relatively longer (\~45--60 min) compared to the average beam‐on‐time in our study (\~20 min). Our prescription dose was 80 Gy to all patients. By identifying the optimal arc arrangement with a 4 mm circular cone size and differential arc weighting, the dose to the brainstem was minimized such that the average maximum brainstem point dose was less than 16 Gy (20% IDL). There was no posttreatment toxicity such as cranial nerve deficit, numbness, or brain necrosis observed in our study. In summary, we have presented our faster and robust technique for the treatment of TN utilizing linac‐based SRS. Our overall clinical outcomes suggest that this technique is both safe (less radiation‐induced toxicity) and effective for patients with *typical* TN SRS treatment. However, the lack of success rates for those patients who underwent for *atypical* TN SRS (similar results were presented by Smith et al.[19](#acm212057-bib-0019){ref-type="ref"}) needs further investigations. Due to the advanced of image‐guidance system, linac‐based TN SRS with frameless radiosurgery setup[24](#acm212057-bib-0024){ref-type="ref"}, [25](#acm212057-bib-0025){ref-type="ref"}, [26](#acm212057-bib-0026){ref-type="ref"} merits further investigation. 5. Conclusion {#acm212057-sec-0018} ============= In this paper, we have presented our linac‐based SRS treatment procedure for TN and the corresponding clinical outcomes. Our overall response rate was 82% in patients with *typical* TN with half of those patients achieving complete pain relief. Of the patients who responded to treatment, actuarial pain recurrence free survival rates were approximately 100%, 75%, and 50% at 12 months, 15 months, and 24 months, respectively. None of the *atypical* TN patients included in this study had a response to treatment. However, there was no treatment‐related neurological toxicity observed in this study. Longer follow‐up of these patients is anticipated to confirm our clinical observations. Conflict of Interest {#acm212057-sec-0019} ==================== No conflict of interest.

FrameByFrame Alternatives FrameByFrame Description FrameByFrame is a stop motion movie production software for the Mac OS X operating systems that will make you able to create the QuickTime movies from pictures that you have already taken from the webcam and cameras connected to your Mac device… read more 13 FrameByFrame Alternatives & Similar Software qStopMotion is a free platform for creating the free stop-motion animated movies. If you have the best idea and storyline, then qStopMotion will assist you in creating a perfect one stop-motion animation movie. AnimatorHD is an ultra-modern and high-tech 3D stop animation platform for Windows operating systems only. In addition to 3D stop-motion animation, this program can also be used as a previsualization tool in the film production as well. Zu3D is a best in the class stop animation software specially designed for educating the children. The best place for using the Zu3D is the educational sector. It can also be used for commercial purposes but is not capable of handling the mega production projects. Designed for the Windows PC only, Stop Motion Pro is the name of a stop motion animation platform for the professional and advanced level of clay animation, stop motion animation, pixelation and cut out animation. iKITMovie is a high-tech stop motion software by using which you can easily create the Claymation and LEGO stop motion movies for the Windows system. By using this, you can create the stop motion and Claymation movies. MonkeyJam is a multiplatform and multifunctional stop motion software for creating all kind of stop animation program. It will make you able to capture images from any camcorder, webcam or a scanner and then assemble them as separate frames to create the animation. iStopMotion is a dedicated stop motion platform for the Mac OS X operating systems for creating the both time lapse and stop motion animations. The latest version of iStopMotion support for the Nikon and Canon DSLRs that are said to be one of the most popular cameras for stop motion animation. Frames is a multi-platform software that works as both stop-motion animation and digital storytelling software. Frames is designed for both professionals and beginners to create the stop motion animation, Claymation and digital storytelling in a very easy to use style. Linux Stopmotion is a dedicated stop motion platform for the Linux operating system for creating the both stop motion animation and Claymation animation movies. Being an open source stop motion animation platform, Linux Stopmotion makes its users able to change the code of origin to create the stop motion animations in their own way. Stop Motion Animator is an easy to use animation program that can be used to bring any animated characters to life with the minimal amount of work. This program is particularly designed for creating the stop motion animation. StopMojo is a cross-platform stop animation tool designed to assist the filmmakers in creating the stop motion animations. This software is capable of capturing program supporting capture of image files from various cameras and video capturing devices. CellSoft Take5 is a simple to use and a highly affordable stop motion animation program for both individual and professional purposes. It is designed to assist both educational institutes and professional filmmakers in creating the professional stop animation movies. Stop Motion Pro Eclipse is designed for the education sectors, animators and animation producers to create the stop motion animation in a more comprehensive and easy to use way. Stop Motion Pro Eclipse is not about capturing the frames of animation only but it has a proper audio scrubbing and Chroma key system as well. More About FrameByFrame FrameByFrame is a stop motion movie production software for the Mac OS X operating systems that will make you able to create the QuickTime movies from pictures that you have already taken from the webcam and cameras connected to your Mac device. There is an integrated iSight cameras system in the FrameByFrame. The easy to use and user friendly interface of FrameByFrame has made the stop motion movie production made easy. FrameByFrame will make you able to first stop animation in few seconds using any camera or webcam already connected to your PC. After creating the movie you can export it into QuickTime formats only. The native software system of FrameByFrame will keep the original pictures making you able to apply any post production you can imagine. The recent version of FrameByFrame contains more improved features that will make you able to produce proper movies for images of any resolution. That means you are no longer restricted to using 72 dpi images. In short, FrameByFrame has made the stop-motion animation making simply easy and simple.

Severe chondrolysis after shoulder arthroscopy: a case series. Chondrolysis has been observed after shoulder arthroscopy and results in severe glenohumeral complications. Twenty three cases of post-arthroscopic glenohumeral chondrolysis, occurring between 2005-2006, are reported following a variety of arthroscopic shoulder procedures. Presenting complaints, signs and symptoms, associated operative findings, and potential etiological factors are reviewed. Management options are summarized. Of the 23 cases of chondrolysis identified in our practice over a two year period, 14 occurred in patients following labral repair using a bioabsorbable device. Seventeen of the 23 patients used a high volume intra-articular pain pump for 48 hours after surgery. Seven of the 23 cases had documented use of a thermal probe. Four cases occurred in shoulders with no reported use of fixation anchors, pain pumps, or thermal probes. All cases had at least a 20 cc intra-articular bolus injection of 0.25% bupivicaine with epinephrine. This case series identifies several common factors that could be responsible for post-arthroscopic glenohumeral chondrolysis. No single mechanism can be implicated based on the results of this study. Although strong concerns are raised over the use of intra-articular local anesthetics, glenohumeral chondrolysis appears to be an unfortunate convergence of multiple factors that may initiate rapid dissolution of articular cartilage and degenerative changes. Chondrolysis is a devastating complication of arthroscopic shoulder surgery that can result in long-term disabling consequences. Further research is required to specifically identify causative factors. Until this is a available, we strongly advise against the use of large doses of intra-articular placement of local anesthetics.

Brain and neck tumors among physicians performing interventional procedures. Physicians performing interventional procedures are chronically exposed to ionizing radiation, which is known to pose increased cancer risks. We recently reported 9 cases of brain cancer in interventional cardiologists. Subsequently, we received 22 additional cases from around the world, comprising an expanded 31 case cohort. Data were transmitted to us during the past few months. For all cases, where possible, we endeavored to obtain the baseline data, including age, gender, tumor type, and side involved, specialty (cardiologist vs radiologist), and number of years in practice. These data were obtained from the medical records, interviews with patients, when possible, or with family members and/or colleagues. The present report documented brain and neck tumors occurring in 31 physicians: 23 interventional cardiologists, 2 electrophysiologists, and 6 interventional radiologists. All physicians had worked for prolonged periods (latency period 12 to 32 years, mean 23.5 ± 5.9) in active interventional practice with exposure to ionizing radiation in the catheterization laboratory. The tumors included 17 cases (55%) of glioblastoma multiforme (GBM), 2 astrocytomas (7%), and 5 meningiomas (16%). In 26 of 31 cases, data were available regarding the side of the brain involved. The malignancy was left sided in 22 (85%), midline in 1, and right sided in 3 operators. In conclusion, these results raise additional concerns regarding brain cancer developing in physicians performing interventional procedures. Given that the brain is relatively unprotected and the left side of the head is known to be more exposed to radiation than the right, these findings of disproportionate reports of left-sided tumors suggest the possibility of a causal relation to occupational radiation exposure.

This invention relates to a remote copy technology in a storage system, and more particularly to a technology which shares a network used for remote copy among plural storage systems. In recent years, computer systems have been used extensively by enterprises for business purposes and data stored in a storage system has been gaining increasing importance. For the storage system, a technology which renders a storage area redundant, such as RAID, has been used to protect data to be stored. Even though the storage area in the storage system is rendered redundant, the whole storage system may be lost in the event of a large-scale disaster. Accordingly, the storage system uses a remote copy technology to protect data even in the event of a large-scale disaster. The remote copy technology renders the whole storage system redundant, disposes the redundant storage system at a remote place, and copies data in the storage system (backup storage system) at the remote place. Upon receipt of a request to write data from a host computer, a primary storage system stores the data write request therein and then copies the data in the backup storage system via a network. The remote copy technology uses a synchronous method or an asynchronous method. In the synchronous method, when the copying of data in the backup storage system is completed, the primary storage system notifies the host computer that the writing of the data has been completed. In the asynchronous method, the primary storage system notifies the host computer that the writing of the data has been completed without awaiting the completion of the copying of the data in the backup storage system. A description will be given hereinbelow to the operation of the primary storage system using the asynchronous method. Upon receipt of a data write request from the host computer, the primary storage system temporarily stores the data in its cache. Then, the primary storage system notifies the host computer that the writing of the data has been completed and transmits the cached data to the backup storage system via the network. In the remote copy using the asynchronous method, when the primary storage system receives a request to write data in an amount exceeding an transmission capacity the network used for the remote copy, the data overflows the cache of the primary storage system to be lost. To prevent such a data loss, the primary storage system refuses a data write request from the host computer when the cache is likely to be overflown. A consideration will be given hereinbelow to the case where plural host computers share the primary storage system. In this case, when the primary storage system receives a request to write a large amount of data from a given one of the host computers, it refuses data write requests from all the host computers so that the usability of the system lowers. To prevent such a problem, a technology has been known which sets a threshold value for the capacity of the cache usable for each of the host computers which write data in the primary storage system (see, e.g., JP 2002-334049 A). The primary storage system monitors the used capacity of the cache on a per host-computer basis and limits data write requests from the host computers based on the threshold values. This protects the cache of the primary storage system from being overflown and prevents a reduction in the usability of the system.

Q: Testing Software on Windows 7. Install vs Virtualization We want to test a software we develop on XP on Windows 7. We have Vista installed on a PC. But we would like to avoid installing Windows 7 and rather run it in a Virtual PC or VMWare. Questions: Q1: Is it possible to run Windows 7 in a Virtual PC or Ware ? Q2: Is it possible to run a 64 bit Vista or Windows 7 in Virtual PC or VmWare on a 32 bit system ? Q3: Are there any ready-to-download Windows 7 images (e.g. from Microsoft) similar to the IE7 images which MS made available before? Thanks a lot! Paul A: Q1. Yes you can run Windows 7 in VMWare, Virtual PC, Virtual Box, and in other virtualization platforms. Q2. It is usually not possible to run a 64-bit guest on 32-bit hardware, with normal virtualization. Q3. No idea, but I doubt it. Windows was not free, the last time I checked. A: Q1) Yes. I personally ran the Windows 7 beta using Virtual PC in Vista, and there's no reason the final build wouldn't work. Q2) That would require the emulator to implement the CPU entirely in software, and that's going to be too slow to be usable. You need a 64-bit OS to emulate a 64-bit PC. EDIT: According to various comments, some virtualization software can do it if you have a 64-bit CPU. You'll need to check the documentation for whatever software you're going to use. Q3) It would be a trial if there is one, and so it will eventually expire. You don't want that. A far more common approach is to emulate XP on 7, and you're almost certainly going to get a better result that way: XP is less resource hungry than Vista and 7, so the performance impact from virtualization will be less noticable.

Q: Creating a window without caption and border How can you create a window without caption and border using CreateWindowEx()? And I why do you use '|' OR operator to combine styles instead of '&' And? A: int WINAPI WinMain(....) { MSG msg; WNDCLASS wc={0}; wc.lpszClassName="MyClass"; wc.lpfnWndProc=DefWindowProc;//You MUST use your own wndproc here wc.hInstance=hInstance; wc.hbrBackground=(HBRUSH)(COLOR_3DFACE+1); wc.hCursor=LoadCursor(NULL,IDC_ARROW); if (!RegisterClass(&wc)) {/*Handle Error*/} HWND hwnd; hwnd=CreateWindowEx(0,wc.lpszClassName,0,WS_POPUP|WS_VISIBLE|WS_SYSMENU,9,9,99,99,0,0,0,0); if (!hwnd) {/*Handle Error*/} while(GetMessage(&msg,0,0,0)>0)DispatchMessage(&msg); return 0; } If you want a border, you can add WS_BORDER or WS_DLGFRAME (Not both). If you don't want to show the window in the taskbar, add the WS_EX_TOOLWINDOW extended style. As to why you need to bitwise OR the styles; OR will combine all the style values, AND is used (by windows) to check which styles are set. Say we had two styles (WS_FOO=1,WS_BAR=2): 1 AND 2 = 0 (Binary: 01 AND 10 = 00) 1 OR 2 = 3 (Binary: 01 OR 10 = 11) See wikipedia for more info.

Challenging the cost of clinical negligence. Healthcare professionals in South Africa (SA) are facing challenging times. As the clinical negligence claims environment in SA deteriorates, the impact is being felt by healthcare professionals, but also by the wider public owing to the strain that costs place on the public purse. The authors look at the current claims environment, and explain why a debate about reform is so important.

Q: Error when insert data on database I'm trying to save some values in a local database. This database has 3 columns besides de ID column, that are: NAME, GROUP, ORDER. public class TravelOrderDatabaseHelper extends SQLiteOpenHelper { public void onCreate(SQLiteDatabase db) { db.execSQL("CREATE TABLE " + TABLE_NAME + " (" + TravelOrder._ID + " INTEGER PRIMARY KEY AUTOINCREMENT," + TravelOrder.NAME + " TEXT NOT NULL, " + TravelOrder.GROUP + " TEXT NOT NULL, " + TravelOrder.ORDER + " TEXT NOT NULL " + ");"); } - public class TravelOrderProvider extends ContentProvider { public class TravelOrder implements BaseColumns { public static final String NAME = "name"; public static final String GROUP = "group"; public static final String ORDER = "order"; } In the main activity, tho save the values in the db I use this method: private void insertTraveler(String name, String group, String order){ ContentValues newValues = new ContentValues(); newValues.put(TravelOrder.NAME, name); newValues.put(TravelOrder.GROUP, group); newValues.put(TravelOrder.ORDER, order); getContentResolver().insert(TravelOrderProvider.CONTENT_URI, newValues); } But this is what I'm getting on the LogCat: 06-04 15:21:46.919: E/AndroidRuntime(30381): FATAL EXCEPTION: main 06-04 15:21:46.919: E/AndroidRuntime(30381): android.database.sqlite.SQLiteException: near "group": syntax error: , while compiling: CREATE TABLE travelorder (_id INTEGER PRIMARY KEY AUTOINCREMENT,name TEXT NOT NULL, group TEXT NOT NULL, order TEXT NOT NULL ); 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteCompiledSql.native_compile(Native Method) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteCompiledSql.<init>(SQLiteCompiledSql.java:68) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteProgram.compileSql(SQLiteProgram.java:134) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteProgram.compileAndbindAllArgs(SQLiteProgram.java:361) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteStatement.acquireAndLock(SQLiteStatement.java:260) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteStatement.executeUpdateDelete(SQLiteStatement.java:84) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteDatabase.executeSql(SQLiteDatabase.java:2006) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteDatabase.execSQL(SQLiteDatabase.java:1946) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.data.TravelOrderDatabaseHelper.onCreate(TravelOrderDatabaseHelper.java:29) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.database.sqlite.SQLiteOpenHelper.getWritableDatabase(SQLiteOpenHelper.java:165) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.data.TravelOrderProvider.insert(TravelOrderProvider.java:73) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.content.ContentProvider$Transport.insert(ContentProvider.java:200) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.content.ContentResolver.insert(ContentResolver.java:734) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.NewTravelActivity.insertTraveler(NewTravelActivity.java:429) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.NewTravelActivity.sendTravelersToOrder(NewTravelActivity.java:243) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.NewTravelActivity.access$1(NewTravelActivity.java:238) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.irvvin.carpooling.NewTravelActivity$1.onClick(NewTravelActivity.java:145) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.view.View.performClick(View.java:3549) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.view.View$PerformClick.run(View.java:14393) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.os.Handler.handleCallback(Handler.java:605) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.os.Handler.dispatchMessage(Handler.java:92) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.os.Looper.loop(Looper.java:154) 06-04 15:21:46.919: E/AndroidRuntime(30381): at android.app.ActivityThread.main(ActivityThread.java:4944) 06-04 15:21:46.919: E/AndroidRuntime(30381): at java.lang.reflect.Method.invokeNative(Native Method) 06-04 15:21:46.919: E/AndroidRuntime(30381): at java.lang.reflect.Method.invoke(Method.java:511) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:784) 06-04 15:21:46.919: E/AndroidRuntime(30381): at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:551) 06-04 15:21:46.919: E/AndroidRuntime(30381): at dalvik.system.NativeStart.main(Native Method) A: You cannot use group as a table name as it is inbuilt in keyword of sqlite.. use different keyword A: group and order both are keywoerds into Sqlite. You can not use them as column name here. Change the value of public static final String GROUP = "group"; // "group_column" public static final String ORDER = "order"; // "order_colomn" See the list of SQLite Keywords.

How I used Stack Overflow and GitHub to get dream job before 19 without degree - kuzirashi https://hackernoon.com/how-i-used-stack-overflow-github-to-get-dream-job-before-19-without-degree-8cb5184e2bec ====== magma17 Education system really works.

<?php /** * This file implements the UI view for the user picture crop form. * * This file is part of the evoCore framework - {@link http://evocore.net/} * See also {@link https://github.com/b2evolution/b2evolution}. * * @license GNU GPL v2 - {@link http://b2evolution.net/about/gnu-gpl-license} * * @copyright (c)2003-2020 by Francois Planque - {@link http://fplanque.com/} * * @package admin */ if( !defined('EVO_MAIN_INIT') ) die( 'Please, do not access this page directly.' ); global $display_mode, $Settings; /** * @var instance of User class */ global $edited_User; /** * @var current action */ global $action; /** * @var user permission, if user is only allowed to edit his profile */ global $user_profile_only; /** * @var the action destination of the form (NULL for pagenow) */ global $form_action; /** * @var instance of User class */ global $current_User; /** * @var File that should be cropped */ global $cropped_File; global $aspect_ratio, $content_width, $content_height; $aspect_ratio = param( 'aspect_ratio', 'double' ); $content_width = param( 'content_width', 'integer' ); $content_height = param( 'content_height', 'integer' ); $original_image_size = explode( 'x', $cropped_File->get_image_size() ); $image_height = $original_image_size[1]; $image_width = $original_image_size[0];; $min_avatar_size = $Settings->get( 'min_picture_size' ); $can_crop = ( $image_height >= $min_avatar_size && $image_width >= $min_avatar_size && ! ( $image_height == $min_avatar_size && $image_width == $min_avatar_size ) ); if( $display_mode != 'js' ) { // ------------------- PREV/NEXT USER LINKS ------------------- user_prevnext_links( array( 'user_tab' => 'avatar' ) ); // ------------- END OF PREV/NEXT USER LINKS ------------------- } $Form = new Form( $form_action, 'user_checkchanges' ); if( is_admin_page() ) { $ctrl_param = '?ctrl=user&amp;user_tab=avatar&amp;user_ID='.$edited_User->ID; $form_title = ''; $form_class = 'fform'; } else { global $Collection, $Blog; $form_title = ''; $form_class = 'bComment'; $ctrl_param = url_add_param( $Blog->gen_blogurl(), 'disp='.$disp ); } if( $display_mode != 'js' && is_admin_page() ) { if( !$user_profile_only ) { echo_user_actions( $Form, $edited_User, $action ); } $form_text_title = TB_( 'Crop profile picture' ); // used for js confirmation message on leave the changed form $form_title = get_usertab_header( $edited_User, '', $form_text_title ); } $cropped_image_tag = $cropped_File->get_tag( '', '', '', '', 'original', '', '', '', '', '', '', '', '', '', 'none' ); // Display this error when JS is not enabled echo '<noscript>' .'<p class="error text-danger">'.TB_('Please activate Javascript in your browser in order to use this feature.').'</p>' .'<style type="text/css">form#user_checkchanges { display:none }</style>' .'</noscript>'; $Form->begin_form( $form_class, $form_title, array( 'title' => ( isset( $form_text_title ) ? $form_text_title : $form_title ) ) ); if( is_admin_page() ) { $Form->hidden_ctrl(); } else { $Form->hidden( 'disp', $disp ); $Form->hidden( 'action', $action ); } $Form->add_crumb( 'user' ); $Form->hidden( 'user_tab', param( 'user_tab_from', 'string', 'avatar' ) ); $Form->hidden( 'user_ID', isset( $edited_User ) ? $edited_User->ID : $current_User->ID ); $Form->hidden( 'file_ID', $cropped_File->ID ); $Form->hidden( 'image_crop_data', '' ); if( isset( $Blog ) ) { $Form->hidden( 'blog', $Blog->ID ); } $close_icon = ''; if( $display_mode == 'js' ) { // Display a close link for popup window $close_icon = action_icon( TB_('Close this window'), 'close', '', '', 0, 0, array( 'id' => 'close_button', 'class' => 'floatright' ) ); } // Start displaying content if( ! $can_crop ) { echo '<div style="height: '.$content_height.'px; width: '. $content_width.'px; position=relative; text-align: center;">'; echo '<div style="position: absolute; top: 50%; left: 50%; transform: translate(-50%, -50%);">'.sprintf( TB_('Only images larger than %dx%d pixels can be cropped.'), $min_avatar_size, $min_avatar_size ).'</div>'; echo '</div>'; } else { echo '<div id="user_crop_content" style="height: '.$content_height.'px; width: '. $content_width.'px;">'; echo '</div>'; $Form->button( array( 'type' => 'submit', 'name'=>'actionArray[crop]', 'value'=> TB_('Apply'), 'class' => 'SaveButton btn-primary' ) ); } $Form->end_form(); if( $can_crop ) { ?> <style> #user_crop_workarea img { object-fit: contain; width: <?php echo $content_width;?>; height: <?php echo $content_height;?>; visibility: hidden; } div.jcrop-holder { margin: 0 auto; position: relative; top: 50%; transform: translateY(-50%); } </style> <script> var jcrop_api; var image_url = '<?php echo format_to_js( $cropped_File->get_url() ); ?>'; var gutter = 10; var padding = 0; var content_height = <?php echo $content_height;?>; var content_width = <?php echo $content_width;?>; var show_large_preview = true; var show_small_preview = true; var large_preview_size = 128; var small_preview_size = 64; var preview_size; var preview_margin = 10; var preview_orientation = 'portrait'; var aspect_ratio = <?php echo $aspect_ratio;?>; var render_mode = 'landscape'; var workarea_height; var workarea_width; var workarea_aspect_ratio; var original_image_height = <?php echo $image_height;?>; var original_image_width = <?php echo $image_width;?>; var original_image_aspect_ratio = original_image_height / original_image_width; var size_ratio = 1; var working_image_height = null; var working_image_width = null; var working_image_aspect_ratio = null; var initial_crop_selection = []; //console.debug( 'Content: ', content_width, content_height ); //console.debug( 'Original Image: ', original_image_width, original_image_height ); function get_working_image_dimensions( w_height, w_width, image_aspect_ratio ) { // We'll use this function to determine which mode will provide the larger working image var w_aspect_ratio = w_height / w_width; if( w_aspect_ratio > image_aspect_ratio ) { // width is limiting dimension i_width = w_width; i_height = w_width * original_image_aspect_ratio; } else { // height is limiting dimension i_height = w_height; i_width = w_height / original_image_aspect_ratio; } i_area = i_width * i_height; return { height: i_height, width: i_width, area: i_area }; } function init_layout() { var lw_width, lw_height, lw_area, lw_aspect_ratio; var pw_width, pw_height, pw_area, pw_aspect_ratio; var i_width, i_height, i_area; if( aspect_ratio < 1 && ( large_preview_size > ( content_width / 3 ) ) || aspect_ratio > 1 && ( large_preview_size > ( content_height / 3 ) ) ) { preview_size = small_preview_size; } else { preview_size = large_preview_size; } // Try landscape mode first lw_width = content_width - preview_size - gutter - ( padding * 2 ); lw_height = content_height - ( padding * 2 ); lw_area = lw_width * lw_height; lw_aspect_ratio = lw_height / lw_width; var l_view = get_working_image_dimensions( lw_height, lw_width, original_image_aspect_ratio ); // Now let's try portrait mode pw_height = content_height - preview_size - gutter - ( padding * 2 ); pw_width = content_width - ( padding * 2 ); pw_area = pw_width * pw_height; pw_aspect_ratio = pw_height / pw_width; var p_view = get_working_image_dimensions( pw_height, pw_width, original_image_aspect_ratio ); // See what mode provides the largest working image if ( l_view.area > p_view.area ) { render_mode = 'landscape'; } else { render_mode = 'portrait'; } // Determine if we can show all the preview images if( render_mode == 'portrait' ) { if( ( content_width < original_image_width ? content_width : original_image_width ) > ( ( ( large_preview_size + preview_margin ) * 2 ) + ( ( small_preview_size + preview_margin ) * 2 ) ) ) { show_large_preview = true; show_small_preview = true; preview_size = large_preview_size; } else if( ( ( content_width < original_image_width ? content_width : original_image_width ) > ( ( large_preview_size + preview_margin ) * 2 ) ) && ( ( 0.25 * content_height ) >= large_preview_size ) ) { show_large_preview = true; show_small_preview = false; preview_size = large_preview_size; } else if( ( ( content_width < original_image_width ? content_width : original_image_width ) > ( ( small_preview_size + preview_margin ) * 2 ) ) && ( ( 0.25 * ( render_mode == 'landscape' ? content_width : content_height ) ) >= small_preview_size ) ) { show_large_preview = false; show_small_preview = true; preview_size = small_preview_size; } else { show_large_preview = false; show_small_preview = false; preview_size = 0; } } else { if( ( content_height < original_image_height ? content_height : original_image_height ) > ( ( ( large_preview_size + preview_margin ) * 2 ) + ( ( small_preview_size + preview_margin ) * 2 ) ) ) { show_large_preview = true; show_small_preview = true; preview_size = large_preview_size; } else if( ( ( content_height < original_image_height ? content_height : original_image_height ) > ( ( large_preview_size + preview_margin ) * 2 ) ) && ( ( 0.25 * ( content_width < content_height ? content_width : content_height ) ) >= large_preview_size ) ) { show_large_preview = true; show_small_preview = false; preview_size = large_preview_size; } else if( ( ( content_height < original_image_height ? content_height : original_image_height ) > ( ( small_preview_size + preview_margin ) * 2 ) ) && ( ( 0.25 * ( content_width < content_height ? content_width : content_height ) ) >= small_preview_size ) ) { show_large_preview = false; show_small_preview = true; preview_size = small_preview_size; } else if( ( content_height < original_image_height ? content_height : original_image_height ) > ( ( small_preview_size ) ) && ( ( content_width ) > ( lw_width + ( small_preview_size * 2 ) ) ) ) { // This is a special case when there is not enough room for a vertical preview but a horizontal preview can be accomodated show_large_preview = false; show_small_preview = true; preview_size = small_preview_size; preview_orientation = 'landscape'; } else { show_large_preview = false; show_small_preview = false; preview_size = 0; } } //console.debug( 'Render mode: ', render_mode ); //console.debug( 'Show large preview: ', show_large_preview ); //console.debug( 'Show small preview: ', show_small_preview ); //console.debug( 'Preview Size: ', preview_size ); } function init_workarea() { if( render_mode == 'portrait' ) { workarea_height = content_height - preview_size - gutter - ( padding * 2 ); workarea_width = content_width - ( padding * 2 ); } else { workarea_height = ( content_height - ( padding * 2 ) ); workarea_width = content_width - preview_size - gutter - ( padding * 2 ); } workarea_aspect_ratio = ( workarea_height / workarea_width ); //console.debug( 'Workarea: ', workarea_width, workarea_height ); } function init_working_image() { if( workarea_aspect_ratio == original_image_aspect_ratio ) { working_image_height = workarea_height; working_image_width = workarea_width; size_ratio = original_image_height / working_image_width; //console.debug( 'Limiting dimension: ', 'both' ); } else if( workarea_aspect_ratio > original_image_aspect_ratio ) { working_image_width = workarea_width; working_image_height = workarea_width * original_image_aspect_ratio; size_ratio = original_image_width / working_image_width; //console.debug( 'Limiting dimension: ', 'width' ); } else { working_image_height = workarea_height; working_image_width = workarea_height / original_image_aspect_ratio; size_ratio = original_image_height / working_image_height; //console.debug( 'Limiting dimension: ', 'height' ); } // Should be always equal to original image aspect ratio working_image_aspect_ratio = working_image_height / working_image_width; //console.debug( 'Image: ', working_image_width, working_image_height ); } function init_crop_selection() { var crop_size; initial_crop_selection = []; if( original_image_aspect_ratio > 1 ) { crop_size = original_image_width; } else { crop_size = original_image_height; } if( original_image_aspect_ratio > 1 ) { initial_crop_selection.push( 0 ); initial_crop_selection.push( 0 ); initial_crop_selection.push( crop_size ); initial_crop_selection.push( crop_size ); } else { initial_crop_selection.push( ( original_image_width / 2 ) - ( crop_size / 2 ) ); initial_crop_selection.push( ( original_image_height / 2 ) - ( crop_size / 2 ) ); initial_crop_selection.push( ( original_image_width / 2 ) + ( crop_size / 2 ) ); initial_crop_selection.push( ( original_image_height / 2 ) + ( crop_size / 2 ) ); } //console.debug( 'Initial selection: ', initial_crop_selection ); } function render_content() { var content = jQuery( 'div#user_crop_content' ); var working_image = jQuery( '<img />', { src: image_url }); var workarea = jQuery( '<div />', { id: 'user_crop_workarea', style: { 'background-color': '#f2f2f2', height: workarea_height + 'px', width: workarea_width + 'px' } }); var previews = jQuery( '<div />', { id: 'user_crop_preview', style: { 'background-color': 'white', height: preview_size + 'px', 'text-align': 'center', 'margin-top': gutter + 'px' } }); // Large square preview var preview_lg_sq = jQuery( '<div />', { class: 'preview_cropped_image' }).css({ width: '128px', height: '128px', }).append( working_image.clone() ); // Small square preview var preview_sm_sq = jQuery( '<div />', { class: 'preview_cropped_image' }).css({ width: '64px', height: '64px', }).append( working_image.clone() ); // Large circle preview var preview_lg_c = jQuery( '<div />', { class: 'preview_cropped_image circle' }).css({ width: '128px', height: '128px', }).append( working_image.clone() ); // Small circle preview var preview_sm_c = jQuery( '<div />', { class: 'preview_cropped_image circle' }).css({ width: '64px', height: '64px', }).append( working_image.clone() ); if( render_mode == 'portrait' ) { // Portrait workarea.css({ 'background-color': '#f2f2f2', height: workarea_height + 'px', width: workarea_width + 'px' }); previews.css({ 'background-color': '#ffffff', height: preview_size + 'px', width: content_width + 'px', 'text-align': 'center', 'margin-top': gutter + 'px' }); if( show_large_preview ) { previews.append( preview_lg_sq ); } if( show_small_preview ) { previews.append( preview_sm_sq ); } if( show_large_preview ) { previews.append( preview_lg_c ); } if( show_small_preview ) { previews.append( preview_sm_c ); } workarea.prepend( working_image ); content.append( workarea ); content.append( previews ); preview_images = jQuery( 'div#user_crop_preview div.preview_cropped_image:not(:last-child)' ); preview_images.css( 'margin-bottom', 0 ); preview_images.css( 'margin-right', preview_margin + 'px' ); } else { // Landscape workarea.css({ 'background-color': '#f2f2f2', float: 'left', width: workarea_width + 'px', height: workarea_height + 'px' }); previews.css({ 'background-color': '#ffffff', float: 'left', width: ( preview_orientation == 'portrait' ? preview_size : ( ( preview_size + preview_margin ) * 2 ) ) + 'px', height: content_height + 'px', 'text-align': 'center', 'margin-left': gutter + 'px' }); workarea.prepend( working_image ); var preview_wrapper = jQuery( '<div />' ); if( show_large_preview ) { preview_wrapper.append( preview_lg_sq ); } if( show_small_preview ) { preview_wrapper.append( preview_sm_sq ); } if( show_large_preview ) { preview_wrapper.append( preview_lg_c ); } if( show_small_preview ) { preview_wrapper.append( preview_sm_c ); } previews.append( preview_wrapper ); content.append( workarea ); content.append( previews ); content.append( jQuery( '<div />', { style: { clear: 'both' } } ) ); preview_images = jQuery( 'div#user_crop_preview div.preview_cropped_image:not(:last-child)' ); if( preview_orientation == 'portrait' ) { preview_images.css( 'margin-bottom', preview_margin + 'px' ); preview_images.css( 'margin-right', 0 ); } else { preview_images.css( 'margin-bottom', 0 ); preview_images.css( 'margin-right', preview_margin + 'px' ); } } // Adjust modal dimensions and content to minimize workarea margin var wah_margin = workarea_height - ( original_image_height < working_image_height ? original_image_height : working_image_height ); var waw_margin = workarea_width - ( original_image_width < working_image_width ? original_image_width : working_image_width ); var modal_dialog = jQuery( 'div.modal-dialog' ).length ? jQuery( 'div.modal-dialog' ) : jQuery( '#overlay_wrap' ); var modal_body = jQuery( 'div.modal-body' ).length ? jQuery( 'div.modal-body' ) : jQuery( '#overlay_page' ); var content = jQuery( 'div#user_crop_content' ); //console.debug( 'Margin: ', wah_margin, waw_margin ); //console.debug( 'Workarea: ', workarea_height, workarea_width ); //console.debug( 'Working Image: ', working_image_height, working_image_width ); workarea.css( 'height', ( workarea_height - wah_margin ) + 'px' ); workarea.css( 'width', ( workarea_width - waw_margin ) + 'px' ); content.css( 'height', ( content_height - wah_margin ) + 'px' ); content.css( 'width', ( content_width - waw_margin + ( preview_orientation == 'landscape' ? preview_size + preview_margin + gutter : 0 ) ) + 'px' ); modal_body.css( 'height', ( parseInt( modal_body.css( 'height' ) ) - wah_margin ) + 'px' ); modal_body.css( 'min-height', ( parseInt( modal_body.css( 'min-height' ) ) - wah_margin ) + 'px' ); modal_body.css( 'width', ( parseInt( modal_body.css( 'width' ) ) - waw_margin + ( preview_orientation == 'landscape' ? preview_size + preview_margin + gutter : 0 ) ) + 'px' ); modal_dialog.css( 'height', ( parseInt( modal_dialog.css( 'height' ) ) - wah_margin ) + 'px' ); modal_dialog.css( 'width', ( parseInt( modal_dialog.css( 'width' ) ) - waw_margin + ( preview_orientation == 'landscape' ? preview_size + preview_margin + gutter : 0 ) ) + 'px' ); if( render_mode == 'portrait' ) { previews.css( 'width', ( parseInt( previews.css( 'width' ) ) - waw_margin ) + 'px' ); } else { previews.css( 'height', ( parseInt( previews.css( 'height' ) ) - wah_margin ) + 'px' ); } } function update_preview( coords ) { var target_cropped_image_width = original_image_width; var target_cropped_image_height = original_image_height; var percent_width = Math.ceil( coords.w / target_cropped_image_width * 10000 ) / 100; var percent_height = Math.ceil( coords.h / target_cropped_image_height * 10000 ) / 100; var percent_top = Math.ceil( coords.x / target_cropped_image_width * 10000 ) / 100; var percent_left = Math.ceil( coords.y / target_cropped_image_height * 10000 ) / 100; jQuery( 'input[name=image_crop_data]' ).val( percent_top + ':' + percent_left + ':' + percent_width + ':' + percent_height ); var top = coords.y; var left = coords.x; if( coords.w > coords.h ) { // Center a cropping area of horizontal image left += ( coords.w / 2 ) - ( coords.h / 2 ); } else { var top_shift = ( coords.h - coords.w ) * 0.15; if( top + top_shift + coords.w < top + coords.h ) { // top - 15% top += top_shift; } } jQuery( '.preview_cropped_image img' ).each( function() { var ratio = jQuery( this ).parent().width() / ( coords.w < coords.h ? coords.w : coords.h ); jQuery( this ).css({ width: Math.round( ratio * target_cropped_image_width ) + 'px', height: Math.round( ratio * target_cropped_image_height ) + 'px', marginLeft: '-' + Math.round( ratio * left ) + 'px', marginTop: '-' + Math.round( ratio * top ) + 'px' }); }); } function selection_release() { // Reset to initial selection if selection is released jcrop_api.setSelect( initial_crop_selection ); } function init_jcrop_tool( image ) { var min_avatar_size = <?php echo $min_avatar_size;?>; options = { boxWidth: working_image_width, boxHeight: working_image_height, aspectRatio: 1, minSize: [ min_avatar_size, min_avatar_size ], onChange: update_preview, onSelect: update_preview, onRelease: selection_release }; image.Jcrop( options, function() { jcrop_api = this; jcrop_api.setSelect( initial_crop_selection ); image.css({ visibility: 'visible' }); }); } // Initialize content, workarea and working image init_layout(); init_workarea(); init_working_image(); init_crop_selection(); // Render everything render_content(); // Initialize jcrop tool only after the image is fully loaded jQuery( '#user_crop_workarea img' ).on( 'load', function() { init_jcrop_tool( jQuery( this ) ); }); </script> <?php } ?>

Education for outreach: development of an interprofessional MSc in critical care. This article describes the development of a new multi-professional Master's degree programme, which is aimed at supporting the education of practitioners in meeting the needs of Level 1 critically ill patients. The main drivers that influenced its development were Comprehensive Critical Care (Department of Health, 2000a) and the National Health Service Modernisation Agency for Critical Care. The programme was aimed primarily at junior doctors, senior ward nurses, senior physiotherapists and members of Outreach teams. The programme was designed to maximize work-based learning and is framed around two clinically focused units. The programme is informed by principles of quality improvement and aims to prepare practitioners to meet the needs of Level 1 patients 'at the point of need'.

import Foundation /// A Nimble matcher that succeeds when the actual expression evaluates to an /// error from the specified case. /// /// Errors are tried to be compared by their implementation of Equatable, /// otherwise they fallback to comparison by _domain and _code. public func matchError<T: Error>(_ error: T) -> Predicate<Error> { return Predicate.fromDeprecatedClosure { actualExpression, failureMessage in let actualError: Error? = try actualExpression.evaluate() setFailureMessageForError(failureMessage, postfixMessageVerb: "match", actualError: actualError, error: error) var matches = false if let actualError = actualError, errorMatchesExpectedError(actualError, expectedError: error) { matches = true } return matches }.requireNonNil } /// A Nimble matcher that succeeds when the actual expression evaluates to an /// error from the specified case. /// /// Errors are tried to be compared by their implementation of Equatable, /// otherwise they fallback to comparision by _domain and _code. public func matchError<T: Error & Equatable>(_ error: T) -> Predicate<Error> { return Predicate.fromDeprecatedClosure { actualExpression, failureMessage in let actualError: Error? = try actualExpression.evaluate() setFailureMessageForError(failureMessage, postfixMessageVerb: "match", actualError: actualError, error: error) var matches = false if let actualError = actualError as? T, error == actualError { matches = true } return matches }.requireNonNil } /// A Nimble matcher that succeeds when the actual expression evaluates to an /// error of the specified type public func matchError<T: Error>(_ errorType: T.Type) -> Predicate<Error> { return Predicate.fromDeprecatedClosure { actualExpression, failureMessage in let actualError: Error? = try actualExpression.evaluate() setFailureMessageForError( failureMessage, postfixMessageVerb: "match", actualError: actualError, errorType: errorType ) var matches = false if actualError as? T != nil { matches = true } return matches }.requireNonNil }

Domino\'s Pizza First Half Profits Drop by 46% on German Business Costs Domino's Pizza posted a 46% fall in first-half profits owing to poor growth in the UK business and charges against its German arm. Although the pre-tax profit increased by 10.3% to £25.7m (€30m, $39.3m) from the previous year, exceptional items ate into the group's cash pile. The global fast-food delivery brand's reported a pre-tax profit of £11.6m for the six months to 30 June, after exceptional items, which is a drop from £21.5m in 2012. Exceptional items are related to an £11.1m impartial charge on the German master franchise agreement and corporate store costs. Domino's also warned the market that its losses, from its two-year old German subsidiary, could outgrow its core British business by as much as £6m in 2013. "In our fledgling German business, as in the UK, our franchisees are those with the expertise to run great stores," said Lance Batchelor, Chief Executive Officer. "We know the best way to get great results from stores is to put them in the hands of franchisees - and with five world-class franchisees now operating in the German market and more arriving shortly, we are excited about the future in this territory." Overall Growth The underlying business in the UK and Ireland continued to grow, with sales increasing by 11.2% to £317.9m, compared to £285.9m year-on-year. Britain's biggest pizza delivery firm reported improved like-for-like sales with an increase of 6.4%. In the Republic of Ireland, sales increased by 6.5%, compared to the previous year's 2.9%. In Germany it posted a surge of 23.8% while in Switzerland the like-for-like was reported at 7.8%. Group sales also increased 13.8% to £326.5m. Earnings per share climbed 5.4% to 10.71p where the company's interim dividend increased by 7.6% this year. The company which has nearly 800 stores has grown as a result of new products, promotions and online demand.

Author's page István ÖRKÉNY ( 1912 - 1979 ) Biography 1912 born in Budapest 1932-1934 studies Pharmacology at Semmelweiss University, Budapest 1939 lives in Paris 1940 returns to Hungary 1943-1946 prisoner of war in the USSR 1949-1951 works as a scriptwriter 1958-1963 banned from publication for his participation in the 1956 Revolution 1971 Collected Works begin to be published 1979 dies in Budapest As a dramatist, novelist and short story writer, István Örkény is one of the most popular authors in Hungary; his sense of the absurd and his special genre, the "one-minute story", have a place of their own in world literature. He is distinguished from the other masters of the absurd in his special choice of setting and characters: Örkény's writings are about Hungarians struggling with the absurdities of twentieth-century history. His first short stories were published by Attila József's literary journal; his story "Tengertánc" foretold the Nazi terror. In the title story the inmates of an insane asylum seize power over a city and introduce many insane, dictatorial measures which are impossible to abide by - among others, a linguistic reform. These two grotesque tragicomedies are perhaps Örkény's most famous and most frequently performed plays. Both are based on prose written in the early Sixties, and were later developed into dramas and film scripts. Thanks to Örkény's mosaic-like compositional technique, they were equally successful in book form as on stage and screen. Both feature ordinary people as their main characters and both are set in Hungary. Macskajáték (1963, Cats Play) is a comedy about an aged Budapest widow whose only joy is to treat a once celebrated tenor to dinner. But another woman who is even older manages to seduce the object of her affections. The drama version is even more efficient than the novel. Örkény's authorial instruction to the drama: "This play should be acted throughout as if it consisted of a single sentence. It does not allow for any slowing down, interval or change of scene, for from beginning to end it is nothing but an increasingly tense and nervous argument which the increasingly tense and nervous heroine, Mrs Orbán, carries on with herself, her sister, the whole world." Cats Play is in fact a drama about the freedom of the individual. Tóték (1967, The Tót Family) resembles Cats Play in many respects; it is simultaneously funny, tragic and absurd; it presents the struggle of zealous middle-class people to survive. The book centres on a village couple, the Tóts, who are visited by a neurotic army officer. He gradually engulfs his hosts in terror. The family's endurance is due solely to their belief that they can help their only son, who is on the Russian front, to a better and safer service by pleasing his commanding officer. The absurdity of the situation comes from the fact that their son - unbeknownst to them - is already dead. According to critic Anna Földes, the major "represents the system itself, keeping the ordinary man in fright and humiliating subjugation." This book is a collection of Örkény's most popular genre, the so-called "one-minute story". Critic László Varga explains: "These stories are generally between two and four hundred words long and often run only a few lines, but their spare structure bursts with power. The essence of Örkény's experiment is to break down the short story into its basic constituents: he keeps stripping off individual layers until he reaches the core, which is an integrated work.... These one-minute stories are little gems representing all shades of humour from gentle irony to morbid "black humour"; they are epigrams in prose." Pisti in the Bloodbath This is probably Örkény’s most complex play, written in 1969. Still, it could not be published until 1972, and it was performed only in 1979. As Örkény said in an interview: „Actually, this is a biographical work, a grotesque biography based on our shared experiences. This play has neither plot - in its traditional sense - nor dramatic characters any longer. … The characters in Pisti are shaped only by the given situation, moreover, one can be replaced by the other seomtimes.” (Dialogue on the Grotesque, Palatinus, 2000, p.219) So the dramaturgical structure of the piece is rather complicated. In the 1979 version there is not only on Pisti, but there are four. Their thousand faces are represented in absurd, though apparently realistic situations ranging from the nineteen-thirties through the years of World War II and the Rákosy-period till the consolidation after the supressed revolution. The life situations and possible attitudes of Örkény’s generation are shocking and absurd, just like the political history of the twentieth century. Pisti represents the weird twists of history from the perspective of Everyman. This must have made the drama problematic for the cultural politics of its age, since history written from a bottom-view does not obediently repeat the dogmas of resolute saviours of the world. Although the drama is intertwined with seeking redemption, it refers to the Bible several times (e.g. walking on water, the three magi), and at a certain point even the chief rabbi of New York admits that Pisit is the Messiah. However, the grotesque tone of the play does not hide its primary suggestion: one must survive and restart, by all means. Our social roles are manifold: Örkény’s representation of people is dynamic. Any time, anything can be restarted.

899 F.Supp. 275 (1995) Charles K. SCOTT, Plaintiff, v. COMMISSIONER OF SOCIAL SECURITY, Defendant. Civ. A. No. 2:95-0078. United States District Court, S.D. West Virginia, Charleston Division. September 27, 1995. *276 Montie VanNostrand, Webster Springs, WV, for plaintiff. Carol Casto, Assistant U.S. Attorney, Charleston, WV, Rebecca Betts, United States Attorney, Charleston, WV, Charlotte Hardnett, Office of Regional Attorney, Philadelphia, PA, for defendant. JUDGMENT ORDER HADEN, Chief Judge. This action was referred to the Honorable Jerry D. Hogg, United States Magistrate Judge, who has submitted his proposed findings of fact and recommendation for disposition pursuant to the provisions of 28 U.S.C. § 636(b)(1)(B). The Magistrate Judge's Report-Recommendation was filed on September 8, 1995. There have been no objections to the Magistrate Judge's Report-Recommendation. Accordingly, the Court adopts and incorporates herein the Magistrate Judge's Report-Recommendation and DENIES the Defendant's motion for summary judgment. The Court further GRANTS the Plaintiff's motion for summary judgment and REVERSES the second decision of the administrative law judge dated April 6, 1994 insofar as the Plaintiff's disability onset date was set as of August 1, 1990. The Court ORDERS the Plaintiff be awarded disability benefits as of the original onset date of April 4, 1989. The Court ORDERS this action be stricken from the docket of the Court. REPORT-RECOMMENDATION HOGG, United States Magistrate Judge. This is an action seeking review of the final decision of the Commissioner of Social Security denying the plaintiff's application for a period of disability and disability insurance benefits filed pursuant to the provisions of the Social Security Act, as amended. This case is presently pending on the parties' cross-motions for summary judgment. By standing order docketed February 3, 1995, this action was referred to the undersigned United States Magistrate Judge, who was designated to consider the pleadings and evidence and to submit his proposed findings of fact and recommendation for the disposition thereof pursuant to the provisions of 28 U.S.C. § 636. The plaintiff, Charles K. Scott, filed an application for a period of disability and disability insurance benefits on May 8, 1989, alleging that he had been disabled since April 4, 1989. On November 27, 1990, an administrative law judge held that the plaintiff was entitled to a period of disability commencing April 4, 1989, and to disability insurance benefits. By memorandum dated May 28, 1992, the Office of Disability and International Operations returned the case to the administrative law judge stating that information had been received from the Social Security field office dated April 15, 1992, showing the claimant had worked from December 1989, through July 1990, earning $2,000.00 a month. Following a subsequent hearing, the administrative law judge concluded that the plaintiff was disabled, but that such disability commenced August 1, 1990. A claimant for disability benefits bears the burden of proving a disability, 42 U.S.C. § 423(d)(5); Blalock v. Richardson, *277 483 F.2d 773, 774 (4th Cir.1972), which is defined as the "inability to engage in any substantial gainful activity by reason of any medically determinable impairment which can be expected to last for a continuous period of not less than 12 months...." 42 U.S.C. § 423(d)(1)(A). Once the claimant makes a prima facie showing of a physical impairment which effectively precludes him from returning to his past relevant work, the burden of going forward shifts to the Secretary. The Commissioner then must show two things: (1) that the claimant, considering his age, education, work experience, skills, and physical shortcomings, has the capacity to perform an alternative job and (2) that this specific job exists in the national economy. McLain v. Schweiker, 715 F.2d 866 (4th Cir.1983); McLamore v. Weinberger, 538 F.2d 572 (4th Cir.1976). The Social Security Regulations establish a "sequential evaluation" for the adjudication of disability claims. 20 C.F.R. § 404.1520 If an individual is found "not disabled" at any step, further inquiry is unnecessary. Id. § 404.1520(a). The first inquiry under the sequence concerns whether a claimant is currently engaged in substantial gainful employment. If he is not, the second inquiry is whether the claimant has a "severe" impairment. If a severe impairment is present, the third inquiry is whether such impairment meets or equals any of the impairments listed in Appendix 1 to Subpart P of the Administrative Regulations No. 4. If it does, the claimant is found disabled and is awarded benefits. If it does not, the fourth inquiry is whether the claimant's impairments prevent him from performing his past relevant work. By satisfying inquiry 4, the claimant establishes a prima facie case of disability. The burden then shifts to the Commissioner and leads to the fifth and final inquiry: whether the claimant is able to perform other forms of substantial gainful activity, considering both his remaining physical and mental capacities and his age, education, and prior work experience. In this particular case, the administrative law judge found that the medical evidence establishes that "the claimant has a history of a motor vehicle accident in which he sustained multiple injuries including fractures of the skull, right clavicle, rib injuries, right acetabulum with open reduction and internal fixation, compound comminuted fracture of the right patella with post open reduction, chronic low back pain syndrome, and history of right total hip replacement, but that he does not have an impairment or combination of impairments listed in, or medically equal to one listed in Appendix 1, Subpart P, Regulations No. 4." (TR. 26) The administrative law judge found claimant's subjective complaints of pain to be credible. It was further determined that the claimant has the residual functional capacity to perform the physical exertional and nonexertional requirements of work except for lifting and carrying more than 10 pounds. He is also limited in standing, stooping, kneeling, crawling, walking, standing, or sitting for prolonged periods due to his physical limitations. He is precluded from working around heights, moving machinery, or a polluted environment. The administrative law judge found that the claimant is unable to perform his past relevant work and, considering his nonexertional limitations, cannot be expected to make a vocational adjustment to "other work" and is disabled within the meaning of the Social Security Act. The administrative law judge found that claimant has been under a "disability" since August 1, 1990, but not prior thereto. That decision became the final decision of the Commissioner when the Appeals Council denied the plaintiff's request for review on January 1, 1995. The sole issue before the court is whether the final decision of the Commissioner denying the plaintiff's claim is supported by substantial evidence. In Blalock v. Richardson, supra at 776, substantial evidence was defined to be "evidence which a reasoning mind would accept as sufficient to support a particular conclusion. It consists of more than a mere scintilla of evidence, but may be somewhat less than a preponderance. If there is evidence to justify a refusal to direct a verdict were the case before the jury, then there is `substantial evidence'." Additionally, the Secretary, not the Court, is charged with resolving conflicts in the evidence. Nevertheless, the Courts "must not abdicate their traditional functions; they cannot *278 escape their duty to scrutinize the record as a whole to determine whether the conclusions reached are rational." Oppenheim v. Finch, 495 F.2d 396, 397 (4th Cir.1974). As a point of reference, it is pertinent to note that the plaintiff was originally awarded disability benefits commencing April 4, 1989, but when the favorable decision was processed, it was discovered that he had received $2,000.00 per month between December 1989 and July 1990. The Commissioner apparently concluded that plaintiff had engaged in substantial gainful activity. Plaintiff brought this action asserting that he made an unsuccessful work attempt, but was, nonetheless, disabled as of April 4, 1989. Therefore, the period under consideration is from April 4, 1989 until August 1990. A careful review of the record in this matter establishes that the Commissioner's final decision is not supported by substantial evidence. For the reasons stated more fully below, it is the recommendation of the undersigned magistrate judge that the Commissioner's final decision be reversed. Social Security Ruling 84-25 provides that an "unsuccessful work attempt" is measured by the duration and the conditions of the "work effort" and is used to evaluate any return to work after the first significant break in the continuity of a person's work. The ruling indicates that a work effort of 3 months or less which ends or is reduced to less than the substantial gainful activity earnings level due to the person's impairment or to the removal of special conditions is an unsuccessful work attempt. For "work efforts" of between 3 and 6 months, ending or reduced to below substantial gainful activity due to the impairment or to the removal of special conditions essential to the further performance, additional requirements must be met for the work effort to be considered unsuccessful: a. frequent absences from work, or b. work was performed unsatisfactorily due to impairment, or c. work was performed during a temporary remission, or d. work was performed under special conditions. The special conditions include special arrangements made to accommodate an impairment or an unusual job opportunity, and include a. special assistance from other employees, or b. leave to work irregular hours or take frequent rests, or c. special equipment or assigned work especially suited to the impairment, or d. was able to work only within a framework of special circumstances such as where other persons helped him or her to get to and from work, or e. was permitted to perform at a lower standard or productivity or efficiency than other employees, or f. was granted the opportunity to work despite his or her handicap, because of family relationship or other altruistic reasons. The ruling also specifies a "work effort" of over 6 months cannot be an unsuccessful work attempt regardless of why it ended. The plaintiff notes, however, the ruling specifies that it is the "work effort" that is being evaluated and not the period of earnings as the administrative law judge suggests. (TR. 25) The undersigned magistrate judge must concur with the plaintiff's assertion that it is the "work effort" that must be considered. The evidence of record indicates that in November 1989, plaintiff told his treating orthopedic surgeon, Dr. William G. Sale, III, that he had an opportunity to go back to work. In February 1990, plaintiff again told Dr. Sale that he was planning to return to work. Since plaintiff did not say that he had started such work, the clear implication is that he had not attempted such work opportunity. The undersigned magistrate judge has considered such documentation and concludes that, even though plaintiff told Dr. Sale in November 1989 that he was planning to return to work, he made the same statement in February 1990, and there is nothing in the record, other than the earnings posted for 1989 and 1990, which indicates that such work effort was attempted prior to February 1990. It is uncontested that the work activity ended by August 1, 1990, which represents *279 an actual "work effort" of less than six months.[1] The administrative law judge improperly considered the period of payment which did not entirely coincide with the work effort. In considering the work effort, the evidence of record further indicates that the work performed was under special conditions; that he was permitted to work irregular hours; that it was an unusual job opportunity because of a family relationship. Specifically, plaintiff's only job requirement apparently was to answer the telephone and the work activity involved 12 hours per week. He stopped because he proved to be a liability to the company. (TR. 178-179) According to the Report of Work Activity, the partnership had been created for him by relatives because of his physical and, apparently, financial condition. He also indicated that he reported such work activity to Social Security before he was approved for benefits. (TR. 181) In addition, the administrative law judge misapplied 20 C.F.R. § 404.1574(a)(2) and failed to evaluate the extent to which payments to plaintiff were not related to actual productivity but to a subsidy by the company. The regulation provides: Section 404.1574 Evaluation guides if you are an employee. (2) We consider only the amounts you earn. We do not consider any income not directly related to your productivity when we decide whether you have done substantial gainful activity. If your earnings are being subsidized, the amount of the subsidy is not counted when we determine whether or not your work is substantial gainful activity. Thus, where work is done under special conditions, we only consider the part of your pay which you actually `earn'. The evidence of record establishes that the earnings reported to plaintiff's record were most probably subsidized by his relatives. Specifically, plaintiff reported attempting to work 12 hours per week and earning $2,000.00 per month. (TR. 179) Such earnings appear inflated for the services rendered, i.e., answering the telephone. Furthermore, the record contains affidavits from plaintiff's business partners indicating that he did not receive wages for hours or days worked. The affidavit states that "any monies received ... were for his partnership in the company." (TR.186, 189) Clearly, the administrative law judge failed to properly evaluate to what extent the payments were actually related to plaintiff's productivity as opposed to being subsidized by the employer corporation. Inasmuch as the administrative law judge misapplied Ruling 84-25 by computing the six month time period according to the period during which payments were received rather than the period during which work effort occurred and since the administrative law judge failed to evaluate to what extent the employment earnings represented a subsidy by the employer and were not directly related to productivity and actually earned, the Commissioner's final decision is not supported by substantial evidence. Moreover, it is the finding of the undersigned magistrate judge that the plaintiff should be awarded disability insurance benefits, as originally awarded, commencing April 4, 1989. For the reasons state herein, it is, accordingly, RECOMMENDED that the defendant's motion for summary judgment be denied, the plaintiff's motion for summary judgment be granted, and the second decision of the administrative law judge, dated April 6, 1994, be reversed insofar as the plaintiff's disability onset date was set as of August 1, 1990, and the plaintiff awarded disability benefits as of the original onset date of April 4, 1989. Plaintiff and defendant are hereby notified that a copy of this REPORT-RECOMMENDATION will be submitted to the Honorable Charles H. Haden II, Chief Judge, and that, in accordance with the provisions of Rule 72(b), Federal Rules of Civil Procedure, the parties may, within 13 days of the date of filing of this REPORT-RECOMMENDA- *280 TION serve and file written objections with the Clerk of this court identifying the portions of the REPORT-RECOMMENDATION to which objection is made and the basis for such objections. The Judge will make a de novo determination of those portions of the REPORT-RECOMMENDATION to which objection is made in accordance with the provisions of 28 U.S.C. § 636(b) and the parties are advised that failure to file timely objections will result in a waiver of their right to appeal from a judgment of the district court based on such REPORT-RECOMMENDATION. Copies of objections shall be served on all parties with copies of the same to Judge Haden and this magistrate judge. The Clerk is directed to file this REPORT-RECOMMENDATION and mail a copy of the same to all counsel of record. DATED: September 8, 1995. NOTES [1] Plaintiff asserts that work ended July 14, 1990. In either case, the February 1990 until August 1, 1990 period still represents only six months.

To celebrate National Picnic we’ve created a map of the UK containing all our favourite picnic spots and we’ve also decided to give away 5 Icemax picnic coolbags! All you have to do is head over to our Facebook page, find our picnic map, recommend your favourite picnic spot and get as many people as...CONTINUE READING

This global FTSE 100 is currently undergoing an ambitious Global Transformation in their Procurement Function, aligning all their global procurement activities out of their London Headquarters and transforming their Procurement Function into a world-class team of over 150 individuals. As a result of this transformation, Marketing Procurement is naturally at the forefront of the Global change, and as such they are looking for a talented and ambitious Procurement professional to join and specialise within IS Infrastructure as they go on this global journey. What will this role offer me? The chance to work closely with Senior Procurement Managers with Blue Chip backgrounds to develop global procurement strategies for specific spend areas, leading on £Multi-Million tenders Witness and influence first-hand the globalisation and transformation of a FTSE 100 Procurement Function Rapid progression with mentorship schemes and the opportunity for CIPS Sponsorship as you are developed into a future procurement leader of the business What else will I be doing? Building relationships with Director-level Marketing Stakeholders both in the UK and internationally to identify the key business needs and work in partnership with them to develop best practice procurement Managing contracts and leading tenders within Marketing, including Media and TV Production Collaborating with cross-category teams to broaden your experience across the indirect categories, particularly IT and Digital Innovation Engaging critical suppliers to develop and manage commercially fruitful relationships in the long term, including reviewing key contracts and associated risks and indemnities What are they looking for? Experience across the full end to end procurement process; from RFQ through tender, contract award and post contract supplier management, preferably in a Blue Chip environment Experience in marketing procurement is essential, either as part of a generalist indirect background or a specialism in marketing procurement Individuals from a Procurement Graduate Scheme background are highly preferred A strategic, commercially aware Procurement professional with experience of influencing Senior Level Stakeholders, with a desire to progress and help lead change in a FTSE 100 To express your interest in this position, please send a copy of your CV to Emily at[email removed] or apply directly through the job board.

Gareth Edwards to Direct ‘Godzilla 2′ After ‘Star Wars’ Spinoff Reboots of popular film titles can go one of two ways: they can either take the best parts of the original and spin it in an exciting new direction or they can completely miss the mark. Thankfully for director Gareth Edwards, his new reboot of Godzilla was able to intrigue audiences enough to become king of the box office this past weekend to the tune of $93.2 million, the second highest stateside opening this year after Marvel’s Captain America: The Winter Soldier. After filming documentaries, a TV movie and a short, Edwards first broke onto the film scene in 2010 with a small science fiction thriller, Monsters, before crossing over to the mainstream and making Godzilla. Now the British director is on his way to helming more box office tentpoles, including the just announced Star Wars spinoff, due in December 2016. According to THR, today’s Star Wars spinoff news means the previously confirmed Godzilla 2 sequel, which Edwards is also attached to direct, will not begin filming until after the Star Wars standalone film is completed. In case this information is a little discouraging for Godzilla fans, apparently, as per Legendary Entertainment’s chairman and CEO Thomas Tull, this was always going to be the case: “Gareth’s filmmaking talent makes him one of his generation’s most creative and visionary directors. The plan has always been for Gareth to direct a different film before we started on another Godzilla, but who knew it would be a ‘Star Wars’ installment? We have a great plan in store for ‘Godzilla’ fans and I am looking forward to seeing Gareth’s imprint on the ‘Star Wars’ universe.” As is pretty standard for films with sequel potential, story ideas for Godzilla 2 and even Godzilla 3 already exist, but no scripts have been written yet and Godzilla 2 is still in early development. So this news allows at least a year for Legendary to hire a writer and complete a script before Edwards can even begin to think about returning for a sequel. Basically, don’t expect another Godzilla until 2017 at the earliest. Of course, this only matters if Edwards remains attached to the Godzilla sequels. At this point, that remains the case, but it’s not hard to imagine scheduling conflicts if the release dates occur within a year of each other. Problems could push Godzilla back even further as Star Wars takes precedent, or Legendary could simply decide to change director. It should be interesting to see how the timing all works out. Are you happy that Godzilla 2 and 3 are still in Edwards’ future? Are you sad that we’ll have to wait at least three years for another film? Let us know in the comments. Godzilla 2currently has no release date. Stay tuned to Screen Rant for the latest.

The Sundering Worlds Both literal and figurative The adventurers decided to leave Elzalath outside the village to scout and to avoid unpleasant introductions. Corvus and Peke made their way to the city gate where they were greeted with nocked arrows and barked interrogatives. However, Clan Rooster accepted them quickly when they discovered the heroes made it through the forest at night. They were introduced to the leader Shizue, who was keenly interested in how they overcame the wights in the forest. Peke managed to dodge Shizue’s questions, not intent to give away Elzalath’s presence unless absolutely necessary. Shizue was very straightforward, revealing the tale that led to Endo’s assassination. A Clan Crane Ambassador kidnapped, raped and killed Shizue’s daughter, Shia, heiress and rightful Emperess-in-Waiting of Clan Rooster. The violent and tragic circumstances of her death manifested her spirit as a malevolent Yokai. She haunted the village of Clan Crane, enacting her vengeance by harnessing the power available through the rift. Additionally, the wights of the forest were fueled by her unrest. She killed Endo as vengeance. Meanwhile, Elzalath explored the surrounding region. She discovered an emergency exit from the city perhaps used to evacuate certain members of Clan Rooster in times of crisis. On her return to the front of the city, she spotted a wight flitting through the trees. Elzalath snuck up on the spirit and harnessed the Edict of Nhalgos to bind it. Through interrogation, Elzalath learned that this phantasm was subject to a greater will than her own intent on bringing destruction. The team reunited and rested within the confines of Clan Rooster. In the morning, they made their way uneventfully back to Clan Crane. Upon entering the compound, it was immediately apparent something significant had transpired in their absence. The once bustling village was empty and silent. The troupe noticed oddities emanating from the Elder Temple in the center portion of the compound, so they made their way there. At the entrance, the door was hanging from its hinges, and Elzalath could sense the presence of an intense aether trap. In an attempt to be cautious, Peke summoned a dryad to enter the temple first. The dryad walked in unimpeded. Then Peke attempted to conjure countermeasures to disarm the trap but failed. Corvus asked if his mysterious spikey cypher could be of any use. Peke’s patience had run out, so he took the enigmatic device, yanked on each spike, and tossed it into the temple. A horrific bang shook the temple as all magic embedded in the device was ripped away. The blast knocked Peke back 10 meters. Peke shook off the pain, as Elzalath noticed the dissipation of the force behind the trap she sensed earlier. The group headed into the Elder Temple. Inside, they found another empty room with a staircase. Peke sent the dryad up the staircase, but as soon as it touched the first step, the shrieking spirit of Shia came barreling down from above toward the dryad. Elzalath shouted at Shia that they could work together. Together, they agreed the best course would be to shut the rift and bring Shia to the Lich Lord to add to the ranks of his army. Unfortunately, Shia was bound to the Temple until they resolved the rift problem. Upon learning where Shinzo Abe was hiding out, the adventurers headed to confront him. They found a morose Abe sitting at the center of the room in a large cloak. Through questioning, they discovered Abe literally had the rift in a device in his pocket. Unfortunately, he had a score to settle. He shed his cloak, unsheathed his sword, and engaged the trio with the exceeding grace and precision of a lifelong warrior. Corvus sprang to action first, swinging his trusty blades at Abe, but Abe effortlessly dodged his way around the blows. He sliced through the remaining meat on Elzalath’s forearm. But Elzalath savored her opportunity to engage Abe in combat and blasted him with a Withering Curse. She could feel the life force slowly draining from Abe and empowering her. She held her arm out for him to see as her wounds healed themselves. Peke transformed into a hawk to fly out of range and retrieve the rift. K’ot snarled at Abe to distract him from Peke. Her snarl worked too well, however, as Abe’s blades dug into her flesh. Meanwhile, Corvus snuck over to the cloak, quickly retrieved the rift device, and popped it open to unleash a portal spewing tainted aether. When Peke saw his dearest friend suffering at the hands of Abe, his fight left him. He dove toward the portal and placed one foot on either side, ready to jump in. “Elzalath, enough! We have to go! This fight is beneath you. All that remains for us on this plane is suffering and death.” Elzalath stood firm. “What about Shia? This wasn’t the plan.” Peke replied, “This isn’t what we hoped it would be. It’s too late. We need to go.” Will Elzalath forsake her quest for power in exchange for survival? Will Peke abandon one of his closest friends? Do fucking major and minor effects apply on goddamn defense rolls? Tune in next time to find out! The trio decided to exercise an increased measure of caution as they set out to explore this new realm. Peke dug his fingers into the loose soil at his feet to question the living world. He could feel the damage the tainted aether was causing. He also found no concerns about the warrior on the hill, so he walked to the man while his companions set up camp. The man on the hill introduced himself as Shinzo Abe of Clan Crane. After formalities, Peke felt Abe could be trusted. The adventuring party accompanied Abe back to Clan Crane to meet their leader, Endo. Endo proved cagey at the party’s desire to close the rift and move on. He viewed the arrival of these planeswalkers as an opportunity to solve the clan’s problems with their rival, Clan Rooster. In the midst of their discussions, an arrow flew through the crowded room and killed Endo. Elzalath attempted to gain more insight by talking to Enzo but Abe forbade communication with the dead. Peke and Corvus attempted to identify the culprit, but none could be found. The crew agreed to talk to Clan Rooster to try to figure this out while truly intending to find the rift on their own. They had enough of getting involved in the affairs of the ungrateful planes they needed to save. So they headed off to Clan Rooster to see if they could gain more information about the rift. In the woods, malevolent specters began flitting in and out of view. When one spirit seemed hostile enough to attack, the team attempted to get rid of her. Unfortunately, their attacks only empowered the poltergeist further. When all other options were exhausted, Elazalath called on the Lich Lord for assistance. He delivered them to the edge of the forest in sight of Clan Rooster’s compound. What new intrigues will our heroes find at Clan Rooster? What malevolent force can harness and manipulate our heroes’ power? Where is this dang rift anyway? Tune in next time to find out! 1 down, 6 to go Our heroes found themselves facing the Head of the Dispossessed, Aysha, and her right-hand man, Lothar. The Weaver continued to mentally commune with Elzalath as it hid in the shadows. Elzalath was the first to move. She reanimated the vemon-riddled corpses of their former allies. Upon revivification, the zombies started moving toward Aysha. Meanwhile, Corvus felt the weight of indecision and slunk into the shadows of the cavern. Pekeporo still felt torn over the desire to give the disenfranchised access to equality, but a vision of Myrna being torn apart by tainted aether moved his hands into the dirt to summon a dryad familiar in anticipation of impending combat. Corvus nocked an arrow as Peke’s dryad familiar charged Lothar alongside K’ot. Unfortunately, years of helping facilitate rebellion made him quite a skilled foe. He effortlessly dodged both attacks and crushed the back of K’ot’s head with his gauntlet. Peke’s fury at this blow quickly manifested into a beam of piercing light that lanced directly into Lothar’s exposed side. Meanwhile, as her arm still crackled from the tainted aether well, Elzalath unleashed her fury against Aysha. Her zombies closed the gap and voraciously attempted to sate their appetites on Aysha’s flesh as decay erupted from Elzalath’s good arm and struck Aysha. Upon seeing his companions embroiled in combat, Corvus released his arrow into Lothar’s shoulder. “ENOUGH!” shouted Aysha. “We are your allies! We can explore this rift together. There is no need for this.” Peke sensed Elzalath’s resolve. He envied her clarity as he waded through the murk of moral ambiguity. Peke knew that he could not accompany Elzalath to the completion of her journey, but for now, they were on the path together. He unleashed another destructive blast on Lothar as his dryad and K’ot continued to demonstrate gross incompetence. Elzalath’s zombies were upon Aysha now, drooling necrotic venom all over her. Corvus considered his options and went all in with his companions by tossing an explosive cypher into the fray. The cypher blew everything to hell. But Aysha felt the worst effects as the blast fused the venoumous zombies to her flesh. In a desperate attempt to continue, she dove into the aether rift. Unfortunately, the blast also reached the Weaver. Elz’s mental connection to the Weaver immediately snapped as it descended in a fury to devour the fools who would allow the blast to touch her. However, the heroes were emboldened by their recent martial success and quickly made the Weaver regret crossing them. They also received a bit of assistance from spearmen in the shadows, but it’s unlikely they really needed help. Upon destroying the Weaver together, the cavern started shaking and rocks began falling from the canopy. The soldiers thanked the heroes for their assistance with the aether and stepped through a portal back to the safety of their castle. Elzalath made a psychic inquiry of the aether buzzing around her arm and discovered that Peke was the key to closing the rift. So Elzalath, Corvus and Peke jumped into the rift in order to seal it off forever. As they were travelling through all the planes, each adventurer had a vision. Peke learned from the Jade Emperor that his body is the key to closing these portals, and there are seven of them. Elzalath received confirmation from the Arch-Lich that she could find a way to harness the aether for god-like power. Corvus had a terrifying encounter with Asmodeus, who bragged about his ability to find Corvus in any plane when it was time to collect his due. The planar jump tossed the group into a field where they quickly assessed their surroundings. Corvus found the charred remains of Aysha. Clearly the jump had not been so kind to her. In the distance, they could see a city and in the opposite direction, a man practicing martial arts on top of a hill. Peke communed with the grass surrounding them to learn that the man is not an adversary. The crew decided Elzalath and Corvus would set up camp out in the plains while Peke sets off to talk to the man on the hill. What exciting adventures will our heroes find on this new plane? Is this the fateful time in which Peke will lose his eye? How much time does Corvus have? Will Corvus and Peke’s controllers make time to actually level up their characters? Tune in next time to find out! 404 - File Not Found Lo, this unworthy scribe must beseech the dear reader for forgiveness for allowing nearly two fortnights to pass before continuing his tale. Thus, the details of our heroes’ journey to the cavernous depths in search of the tainted aether rift are largely lost to the ravages of time and a foggy memory. Yea, a reader might even question the qualifications of a bard who, through his own malfeasance is unable to fulfill his primary duty as a weaver of tales. While sound in its reasoning, this criticism falls fatally to the dearth of bards in the present age. When taken to its logical conclusion, this line of reasoning concludes in stating that you, beloved reader, can fuck right off, because I’m the only game in town. So the odd trio of adventurers decided to go investigate the tainted aether. They ran into some weird underground society who needed help getting rid of a giant spider-demon with nasty venom spits. They sent some of their own along with the team to destroy the Weaver, but Elzalath made a mental connection with the Weaver by leveraging its connection to that one death-god we all know and Elz loves. Then we found the aether rift, Elz stuck her arm in it to gain a shitload of power, and the rebellion lady and her helper came to try to figure shit out with the heroes. The crew simultaneously suspects the cavern is rotten with the city guard folk as well. Will our heroes get out of this jam? Will they empower the disenfranchised or save Faerun? Will I write the next play session in a timely enough manner to actually include details? Tune in next time to find out! Session 13 After arriving in a strange new world, the PCs retired to a nearby inn in the hopes of recuperating their senses. Before long, however, an explosion rocked the city square and overturned the city of Areydan. In the chaos, a hooded figure confronted the PCs. He said his name was Mythar and that he was an autarch participating in the dispossessed rebellion against the Fhrey. Deciding to trust Mythar, the PCs followed him to the hideout of the dispossessed where they met with Aysha, the leader of the dispossessed, who spun a tale of oppression and deceit. The Fhrey, she claimed, rule Areydan with an iron fist. They are aether wielders who oversee all of life in Areydan, which has led to a society split in two. The Fhrey rule the dispossessed and the dispossessed are the slaves of the Fhrey. It is Aysha’s hope that the PCs will help the dispossessed ignite a war against the Fhrey, so that Areydanian society will reset itself. Aysha then told the PCs of a great rift outside of the city wherein the aether of the Fhrey is seeping out. The people of the city now have free and fair access to the magic that has so longed oppressed them. The rift, Aysha tells the PCs, must not be closed. It must, if anything, be permanently opened so that the Fhrey will be held accountable. Later that night, however, the PCs were approached by a man named Fingol. He is a Fhrey and a lieutenant of the Archon, Elise. He told the PCs a vastly different version of events wherein Aysha is a usurper looking only for power and prestige. In order to maintain the world balance, Fingol asked the PCs, on behalf of Elise, to travel to the great rift and close the aether slippage. In order to test their own theories, the PCs then traveled Areydan and asked the dispossessed if their lives were worth living. The PCs, unfortunately, received mixed feedback, and so, they realized, the choice was theirs: side with the dispossessed to overthrow the Archon or travel to the rift and seek to close that which is destroying the planar realms? The Story So Far “Sit, sit,” she said. "Gather ‘round the fire. That’s right. That’s right. Now, quiet down. Yes. Just so. Just so.” The fired danced around the story teller’s face. Her gray hair glistening in the moon light. “Are you ready?” The crowd nodded in anticipation. “Long ago,” she began, “there was a nameless one, called by the gods, who ventured forth from his enclave. He was a traveler, of sorts, who, along with another, decided to accompany the plentiful goods of Neverwinter to a southern village named Phandalin. Once there, however, these two, Pekeporo and Corvus, found a city in disarray. A band of evil men, orcs, and half-elves had taken ownership of Phandalin. But this was nothing to our heroes, for they soon dispatched the invaders deep within their hideout and came to befriend a drow elf, a paladin named, Elzalath.” “Is that Elzalath the forsaken?” a man shouted just beyond the storyteller’s vision. “Just the one,” she nodded. “But do not outpace the very story I am weaving.” The storyteller paused to let her audience adjust. “Yes, it was Elzalath—that troubled soul. She had been taken prisoner by the Red Guard, but not for long. Pekeporo and Corvus rescued her and—” “And saved Phandalin?” a child near the storyteller asked. She shook her head. “No. No, that is not the story I have to tell. In fact, the city fell to a drow army. It was overran soon after a rogue general sided with the drow against Neverwinter. But that is another tale, for another time. The reason our adventurers—the three you know so well—were not present was because Corvus’ adopted parents had been taken prisoner in a village just outside of Thundertree.” “And K’otquchoq saved them? Isn’t that right?” The storyteller stared at a young woman whose face was aglow. “Is that what you heard?” “Yes,” she said, abashed. “I did.” “No,” the storyteller shook her head. “The wolf is a friend of Pekeporo, that is the truth, but a tortured animal. A beast of mixed passions. He was there, to be sure, but he was not alone. The three adventures, and their beast companion, saved Corvus’ family and unlocked the secret of Mandibar’s Mirror, but not before the God of Choas, Asmodeus, touched Corvus.” “His brand, you mean?” another shouted. “Aye, that. He was a haunted man, after. But that comes in later. Saving his parents did him well, but it did nothing to alleviate the dreams that Pekeporo was having, given to him by the God the Verdant Temple.” “Is that why he lost his arm?” “Hush, child!” the storyteller said. “You know not of what you speak. Pekeporo is a great druid, perhaps the greatest throughout the Great Wheel. He slayed the aboleth and traveled across worlds. Be respectful, if nothing else.” She paused again and waited for the attention of those gathered around her. “But, yes, they quickly traveled south to assuage Pekeporo’s dreams. With the help of a well-known sea captain, Pekeporo, Elzalath, and Corvus traveled into the bloodmists and the Moon Isles. Once there, they were told the most important thing that they would ever hear: the aether between the worlds was slipping. The realms of the Great Wheel were sundering.” A silence fell over the crowd. Not a whisper could be heard over the crackling fire. “And so,” the storyteller continued, “they met with the Emerald Emperor and traveled to another world. And, though they did not yet know it, their mission was simple, to seal the cracks of the planar realms, to heal the wounds of the cosmos.” “But Elzalath?” one said. “She was transformed.” “Aye, that is true. And even then,” the storyteller said, “no one knew which way she would turn.” Session 12 With the prompt dispatching of their martial foe, Corvus, Elzalath and Pekeporo set out for Leilon. Upon reaching the bustling borough, they made their way to the docks to seek passage south. After some guidance, they found Wick, Captain of the Lost Mermaid. He agreed to provide passage to Waterdeep despite his hesitation with boarding a drow. The first days of the voyage passed without incident. The adventurers passed the time in their cabin until Captain Wick paid them a visit. As a merchant, he hoped to enlist their help through the dangerous voyage past the Mere of Dead Men. They agreed to provide assistance on the voyage as a basis for developing an ally on the seas. On the third day, Pekeporo was sitting in the crow’s nest in the middle of the night. On the deck below, K’ot’s ears suddenly perked toward the stern, so Peke looked carefully at a rising wave. Almost too late, he realized a terror from the deep was upon them. As he sounded the alarm, Peke thrust an entanglement spell, but the beast was too elusive to be bound. Captain Wick, his first mate, Elzalath and Corvus emerged from their quarters as massive tentacles emerged from the depths, surrounding the vessel. From learning an ancient legend of a master crippling his wayward pupil by gaining the higher ground, Corvus bounded up into the crow’s nest. Elzalath recognized the potency of their foe and quickly tagged it with her Hunter’s Mark before slicing into one of its tentacles with her Divine Smite-enflamed rapier. The beast retaliated by spraying the entire boat in a disgusting ooze. Corvus and Peke shook it off, but something evil in the vile sludge penetrated Elzalath’s psyche. Then the beast brought its tentacles down on the ship, blasting Elzalath against the deck. Corvus and Elzalath unleashed flurries of arrows at the writhing tentacles as K’ot did her best to rend flesh from them. Peke despaired at the seemingly impossible task of felling such a creature through mere damage to extremities. He steeled his nerves, having grown up in a land-locked woods after all, and leapt from the crow’s nest toward the water. As he plummeted, he transformed into a giant squid, plunged into the icy depths and propelled himself at the bulk of his foe. After pummeling the beast with his freshly sprouted tentacles, he wrapped it up in a giant squid hug. As the boat glided off into the night, Peke took comfort in knowing his companions were safe for the time being. Simultaneous to this relief was the dreadful realization that he was now single-handedly combatting an aboleth. Thanks to his uncle’s tales, Peke knew the potential for this beast to destroy not only his body, but his mind as well. “Perhaps if I speed to shore, the aboleth will chase me in an attempt to gain vengeance for removing it from its quarry. Then I can make my way south on land to rejoin my companions.” As Peke swam for shore, his best laid plans were immediately undermined by the aboleth’s determined and renewed pursuit of the Lost Mermaid. The aboleth swam with such speed and fury just below the surface, that a wave rose in front of him. Fortunately, the heroes on the boat did not view their reprieve from battle as an excuse for complacency. Elzalath spotted the approach from a distance and pulled her holy talisman from her cloak. As she bellowed her divine utterances, the token shone brilliantly and blasted the unholy creature in its pursuit. Suddenly, the wave died. The aboleth continued pursuit, but its pace was greatly reduced. Peke built all the squid speed he could muster to catch up to the aboleth. He leapt from the water, transformed back into his natural form and blasted a wave of thunder from his fingertips as he skimmed across the surface of the water. The aboleth maneuvered around much of the wave, but the spell still took its toll. As he lost momentum, Peke transformed back into a giant squid and slid under the surface. Corvus jumped on the opportunity to shoot the aboleth in the face with some arrows. His shots flew true and buried deep in the meat of the beast. The creature retaliated by slamming the ship again, this time destroying the crow’s nest. Fortunately, Corvus anticipated the blast and majestically dove at and clung to a hand-hold high above the ship. Once again, Peke pounded the aboleth with his tentacles and restrained it from further movement. Elzalath spotted its eye amidst the writhing tentacles and released the death blow. As sickening lifeblood poured into the sea from its eye, the aboleth’s corpse sunk down to the murky depths. The journey to Waterdeep continued without further disruption. Upon arrival, Captain Wick offered the adventurers any favors they needed, recognizing the lifedebt he owed these three. He agreed to transport them to the Moon Isles, but he needed two days to repair the damage the aboleth had dealt the Lost Mermaid. The trio headed into Waterdeep to kill some time and ready themselves for the next stage in their adventure. What dangers lie ahead at the Moon Isles for our heroes? What adventures will emerge in Waterdeep? What is the Jade Emperor doing while he waits for them to arrive? Sniffing buttholes? He’s probably sniffing buttholes. Tune in next time to find out! Session 9-11 Our trio made their way to Waterdeep or Neverwinter or something for some reason. While there, they did some research on the Jade Emperor and the Verdant Temple. They stole a book from a lady to get this librarian lady to help them out. Meanwhile, the drow lady kept following them around, because she still wanted Elzalath to come with her. When the team left the library, some general wanted them. They tried to run away, but the librarian lady turned out to be a High Sorcerer of the city or some such title, and she prevented them from running. They met with the dude and agreed to help the battle against the bad drow army and joined some order of the city that will hook them up from now on. Then they went to the sewers to follow some guidance they got from somewhere. They found a powerful-seeming drow dude in a room with floating shit. They became pals and agreed to all go together to infiltrate the drow army or something. On their way south, Pekeporo had his most urgent vision yet, and attempted to abandon everyone to go to the Moon Isles with K’ot. But his pals heard him leaving and decided to go along. But then the army leader fought them for some reason. They whipped his ass and he said something about a Sorcerer or something that showed he’s bad and things are corrupt. Will our heroes make it to the Moon Isles? Are they walking into a trap? Will Ryan write these in a more timely fashion so they aren’t insanely vague and brief? Tune in next time to find out! Session 8 Corvus emerged from the abandoned home of his childhood mentor. Pekeporo met his eye and immediately understood Corvus’ despair. Elzalath slunk into the town’s periphery, recognizing her outsider status as a drow in this village. Time was of the essence and the group had no leads. orc-gruumsh01-1Corvus and Peke headed to the village blacksmith to see if he had any information. The kidnapping and arson happened so fast, there were only rumors and hunches to be had. Yet among this hearsay, Corvus discovered a roving band of orcs calling themselves The Blackbird Red were responsible for this nightmare. He couldn’t say for certain, but the blacksmith thought they were probably posted at the Ruins in the surrounding forest. Corvus thanked the man for the information. Meanwhile, Elzalath wandered the surrounding forest looking for clues. She sensed a presence and whirled to spot a fellow drow in the clearing. “Elzalath, my name is Jaelryn, and I come on behalf of my master, Elendar. He sent me with this letter.” Elzalath knew enough of the ways of the drow to find this encounter quite unremarkable. “Elendar seeks what you seek and desires your tutelage. He can provide what you desire.” Elzalath accepted the letter and Jaelryn disappeared into the brush. Skeptical, Elzalath tracked Jaelryn’s movement far enough to identify her movement west toward Neverwinter. This was certainly an intriguing proposition, but Elendar could wait. Elzalath had developed a strange and unexpected affinity for her journeying partners and wanted to help Corvus see this through. The party reunited in the village and decided to confront the Blackbird Red head-on in the morning. After an uneventful rest, they journeyed to the forest ruins. They decided to rest before their confrontation, and Peke meditated fitfully. At the end of his meditation, he received the most powerful vision he had yet encountered. This time, the voice was understandable, admonishing Peke to seek The Green Emperor in the Verdant Temple. While still cryptic, Peke began to understand that his namequest would take him beyond every reality he had ever experienced. In the morning, the adventurers finished their journey to the Ruins. Peke took advantage of the opportunity to show off the new spell he had been studying and cast the entire party into a veil of invisibility. Easily undetected, they found two watch-orcs high in a treenest. With incredibly precision, they unleashed a trio of arrows at one of the orcs, all three of which sunk deep into the orc to send him toppling out of his perch. Without hesitation, all three heroes immediately released another volley of arrows at the other orc. The first orc hit the ground in a burst of gore and viscera. The second orc managed to bone-crunchingly land on his feet, just conscious enough to taunt the trio with a warning of one-eyed Agar waiting for them. Then he breathed his last. The crew made their way to the clearing leading to the Ruins to find a terrifying scene. Four orcs awaited them, two of which held Corvus’ father and mentor at knife-point. Between them stood a giant of an orc with a patch over one eye. Corvus’ mother was restrained behind his enormous arm with a knife to her throat. This orc had to be Agar, and he was waiting for them. Corvus, Peke and K’ot remained shrouded and slowly snuck up behind the orcs around the backside of the ruins. Meanwhile, Elzalath emerged from Peke’s spell to attempt negotiation with the orcs. “Who are you?” growled Agar. “I’m a drow, and we share a common purpose. Perhaps if we work together, we can successfully enact vengeance upon the scoundrel Corvus Cade.” Elzalath was so convincing, Peke began to worry that Elzalath had been stringing them along this whole time. He certainly didn’t know anything about Elzalath’s life before they met. Agar chuckled deridingly. “This isn’t about revenge.” “Then what is it about?” “Service. I simply follow the will of Asmodeus. He desires Corvus’ destruction, and I am happy to oblige.” “In that case, we remain allies as I serve Elendar.” At this, Agar grew quite agitated. He pointed his dagger at Elzalath. “You are no acolyte of Elendar. Now out with the truth.” Corvus sensed the exchange was headed south, and took advantage of the opportunity to attack Agar while his blade was away from Corvus’ mother’s throat. He drew his swords and unleashed his fury on Agar. The trap was sprung. Elzalath took the cue and cast a Blessing on her companions for battle while eviscerating the orc holding Corvus’ father. Peke used the advantage of surprise to Wild Shape into a brown bear. In a flurry of teeth and claws, he flayed Agar’s back as K’ot put all her strength into knocking Agar over. From the prone position, Agar had no hope of turning the tide of battle. Even worse for him, as he struggled to right himself onto one leg, he saw his faithful followers scatter into the forest. “Goddamn cowards.” Sensing that Corvus had the situation well in hand with his blades at Agar’s throat, Elzalath, Peke and K’ot leapt into the woods after the fleeing orcs. Elzalath and Peke subdued their quarries while K’ot demonstrated no such restraint. She sensed Corvus’ hatred of orcs in general and took this opportunity to decimate her prey. Meanwhile, Corvus interrogated Agar to no avail. He held his sword and a potion of healing in front of Agar’s working eye. “We can do this the easy way or you can die.” “Fuc…” Before Agar could finish his response, he choked on a geyser of blood that erupted as Corvus buried his blade in Agar’s throat. Corvus vaguely sensed the return of his companions but could only muster the strength to gratefully hug his family and mentor. In the group hug, he smiled as he noticed K’ot trot up and drop an orc dick at his feet. Peke attempted to interrogate his captive to no avail. He paid attention as Elzalath demonstrated how interrogations work in Faerun. She enchanted her sword with a white burning heat. She held her blade over her captives head as she stuck her face inches from the orc’s. “We need information, and let me assure you, you will not die quickly if you refuse to cooperate. You will not be so fortunate as your master. I have all day and endless creative ways to make you suffer. So measure your response carefully. Where is Asmodeus?” As the stench of waste emanated from his trousers, the orc was clearly terrified. “I…I am a lowly minion. Agar didn’t trust me with any information. B…But he could read and write. He kept a journal. It has to be around here somewhere. I’ll help you find it.” Elzalath lowered her sword while maintaining eye contact. The blade slid down further until it severed the bonds at the orc’s feet. “Good answer.” She helped the orc to his feet as Peke looted the ruins for treasure and Corvus held his family. Will our heroes have what it takes to confront the threat of Asmodeus head-on? What secrets are our adventurers keeping from one another? Why is the Dungeon Mistress trying to give away all of Peke’s secrets? Tune in next time to find out! Session 7 After Peke checked in with Myrna and Corvus and Elzalath went to patch things up with Phandalin’s inn-keeper, the crew decided it was time to confront Porcius’ irritating bounty head-on. After devising a slap-dash and largely unnecessary plan to confront Porcius, the ventured out toward Corvus’ hometown. On the first morning of the journey after a night’s rest, the adventurers discovered an ominous dark object. Elzalath immediately recognized its demonic origin and was hesitant to bring it along. But where Elzalath saw danger, Peke saw power, and he convinced her to bring it along. The party stayed away from cities to avoid suspicious glances. Upon awakening the second morning of their journey, they found themselves back at their camp of the previous night. Ignorant of whether they had been transported through time or space, Corvus and Peke were sufficiently convinced to heed Elzalath’s caution and leave the damned object behind. After repeating their journey of the previous day, they once again made camp outside Corvus’ town. The following morning, when no ineffable teleportation took place, they headed into town, straight to Porcius’ dwelling. Their ruse worked flawlessly as Porcius welcomed them in, believing Corvus to be captured. When he had his goons start escorting Corvus away, the trap was sprung. Corvus filled the room with fog as he grabbed his swords. Peke cemented their advantage by casting Faerie Fire on their opponents. Elzalath and Corvus made shockingly short work of Porcius’ goons, basically pulling an organs-on-the-outside-of-your-body-now situation. Meanwhile, Porcius grabbed a small object out of his desk and tried to escape. His voluminous body made escape a slow and laborious process, and Peke had plenty of time to slip into a form a bit more serpentine and put the squeeze on Porcius. As the fog cleared, Porcius accepted defeat and demonstrated his true impetus for seeking Corvus. “This is much bigger than you can imagine. I was only doing what Asmodeus made me do. Please let me go. It won’t end with me. You’re screwed and so is your family.” In Porcius’ hand was a small glass bird containing a swirling black cloud. Elzalath bristled at the evil it contained. She recognized the dark machinations of a object of Asmodeus. Even the thought of that name filled her with righteous rage. K’ot also sensed the evil in Porcius’ hand. She instinctively attempted to separate Porcius from the glass bird by biting off his hand. But in doing so, her teeth came down on the bird, shattering its fragile body in her mouth. An ominous cloud burst from her mouth, filling Porcius’ cavities and melting his entire head into a gooey blob. Simultaneously, a burst of energy erupted from K’ot, turning her greay coat deep crimson and seemingly sucking the life from her. Pekeporo blanched. He ran to her side as the cloud dissipated in the room. K’ot was barely breathing, but Peke dug deep in her psyche and found a remnant of her former, virtuous self buried beneath the fog of evil. He recognized what she had become…a Changer. He remembered sleepless moonlit nights in the Wood of Sharp Teeth, keeping watch with T’Allius for the Changers who would mindlessly eradicate life throughout the Wood. “What happened to her?” Elzalath’s seemingly benign question snapped Peke back to reality. But he had known Elzalath long enough to know the real question she was asking, and K’ot’s life was far more important than some holy agenda to eradicate evil. “She’s hurt is all. That cloud sucked the life out of her.” “You sure?” “Positive. She’ll be fine.” Peke healed his lifelong companion and they walked outside together. When Corvus and Elzalath rejoined them, the group decided to make haste for Corvus’ home. After their recent ordeal, Porcius’ threat seemed far from empty. They ran to Porcius’ home to find it ablaze. The townsfolk had no helpful information regarding the fire or Corvus’ family. Peke and Elzalath helped the residents put out the fire while Corvus ran to his mentor’s home, hoping to find his family. Instead, he found an empty house with another glass bird sitting on the doorstep. He picked it up and swore vengeance upon Asmodeus. Have our heroes stumbled into their greatest foe yet? What will happen when Elzalath learns of Pekeporo’s deceit? What’s the deal with all these glass birds? Where’s the Faerun Seinfeld when you need him? Tune in next time to find out! Session 6 After a bit of persuasion and payment, Halia outfitted Elzalath with her weapons of choice. The trio headed to the inn for some much-needed rest. The inn’s tavern was largely empty, with a lone figure drinking silently in the corner. Pekeporo’s ignorance of social custom shone through as he attempted to befriend the gruff stranger who quickly dismissed Peke’s congeniality. After some drinking and good-natured rabble-rousing, the adventurers headed off to bed, using their Bree technique. But this time, their adversary anticipated their schemes. “Corvus Cade. Your time has come. You can come quietly and nobody gets hurt.” The gruff man’s voice rang through the hallway as he emerged from the stairway. Another figure appeared from behind to surround them. Corvus reached his limit on paying off Portius’ goons, so he drew his swords. 3 on 2 with a trained draw on their side? This would not be a fair fight. And it wasn’t. After slaughtering the unknown figure behind them, the trio allowed the gruff figure to live when he dropped his false bravado. He was simply an aspiring adventurer on a hunt for gold. Despite Elzalath’s misgivings, they decided to allow him to live and accompany them back to the Redbrand hideout as their meat shield. The innkeeper entered the hall and was horrified by the grisly scene. HIs anger was dwarfed by Elzalath’s rage that he allowed this to happen. He stammered the impossible situation he was placed in and in his outrage, he kicked them out of his inn. He didn’t realize his good fortune at escaping the situation with his life. The heroes along with their captive journeyed back into the Redbrand cave. As they entered, their captive felt the oppressive mental prison Corvus experienced on their first entry. Fortunately, this left Elzalath, Corvus, Peke and K’ot the mental clarity to recognize the foe across the chasm that was the source of the anguish. It was a massive nothic out of an eldritch nightmare. They quickly sprang into action. Corvus unleashed a flurry of well-placed arrows as Elzalath harnessed her righteous power to bring down destruction on this beast. Upon arriving in melee range, Corvus and Elzalath beat the life out of the creature. Meanwhile, Peke took careful aim and his arrow flew true, straight into the monster’s teeny-tiny butthole, stimulating the anus and causing the beast to projectile shit all over the cavern. An unholy stench temporarily overwhelmed the nearby combatants, but they quickly recovered and dispatched of the creature. After a bit of healing, the party headed into the unexplored areas of the hideout. In the farthest reaches of the caves, they found a room that was clearly Glasstaff’s. The entire hideout was abandoned. Among the baubles and arcana, they found a rat skittering around the room. Glasstaff addressed the team through the rat, demonstrating its presence as a surveillance simulacrum. When Peke attempted to destroy it, it disappeared. But the most disturbing part was their discovery of Glasstaff’s identity. Glasstaff was their pal’s missing friend. Will our heroes ever track down Glasstaff? What the fuck, friend we rescued earlier? Why can’t Ryan ever remember anyone’s name? Tune in next time to find out! Session 5 Our heroes awoke refreshed in Queline’s entryway. Carp bounced into the room, thrilled at the prospect of adventure. Corvus swept the sleep out of his eyes as Pekeporo packed their equipment. With a final caution from Queline toward safety, the trio set out for the secret entrance to the Redbrand cave with K’ot. They had a leisurely and uneventful stroll to the cave mouth. Carp pointed it out and went on his way, pockets a bit heavier with the gold he earned for his bravery. Corvus, Peke and K’ot cautiously entered the cave. There they found a massive room with a stream running through it, but Corvus was unable to enjoy its magnificence, as he felt an oppressive voice in his mind. He couldn’t shake the voice as despair settled in. Fortunately, Peke noticed the tree from his vision in the southeast corner of the room, so they went down the hall. As they distanced themselves from the entry room, Corvus felt his sanity returning. They found another room containing many doors and a well. Peke sensed something was odd about the well, so he went to inspect it closely. He found a thin strand, which he pulled to reveal the necessary supplies for disguise and escape. Corvus led the way as they explored a storage room beyond the well, where they found three Redbrands. Their attempts at deceit fell pathetically short, so Corvus made quick work of the three foes. Before dispatching the third, Corvus questioned the Redbrand, but he was of little assistance. His lack of useful information was rewarded with the business end of Corvus’ blades. Fearing discovery, Peke weighed the bodies down and dumped them in the well. Further exploration revealed an abandoned manor and a hallway with an ornate door bearing Peke’s vision tree. The voice from his vision urged him on. He noticed a trap in the hall and managed to trigger the collapse of the floor without stepping on it. This left him a narrow ledge to cross. Unfortunately, his poor agility led him to fall in the pit anyway. Corvus danced across the ledge and sent down a rope to help him out. Corvus and Peke explored the next room, where they discovered four strange sarcophagi. Feeling a bit of weariness, they decided to press on in their Redbrand disguises. When they entered the next room, they discovered a prison guarded by two Redbrands. This time, they managed to convince the Rebrands they were replacement guards and even persuaded the Redbrands to give them keys to the cells. Peke was so excited it worked, he didn’t think to change out of his disguise before setting out to befriend the silent Drow Elf sitting in one of the cells. When he discovered his foolishness, he quickly explained himself and recklessly freed the Drow and offered her goodberries to gain her trust. While the drow was hesitant, she did offer her name: Elzalath. Meanwhile, Corvus freed the woman and children in the opposite cell. Through discussion, he learned that this was Mirna, the woman who disappeared with her children when her husband was killed. The group set out to exit the caves when Elzalath recognized a dark presence. Frustrated with her liberators’ lack of thoroughness, she knew a fight with evil was imminent. Fortunately, her combat spell proficiencies had suffered no entropy while imprisoned. As skeletons emerged from the coffins, Elzalath and Corvus made quick work of them with some assistance from Peke and K’ot. They cautiously exited the caves. Fortunately Peke’s earlier actions prevented the Redbrands from discovering their deceit. They headed back to Phandalin to return Myrna and her children to their home, to seek weapons for Elzalath from Halia, and to get some much needed rest after a productive day. Will the Redbrands recognize them palling around town? Will they ever figure out who Glasstaff is? Is Peke moving in on Myrna way too soon after she lost her husband and was imprisoned? Tune in next time to find out! Session 4 K’ot was lying bloody and bruised next to the enraged ogre. Pekeporo’s distaste for violence disappeared as the spirit of vengeance overcame him. With a battle cry, he unleashed an arrow that sunk deep into the ogre’s cheek. His vengeful rage emboldened his companions. Corvus let loose an unrelenting barrage of arrows. K’ot mustered the energy to bite the ogre’s ankle, but the ogre immediately dealt a near-fatal blow, knocking her out indefinitely. Meanwhile, the orcs escaped their verdant chains. Filled with bloodlust, they charged down the corridor. Pekeporo knew this day might come. They were too numerous to deal with individually. Despite his truest friend’s presence in the blast radius, he appealed to Aerdrie and watched sorrowfully as lightning erupted from his fingers, tearing apart the bodies of two of the orcs, but also blasting K’ot against a wall to crumple in a seemingly lifeless heap. The ogre managed to stay outside the arc of the lightning blast. He gathered his last bit of strength to heave his great battle axe at Corvus while the remaining orc heaved a nearby javelin at Pekeporo. The orc’s waning strength was a boon for Pekeporo as the javelin clanged to the ground five feet in front of him. But even a near-dead ogre outmatches most creatures in strength. His axe whipped across the cavern and dug deep into Corvus’ shoulder, parting flesh and muscle. Where many would find despair at suffering such injury, Corvus found a renewed sense of vengeance and purpose. Adrenaline mixed with hatred in his veins as he wrenched the cursed axe from his shoulder. The unavenged spirits of his slain family filled him with a strength reserved only for the bulkiest berserkers as he flung the ogre’s great axe directly back down the corridor. Pekeporo watched in awe, mouth agape, as the axe sailed end-over-end, supernaturally picking up speed as it travelled. The well-sharpened blade flew through the ogre’s skull so cleanly and with such ferocity that it took a moment for the split to appear. But when the split came, Pekeporo realized the axe had cleaved the ogre’s head straight down the middle. The axe continued its inspired journey as though guided by the hand of divinity straight into the chest of the remaining surviving orc. Corvus let out a mighty cry of satisfied absolution. Upon recognizing their victory, Pekeporo couldn’t share in Corvus’ glee. He ran to K’ot and laid hands on her. He called desperately on Angharradh, elven goddess of protection to revive K’ot. Then, to his indescribable relief, he felt her chest move as she began breathing once again. A wave of joy washed over Pekeporo as he clung to K’ot. He moved deeper into the passage to examine its contents with K’ot by his side. But when they passed the ogre’s lifeless body, K’ot stopped and whimpered. Pekeporo kept walking but couldn’t shake the feeling that he and his friends had violated innocent life in their violence against the ogre. As Pekeporo explored the cave, grabbing some well-deserved loot, Corvus set about removing the heads of their dispatched foes. Pekeporo couldn’t help noticing how much happier Corvus seemed.green-nature-trees-forest-grass-plants-magic-orbs-fresh-new-hd-wallpaper2 After such a demanding battle, the trio bedded down until morning to get some much-needed rest. Pekeporo took first watch. Halfway through his watch, he sensed a presence in the forest with him. Growing up isolated among the beasts of the Wood of Sharp Teeth, he was no stranger to this sensation. So he followed his habit in such cases and tapped into Rillifane’s insight to communicate with nearby animals. But his appeal went unanswered by the surrounding woods. This presence was no animal. Then he saw hundreds of orbs of light passing through the forest around him. Trying to make sense of the strange phenomenon, he attempted to touch one with no success. But he heard a clamor of quiet voices. Through the din, he was able to decipher the repetitive whisper of “name…name.” Seeking guidance, Pekeporo poured his thoughts and energy into the mistletoe woven into his druidic focus and touched one of the orbs with it. Suddenly, he was blasted back against the nearest tree. Yet Pekeporo felt nothing as his mind and vision were completely engulfed in a prophetic dream of the largest and most ebullient tree he had ever seen. It seemed beyond anything beyond possibility as he clearly heard, “Portal.” The trio returned to Phandalin in the morning as it burst with life. Children played while neighbors talked idly. Corvus made his way to the Townmaster’s Hall with his friends, until they were abruptly stopped by four Redbrands standing in the commons. Corvus and Pekeporo were ready to get their reward and spend an evening in drunken revelry, but the Rebrands made it clear that they were a problem that could no longer be ignored. K’ot frightened one of the Redbrands off, but the rest bared naked steel in response to threats. Their de facto leader closed the gap in attack. Corvus, still feeling confident from the destruction of orcs, would no longer abide gangs of foes teaming up against them. He found an unprecedented swiftness and strength to engage two foes at a time. He simultaneously embraced this new combat style, noting its efficacy for future engagement. While one of the ruffians buried his sword deep in Pekeporo’s side, Corvus almost single-handedly dispatched of his opponents in a flurry of shortswords. K’ot felt a growing connection to Corvus and sunk her teeth into the final opponent’s manhood. She lapped up some of the blood and looked to Corvus for approval as the man bled out from the groin. Corvus was quite pleased and affirmed K’ot’s tactics. After supernaturally knitting his side back together, Pekeporo rejoined Corvus, and they donned the Red cloaks of their slain enemies as a disguise. They looked around and realized their fight had emptied the town, confirming the grip of fear the Redbrands had instilled in the town. As they approached the Townmaster’s hall, Townmaster Harbin burst through the doors, frantically crying, “What have you done? Oh god, you’ve only made things worse. Glasstaff will repay this tenfold.” Even in the midst of his cowardly tirade, when Harbin noticed the orc heads dangling from Corvus’ belt, his mood slightly lightened. After brushing with death thanks to Harbin’s incompetence, Pekeporo had enough. “Look, we took care of your orc problem. Now we’re taking care of your Redbrand problem . Whether or not you’re competent enough to recognize it, you need us. Now pay what you owe.” “No, your help can’t make up for how much worse you’ve made things.” Harbin turned on his pudgy heel, waddled into the hall and slammed the door. Pekeporo ran down to Halia’s shop to drop off the swords of the slain Redbrands. In return, she offered him information. “Glasstaff is headquartered in the manor east of town with the rest of the local Redbrands.” She thanked him for their help. When Pekeporo returned and told Corvus about Halia’s message, Corvus communed with the largest tree in the commons to learn more about Redbrand activity. Then they went back to the tavern for a night of drinking and relaxation. The innkeeper was very grateful and reminded them of Queline Alderleaf’s son, Carp, discovering a secret passage to the Redbrands hideout. They thanked him for his hospitality and retired for the night… Until Corvus woke up Pekeporo at midnight. “Peke, let’s go scope out that tunnel.” “We don’t know where it is.” “Oh, right. But Carp does.” “You’re right. Let’s go talk to him.” “Now? It’s the middle of the night.” “Exactly. No one will suspect we got help from a little kid in the middle of the night.” They snuck out the window and stealthily walked down to Alderleaf’s farm and awakened Queline with their knocking. She was initially displeased with being awakened, but quickly warmed to the adventurers who were willing to rid her town of its blight. She invited them in but was unwilling to allow her son to show them the cave in the middle of the night. So they posted up in her entryway until morning, waiting for the opportunity to take on the local Redbrand scourge head-on… Will our heroes be prepared to face such a massive threat? Will their allies manage to avoid the wrath of the Redbrands? Will Corvus turn into Dae Hyo from the Hollow Gods trilogy? Does anyone even know what I’m talking about? Tune in next time to find out! Session 3 “Show yourself.” Corvus was in no mood to deal with a new threat. He was exhausted, and Sildar was in no physical condition to engage in a fight. Diplomacy was the only option. “Very well,” called the sinister voice as there was a rustling high in the nearest treetop. A nimble figure clothed in all green was barely visible among the foliage as it leapt from the heights. The masked figure taunted, “You are truly fucked, Corvus. Porcius placed a 100 gold bounty on your head. Every bounty hunter worth their salt is out to get you. Fortunately, for me, you’re shit at hiding.” He noticed Pekeporo moving his hand over his staff to cast Shilelagh. “I wouldn’t do that if I were you, Druid. I may be outnumbered, but you’ll not survive a fight with me.” Pekeporo withdrew his hand as Corvus dropped his blades. If Corvus wasn’t ready to fight, Pekeporo knew this bounty hunter’s words were no idle threat. True to form, though, Corvus wouldn’t go down without a verbal fight. “Who is this Corvus you keep talking about? I’ve never heard of him.” “Ha. You must take me for a fool from your backwoods village. I know very well who you are, and you’re coming with me. Unless you can make it worth my while.” Pekeporo was dismayed. “We can’t just pay off every bounty hunter that comes calling.” Corvus looked around, noting the windfall of valuable goods they were delivering to Phandalin. “No, Peke, we can’t. But we can pay this one.” He turned to the bounty hunter. “I can pay, but there’s no way I’m worth 100 gold to Porcius. He’s offering you 50 at the most, so I’ll give you 75 gold pieces to fuck off.” “Very well.” The bounty hunter held out his hand for compensation. The moment the gold dropped in his hand, he started skipping off, laughing. “You fools,” he taunted, “the offered reward is only 25 gold.” And with that, he disappeared into the woods. Corvus felt a bit foolish but was too relieved to be on his way to care too much. Pekeporo’s wolf sensed the alpha presence of K’ot and sat at her feet. This act of supplication greatly pleased K’ot as she loved Pekeporo but often missed her pack. Pekeporo came over and told the wolf, “I promised you freedom. You repaid me with your protection. There is no debt between us. Be free.” The wolf yipped in gratitude and bounded into the forest. The crew uneventfully journeyed with the goods the rest of the way to Phandalin. On the journey, Corvus advised his friends to call him Markan to avoid such unpleasant encounters in the future. Upon entry in Phandalin, they made their way directly to Halia Thorton’s blacksmith shop, recognizing the symbol on her sign from some of the crates they transported. They delivered her goods to much gratitude and offers of even the most exclusive wares. Corvus and Pekeporo found an instant ally in this strange new town. She warned them of the presence of Redbrands, a group of ruffians casting the town into a state of fear and paranoia. They thanked her and moved on to deliver the rest of the goods. At the inn that night, they drank, laughed and talked. They learned of many local problems and discoveries from the other patrons in the bar. Sildar told them of his true intentions. He made his way to Phandalin to seek answers regarding the disappearance of his friend, a sorcerer, who was last seen on the outskirts of town. They lost the night to friendship. The next morning, Corvus and Pekeporo headed to the Townmaster’s Hall to seek more information on the commission of adventurers to rid the town of a nearby band of orcs descending upon unwary travelers. For some reason Pekeporo didn’t understand, Corvus was hellbent on seeking out these orcs. But he was in this for the adventure, and Corvus was the truest friend he had found since leaving the Wood of Sharp Teeth. When questioned about the Redbrands, the Townmaster grew nervous and undermined the word of the townsfolk as mere hyperbole. Before leaving, Corvus and Pekeporo promised Sildar they would return to assist him in his quest. Corvus, Pekeporo and K’ot made their way on the trail but ducked into the woods as they approached the orc encampment. There were two watchorcs at the opening of the enclosure. Thanks to their stealthy tactics, they quickly dispatched of these guards in a flurry of arrows and direwolf teeth. After hiding the guards’ bodies, the pair slowly approached the entrance. Inside, they found three orcs and a massive ogre, all sleeping. Both adventurers felt a sense of being outmatched and quickly took advantage of their surprise position. They unleashed their arrows at the sleeping ogre as K’ot bounded down the opening at him. The ogre awoke with a cry of pain thereby arousing the orcs. They charged toward the opening, but Pekeporo called on the power of Rillifane Rallathil, elven god of nature. Barbed vines sprouted from the floor and walls and bound the orcs where they stood. Meanwhile, the ogre slowed in the face of Corvus’ relentless arrow strikes and K’ot’s persistent attacks. As the battle seemed all but won, the ogre tossed K’ot against the wall and brought his massive club down in a crushing blow on her skull. Pekeporo cried out in fury as he witnessed his closest companion crumple to the ground. Will our heroes overcome this, the greatest battle they have ever faced? Will Corvus find what he needs to be whole, and will it cost him his friendship with Pekeporo? Will our heroes ever stop ruining perfectly pleasant nights of sleep for their foes? Tune in next time to find out!

A methodology for the collection of supplemental information on agricultural fatalities. Due to the lack of a standardized methodology to identify agriculturally related fatalities and the inaccuracy of data from death certificates, NSC's and NIOSH's estimates of the level of agriculturally related fatalities conflict, and the validity of both is subject to criticism. A follow-up survey with the next of kin of 107 potential agricultural fatalities, from a pool of 150 Pennsylvania cases, has been conducted. This paper details the methods and success in gaining supplemental information. Next of kin were generally willing to provide the information (67%), with both telephone (66%) and mail (68%) methods generating comparable levels of cooperation. The optimal time frame for finding current addresses and telephone numbers as well as for gaining cooperation appeared to be about 1 year after the fatality. However, a retrospective investigation of up to 3 years past the anniversary of the fatality produced an acceptable rate of cooperation. How the collected data compares with fatality data for agriculture published by NSC and N1OSH has been addressed in another paper.

Q: How to modify a boxplot!? (mean, outliers, values: median, ...) I have some issues with my boxplot and I hope someone can help me to fix them. Here my MWE: \documentclass[a4paper]{scrbook} \usepackage[utf8]{inputenc} \usepackage{textcomp} \usepackage[T1]{fontenc} \usepackage[english,ngerman]{babel} \usepackage[babel,german=quotes,threshold=1]{csquotes} \usepackage[language=autobib, backend=biber %%% !!!! ]{biblatex} \usepackage{lmodern} \usepackage{scrlayer-scrpage} \usepackage{tikz} \usepackage{pgfplots} \usepackage{pgfplotstable} \usepgfplotslibrary{statistics} \pgfplotsset{compat=1.8} \pgfplotsset{/pgf/number format/use comma} %%% does not work: comma instead of dot \usepackage{caption} \usepackage{graphicx} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{document} \begin{figure} \textcolor{black!50}{\rule{\linewidth}{0.25pt}} \centering \scriptsize \\[5ex] \begin{tikzpicture} \begin{axis} [ clip=false, boxplot/draw direction=y, boxplot/variable width, boxplot/every median/.style={black,very thick,solid}, width=1.00\textwidth, height=0.33\textheight, ylabel style={align=center}, ytick={0,1,2,3,4,5,6,7}, y tick label style={align=right}, yticklabels={0,1,2,3,4,5,6,7}, xtick={0,1,2,3,4}, x tick label style={align=center}, xticklabels={,1,2,3,4} ] \node at (axis cs:0,0) [anchor=west] {\tiny (- - -)}; \node at (axis cs:0,8) [anchor=west] {\tiny (+++)}; %%% MEAN %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \draw[black, densely dotted] (axis cs:0, 5.50) coordinate (tmp) (current axis.west |- tmp) -- (current axis.east |- tmp); %%% Can the value "5,50" be displayed on the left side of the graph above he dotted line? %%% 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, boxplot prepared={ average=5, lower whisker=2, lower quartile=4.5, median=5, upper quartile=6.5, upper whisker=7, sample size=3 }, ] coordinates {} node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\boxplotvalue{average}}; ; % 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, mark options={fill=white}, boxplot prepared={ average=6.25, lower whisker=5, lower quartile=6, median=6, upper quartile=7, upper whisker=7, sample size=3 }, ] coordinates { (2,1) % mark as "*" (2,4) % mark as "°" } node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\boxplotvalue{average}}; ; % 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, mark options={fill=white}, boxplot prepared={ average=6.3, lower whisker=5, lower quartile=6, median=6, upper quartile=7, upper whisker=7, sample size=3 }, ] coordinates {(3,4)} node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\boxplotvalue{average}}; ; % 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, boxplot prepared={ average=5.50, lower whisker=4, lower quartile=5, median=6, upper quartile=6, upper whisker=7, sample size=3 }, ] coordinates {} node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\boxplotvalue{average}}; \end{axis} \end{tikzpicture} \textcolor{black!50}{\rule{\linewidth}{0.25pt}} \caption[caption]{caption} \end{figure} \end{document} How can I replace the "." with a "," as decimal separator? I thought \pgfplotsset{/pgf/number format/use comma} would be a solution!? Extreme outliers are usually marked with a "*". How can I do so for the outlier with the coordinates (2,1)? The mean over all the groups is 5.5. (dotted line) Is it possible to display this value above the dotted line on the left side of the graph? Can the 50% quantile of the boxes be displayed? (left of each box) A: I had expected a couple of these things to be a bit harder, but wasn't that bad. As mentioned in a comment, \boxplotvalue{average} doesn't pass the number through the pgf number parser, so the /pgf/number format/use comma has no effect. Use \pgfmathprintnumber{\boxplotvalue{average}}. One way of achieving this is to use scatter/classes. To the axis options add scatter/classes={ a={mark=star}, b={mark=*} } To the options of the \addplot in question add scatter, scatter src=explicit symbolic, And finally modify the coordinate stream to have coordinates { (2,1) [a] (2,4) [b] }; To add a node above the start of a line from a to b one can use e.g. \draw (a) node[above right] {...} -- (b);. So in your case \draw[black, densely dotted] (axis cs:0, 5.50) coordinate (tmp) (current axis.west |- tmp) node[above right] {\pgfmathprintnumber{5.5}}-- (current axis.east |- tmp); (Some modification possibly needed, depending on exactly where you wanted that node, not sure I understood you correctly.) The 50% quantile is just the median right, so you could use exactly the same method as you used for the average value. Do you want next to the middle of the box, or next to the median (as seen below) by the way? That said, it's possible to modify how boxplot is drawn in order to add those two nodes by default. The styles that need modification is described in the manual, so I took most of the following from there, adding just two things indicated by comments: \makeatletter \pgfplotsset{ boxplot/draw/average/.code={ % the bulk part of this is the default, found in the manual \draw [/pgfplots/boxplot/every average/.try] \pgfextra % do NOT use \draw[mark=*] plot coordinates because % boxplots uses the same plot handler to draw its % outliers. \pgftransformshift{% % basic level access to ’boxplot box cs’: \pgfplotsboxplotpointabbox {\pgfplotsboxplotvalue{average}} {0.5}% }% \pgfuseplotmark{\tikz@plot@mark}% \endpgfextra ; % the following line is the only new addition \node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\pgfmathprintnumber{\boxplotvalue{average}}}; }, boxplot/draw/median/.code={ \draw [/pgfplots/boxplot/every median/.try] (boxplot box cs:\pgfplotsboxplotvalue{median},0) % the following line is the only difference from the default definition node[left] {\pgfmathprintnumber{\pgfplotsboxplotvalue{median}}} -- (boxplot box cs:\pgfplotsboxplotvalue{median},1); }, } \makeatother In the code below I just added this just before \begin{tikzpicture}. However, if you have multiple boxplots that all should have the same style, it's probably better to either move it to the preamble, or make a new style. \documentclass[a4paper]{scrbook} \usepackage[utf8]{inputenc} \usepackage[T1]{fontenc} \usepackage{pgfplotstable} % loads pgfplots which loads tikz which loads graphicx \usepgfplotslibrary{statistics} \pgfplotsset{compat=1.8} \pgfplotsset{/pgf/number format/use comma} \usepackage{caption} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{document} \begin{figure} \textcolor{black!50}{\rule{\linewidth}{0.25pt}} \centering \scriptsize \\[5ex] \makeatletter \pgfplotsset{ boxplot/draw/average/.code={ % the bulk part of this is the default, found in the manual \draw [/pgfplots/boxplot/every average/.try] \pgfextra % do NOT use \draw[mark=*] plot coordinates because % boxplots uses the same plot handler to draw its % outliers. \pgftransformshift{% % basic level access to ’boxplot box cs’: \pgfplotsboxplotpointabbox {\pgfplotsboxplotvalue{average}} {0.5}% }% \pgfuseplotmark{\tikz@plot@mark}% \endpgfextra ; % the following line is the only new addition \node[right,font=\scriptsize] at (boxplot box cs: \boxplotvalue{average}, 0.95) {\pgfmathprintnumber{\boxplotvalue{average}}}; }, boxplot/draw/median/.code={ \draw [/pgfplots/boxplot/every median/.try] (boxplot box cs:\pgfplotsboxplotvalue{median},0) % the following line is the only difference from the default definition node[left] {\pgfmathprintnumber{\pgfplotsboxplotvalue{median}}} -- (boxplot box cs:\pgfplotsboxplotvalue{median},1); }, } \makeatother \begin{tikzpicture} \begin{axis} [ clip=false, boxplot/draw direction=y, boxplot/variable width, boxplot/every median/.style={black,very thick,solid}, width=1.00\textwidth, height=0.33\textheight, ylabel style={align=center}, ytick={0,1,2,3,4,5,6,7}, y tick label style={align=right}, yticklabels={0,1,2,3,4,5,6,7}, xtick={0,1,2,3,4}, x tick label style={align=center}, xticklabels={,1,2,3,4}, scatter/classes={ a={mark=star}, b={mark=*} } ] \node at (axis cs:0,0) [anchor=west] {\tiny (- - -)}; \node at (axis cs:0,8) [anchor=west] {\tiny (+++)}; %%% MEAN %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \draw[black, densely dotted] (axis cs:0, 5.50) coordinate (tmp) (current axis.west |- tmp) node[above right] {\pgfmathprintnumber{5.5}}-- (current axis.east |- tmp); %%% Can the value "5,50" be displayed on the left side of the graph above he dotted line? %%% 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, boxplot prepared={ average=5, lower whisker=2, lower quartile=4.5, median=5, upper quartile=6.5, upper whisker=7, sample size=3 }, ] coordinates {}; % 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ boxplot, scatter, scatter src=explicit symbolic, mark options={fill=white}, boxplot prepared={ average=6.25, lower whisker=5, lower quartile=6, median=6, upper quartile=7, upper whisker=7, sample size=3 }, ] coordinates { (2,1) [a] (2,4) [b] }; % 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, mark options={fill=white}, boxplot prepared={ average=6.3, lower whisker=5, lower quartile=6, median=6, upper quartile=7, upper whisker=7, sample size=3 }, ] coordinates {(3,4)}; % 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \addplot[ mark=*, boxplot, boxplot prepared={ average=5.50, lower whisker=4, lower quartile=5, median=6, upper quartile=6, upper whisker=7, sample size=3 }, ] coordinates {}; \end{axis} \end{tikzpicture} \textcolor{black!50}{\rule{\linewidth}{0.25pt}} \caption[caption]{caption} \end{figure} \end{document}

package jsat.classifiers.trees; import java.util.*; import jsat.classifiers.CategoricalData; import jsat.classifiers.CategoricalResults; import jsat.classifiers.ClassificationDataSet; import jsat.classifiers.Classifier; import jsat.classifiers.DataPoint; import jsat.classifiers.trees.ImpurityScore.ImpurityMeasure; import jsat.math.OnLineStatistics; import jsat.parameters.Parameterized; import jsat.regression.RegressionDataSet; import jsat.regression.Regressor; import jsat.utils.IntList; import jsat.utils.IntSet; import jsat.utils.ListUtils; import jsat.utils.random.RandomUtil; /** * The ExtraTree is an Extremely Randomized Tree. Splits are chosen at random, * and the features that are selected are also chosen at random for each new * node in the tree. <br> * If set to randomly select one feature for each node, it becomes a <i>Totally * Randomized Tree</i><br> * See: <br> * Geurts, P., Ernst, D.,&amp;Wehenkel, L. (2006). <i>Extremely randomized trees * </i>. Machine learning, 63(1), 3–42. doi:10.1007/s10994-006-6226-1 * * @author Edward Raff */ public class ExtraTree implements Classifier, Regressor, TreeLearner, Parameterized { //TODO in both of the train methods, 2 passes are done for numeric features. This can be done in one pass by fiding the min/max when we split, and passing that info in the argument parameters private static final long serialVersionUID = 7433728970041876327L; private int stopSize; private int selectionCount; protected CategoricalData predicting; private boolean binaryCategoricalSplitting = true; /** * Just stores the number of numeric features that were in the dataset for * that the getFeatures method can be implemented correctly for categorical * variables. */ private int numNumericFeatures; private ImpurityScore.ImpurityMeasure impMeasure = ImpurityMeasure.NMI; private TreeNodeVisitor root; /** * Creates a new Extra Tree that will use all features in the training set */ public ExtraTree() { this(Integer.MAX_VALUE, 5); } /** * Creates a new Extra Tree * * @param selectionCount the number of features to select * @param stopSize the stop size */ public ExtraTree(int selectionCount, int stopSize) { this.stopSize = stopSize; this.selectionCount = selectionCount; this.impMeasure = ImpurityMeasure.NMI; } /** * Copy constructor. * @param toCopy the object to copy */ public ExtraTree(ExtraTree toCopy) { this.stopSize = toCopy.stopSize; this.selectionCount = toCopy.selectionCount; if(toCopy.predicting != null) this.predicting = toCopy.predicting; this.numNumericFeatures = toCopy.numNumericFeatures; this.binaryCategoricalSplitting = toCopy.binaryCategoricalSplitting; this.impMeasure = toCopy.impMeasure; if(toCopy.root != null) this.root = toCopy.root.clone(); } /** * Sets the impurity measure used during classification tree construction to * select the best of the features. * @param impurityMeasure the impurity measure to use */ public void setImpurityMeasure(ImpurityMeasure impurityMeasure) { this.impMeasure = impurityMeasure; } /** * Returns the impurity measure in use * @return the impurity measure in use */ public ImpurityMeasure getImpurityMeasure() { return impMeasure; } /** * Sets the stopping size for tree growth. When a node has less than or * equal to <tt>stopSize</tt> data points to train from, it terminates and * produces a leaf node. * @param stopSize the size of the testing set to refuse to split */ public void setStopSize(int stopSize) { if(stopSize <= 0) throw new ArithmeticException("The stopping size must be a positive value"); this.stopSize = stopSize; } /** * Returns the stopping size for tree growth * @return the stopping size for tree growth */ public int getStopSize() { return stopSize; } /** * The ExtraTree will select the best of a random subset of features at each * level, this sets the number of random features to select. If set larger * than the number of features in the training set, all features will be * eligible for selection at every level. * @param selectionCount the number of random features to select */ public void setSelectionCount(int selectionCount) { this.selectionCount = selectionCount; } /** * Returns the number of random features chosen at each level in the tree * @return the number of random features to chose */ public int getSelectionCount() { return selectionCount; } /** * The normal implementation of ExtraTree always produces binary splits, * including for categorical features. If set to <tt>false</tt> categorical * features will expand out for each value in the category. This reduces the * randomness of the tree. * @param binaryCategoricalSplitting whether or not to use the original * splitting algorithm, or to fully expand nominal features */ public void setBinaryCategoricalSplitting(boolean binaryCategoricalSplitting) { this.binaryCategoricalSplitting = binaryCategoricalSplitting; } /** * Returns whether or not binary splitting is used for nominal features * @return whether or not binary splitting is used for nominal features */ public boolean isBinaryCategoricalSplitting() { return binaryCategoricalSplitting; } @Override public CategoricalResults classify(DataPoint data) { return root.classify(data); } @Override public void train(ClassificationDataSet dataSet, boolean parallel) { Random rand = RandomUtil.getRandom(); IntList features = new IntList(dataSet.getNumFeatures()); ListUtils.addRange(features, 0, dataSet.getNumFeatures(), 1); predicting = dataSet.getPredicting(); ImpurityScore score = new ImpurityScore(predicting.getNumOfCategories(), impMeasure); for(int i = 0; i < dataSet.size(); i++) score.addPoint(dataSet.getWeight(i), dataSet.getDataPointCategory(i)); numNumericFeatures = dataSet.getNumNumericalVars(); root = trainC(score, dataSet, features, dataSet.getCategories(), rand); } /** * Creates a new tree top down * @param setScore the impurity score for the set of data points being evaluated * @param subSet the set of data points to perform a split on * @param features the features available to split on * @param catInfo the categorical information * @param rand the source of randomness * @param reusableLists a stack of already allocated lists that can be added and removed from * @return the new top node created for the given data */ private TreeNodeVisitor trainC(ImpurityScore setScore, ClassificationDataSet subSet, List<Integer> features, CategoricalData[] catInfo, Random rand) { //Should we stop? Stop split(S) if(subSet.size() < stopSize || setScore.getScore() == 0.0) { if(subSet.isEmpty()) return null; return new NodeC(setScore.getResults()); } double bestGain = Double.NEGATIVE_INFINITY; double bestThreshold = Double.NaN; int bestAttribute = -1; ImpurityScore[] bestScores = null; List<ClassificationDataSet> bestSplit = null; Set<Integer> bestLeftSide = null; /* * TODO use smarter random feature selection based on how many features * we need relative to how many are available */ Collections.shuffle(features); //It is possible, if we test all attribute - that one was categorical and no longer an option final int goTo = Math.min(selectionCount, features.size()); for(int i = 0; i < goTo; i++) { double gain; double threshold = Double.NaN; Set<Integer> leftSide = null; ImpurityScore[] scores; int a = features.get(i); List<ClassificationDataSet> aSplit; if(a < catInfo.length) { final int vals = catInfo[a].getNumOfCategories(); if(binaryCategoricalSplitting || vals == 2) { scores = createScores(2); Set<Integer> catsValsInUse = new IntSet(vals*2); for(int j = 0; j < subSet.size(); j++) catsValsInUse.add(subSet.getDataPoint(j).getCategoricalValue(a)); if(catsValsInUse.size() == 1) return new NodeC(setScore.getResults()); leftSide = new IntSet(vals); int toUse = rand.nextInt(catsValsInUse.size()-1)+1; ListUtils.randomSample(catsValsInUse, leftSide, toUse, rand); //Now we have anything in leftSide is path 0, we can do the bining aSplit = new ArrayList<>(2); aSplit.add(subSet.emptyClone()); aSplit.add(subSet.emptyClone()); for(int j = 0; j < subSet.size(); j++) { int dest = leftSide.contains(subSet.getDataPoint(j).getCategoricalValue(a)) ? 0 : 1; scores[dest].addPoint(subSet.getWeight(j), subSet.getDataPointCategory(j)); aSplit.get(dest).addDataPoint(subSet.getDataPoint(j), subSet.getDataPointCategory(j), subSet.getWeight(j)); } } else//split on each value { scores = createScores(vals); //Bin all the points to get their scores aSplit = new ArrayList<>(vals); for(int z = 0; z < vals; z++) aSplit.add(subSet.emptyClone()); for (int j = 0; j < subSet.size(); j++) { DataPoint dp = subSet.getDataPoint(j); int y_j = subSet.getDataPointCategory(j); double w_j = subSet.getWeight(j); scores[dp.getCategoricalValue(a)].addPoint(w_j, y_j); aSplit.get(dp.getCategoricalValue(a)).addDataPoint(dp, y_j, w_j); } } } else { int numerA = a - catInfo.length; double min = Double.POSITIVE_INFINITY, max = Double.NEGATIVE_INFINITY; for (int j = 0; j < subSet.size(); j++) { double val = subSet.getDataPoint(j).getNumericalValues().get(numerA); min = Math.min(min, val); max = Math.max(max, val); } //Uniform random threshold threshold = rand.nextDouble()*(max-min)+min; scores = createScores(2); aSplit = new ArrayList<>(2); aSplit.add(subSet.emptyClone()); aSplit.add(subSet.emptyClone()); for (int j = 0; j < subSet.size(); j++) { double val = subSet.getDataPoint(j).getNumericalValues().get(numerA); double w_j = subSet.getWeight(j); int y_j = subSet.getDataPointCategory(j); int toAddTo = val <= threshold ? 0 : 1; aSplit.get(toAddTo).addDataPoint(subSet.getDataPoint(j), y_j, w_j); scores[toAddTo].addPoint(w_j, y_j); } } gain = ImpurityScore.gain(setScore, scores); if(gain > bestGain) { bestGain = gain; bestAttribute = a; bestThreshold = threshold; bestScores = scores; bestSplit = aSplit; bestLeftSide = leftSide; } } //Best attribute has been selected NodeBase toReturn; if(bestAttribute < 0) return null; if(bestAttribute < catInfo.length) if(bestSplit.size() == 2)//2 paths only toReturn = new NodeCCat(bestAttribute, bestLeftSide, setScore.getResults()); else { toReturn = new NodeCCat(goTo, bestSplit.size(), setScore.getResults()); features.remove(new Integer(bestAttribute));//Feature nolonger viable in this case } else toReturn = new NodeCNum(bestAttribute-catInfo.length, bestThreshold, setScore.getResults()); for(int i = 0; i < toReturn.children.length; i++) { toReturn.children[i] = trainC(bestScores[i], bestSplit.get(i), features, catInfo, rand); } return toReturn; } /** * Creates a new tree top down * @param setScore the impurity score for the set of data points being evaluated * @param subSet the set of data points to perform a split on * @param features the features available to split on * @param catInfo the categorical information * @param rand the source of randomness * @param reusableLists a stack of already allocated lists that can be added and removed from * @return the new top node created for the given data */ private TreeNodeVisitor train(OnLineStatistics setScore, RegressionDataSet subSet, List<Integer> features, CategoricalData[] catInfo, Random rand) { //Should we stop? Stop split(S) if(subSet.size() < stopSize || setScore.getVarance() <= 0.0 || Double.isNaN(setScore.getVarance())) return new NodeR(setScore.getMean()); double bestGain = Double.NEGATIVE_INFINITY; double bestThreshold = Double.NaN; int bestAttribute = -1; OnLineStatistics[] bestScores = null; List<RegressionDataSet> bestSplit = null; Set<Integer> bestLeftSide = null; /* * TODO use smarter random feature selection based on how many features * we need relative to how many are available */ Collections.shuffle(features); //It is possible, if we test all attribute - that one was categorical and no longer an option final int goTo = Math.min(selectionCount, features.size()); for(int i = 0; i < goTo; i++) { double gain; double threshold = Double.NaN; Set<Integer> leftSide = null; OnLineStatistics[] stats; int a = features.get(i); List<RegressionDataSet> aSplit; if(a < catInfo.length) { final int vals = catInfo[a].getNumOfCategories(); if(binaryCategoricalSplitting || vals == 2) { stats = createStats(2); Set<Integer> catsValsInUse = new IntSet(vals*2); for(int j = 0; j < subSet.size(); j++) catsValsInUse.add(subSet.getDataPoint(j).getCategoricalValue(a)); if(catsValsInUse.size() == 1) return new NodeR(setScore.getMean()); leftSide = new IntSet(vals); int toUse = rand.nextInt(catsValsInUse.size()-1)+1; ListUtils.randomSample(catsValsInUse, leftSide, toUse, rand); //Now we have anything in leftSide is path 0, we can do the bining aSplit = new ArrayList<>(2); aSplit.add(subSet.emptyClone()); aSplit.add(subSet.emptyClone()); for(int j = 0; j < subSet.size(); j++) { DataPoint dp = subSet.getDataPoint(j); double w_j = subSet.getWeight(j); double y_j = subSet.getTargetValue(j); int dest = leftSide.contains(dp.getCategoricalValue(a)) ? 0 : 1; stats[dest].add(y_j, w_j); aSplit.get(dest).addDataPoint(dp, y_j, w_j); } } else//split on each value { stats = createStats(vals); //Bin all the points to get their scores aSplit = new ArrayList<>(vals); for(int z = 0; z < vals; z++) aSplit.add(subSet.emptyClone()); for (int j = 0; j < subSet.size(); j++) { DataPoint dp = subSet.getDataPoint(j); double w_j = subSet.getWeight(j); double y_j = subSet.getTargetValue(j); stats[dp.getCategoricalValue(a)].add(y_j, w_j); aSplit.get(dp.getCategoricalValue(a)).addDataPoint(dp, y_j, w_j); } } } else { int numerA = a - catInfo.length; double min = Double.POSITIVE_INFINITY, max = Double.NEGATIVE_INFINITY; for(int j = 0; j < subSet.size(); j++) { DataPoint dp = subSet.getDataPoint(j); double val = dp.getNumericalValues().get(numerA); min = Math.min(min, val); max = Math.max(max, val); } //Uniform random threshold threshold = rand.nextDouble()*(max-min)+min; stats = createStats(2); aSplit = new ArrayList<>(2); aSplit.add(subSet.emptyClone()); aSplit.add(subSet.emptyClone()); for (int j = 0; j < subSet.size(); j++) { DataPoint dp = subSet.getDataPoint(j); double w_j = subSet.getWeight(j); double y_j = subSet.getTargetValue(j); double val = dp.getNumericalValues().get(numerA); int toAddTo = val <= threshold ? 0 : 1; aSplit.get(toAddTo).addDataPoint(dp, y_j, w_j); stats[toAddTo].add(y_j, w_j); } } gain = 1; double varNorm = setScore.getVarance(); double varSum = setScore.getSumOfWeights(); for(OnLineStatistics stat : stats) gain -= stat.getSumOfWeights()/varSum*(stat.getVarance()/varNorm); if(gain > bestGain) { bestGain = gain; bestAttribute = a; bestThreshold = threshold; bestScores = stats; bestSplit = aSplit; bestLeftSide = leftSide; } } //Best attribute has been selected NodeBase toReturn; if (bestAttribute >= 0) { if (bestAttribute < catInfo.length) if (bestSplit.size() == 2)//2 paths only toReturn = new NodeRCat(bestAttribute, bestLeftSide, setScore.getMean()); else { toReturn = new NodeRCat(goTo, bestSplit.size(), setScore.getMean()); features.remove(new Integer(bestAttribute));//Feature nolonger viable in this case } else toReturn = new NodeRNum(bestAttribute - catInfo.length, bestThreshold, setScore.getMean()); for (int i = 0; i < toReturn.children.length; i++) { toReturn.children[i] = train(bestScores[i], bestSplit.get(i), features, catInfo, rand); } return toReturn; } return new NodeR(setScore.getMean()); } @Override public boolean supportsWeightedData() { return true; } @Override public ExtraTree clone() { return new ExtraTree(this); } @Override public TreeNodeVisitor getTreeNodeVisitor() { return root; } /** * Add lists to a list of lists * @param <T> the content type of the list * @param listsToAdd the number of lists to add * @param reusableLists available pre allocated lists for reuse * @param aSplit the list of lists to add to */ static private <T> void fillList(final int listsToAdd, Stack<List<T>> reusableLists, List<List<T>> aSplit) { for(int j = 0; j < listsToAdd; j++) if(reusableLists.isEmpty()) aSplit.add(new ArrayList<>()); else aSplit.add(reusableLists.pop()); } private ImpurityScore[] createScores(int count) { ImpurityScore[] scores = new ImpurityScore[count]; for(int j = 0; j < scores.length; j++) scores[j] = new ImpurityScore(predicting.getNumOfCategories(), impMeasure); return scores; } @Override public double regress(DataPoint data) { return root.regress(data); } @Override public void train(RegressionDataSet dataSet, boolean parallel) { train(dataSet); } @Override public void train(RegressionDataSet dataSet) { Random rand = RandomUtil.getRandom(); IntList features = new IntList(dataSet.getNumFeatures()); ListUtils.addRange(features, 0, dataSet.getNumFeatures(), 1); OnLineStatistics score = new OnLineStatistics(); for (int j = 0; j < dataSet.size(); j++) { double w_j = dataSet.getWeight(j); double y_j = dataSet.getTargetValue(j); score.add(y_j, w_j); } numNumericFeatures = dataSet.getNumNumericalVars(); root = train(score, dataSet, features, dataSet.getCategories(), rand); } private OnLineStatistics[] createStats(int count) { OnLineStatistics[] stats = new OnLineStatistics[count]; for(int i = 0; i < stats.length; i++) stats[i] = new OnLineStatistics(); return stats; } /** * Node for classification that splits on a categorical feature */ private class NodeCCat extends NodeC { private static final long serialVersionUID = 7413428280703235600L; /** * Categorical attribute to split on */ private int catAtt; /** * The cat values that go to the left branch, or null if no binary cats are being used */ private int[] leftBranch; public NodeCCat(int catAtt, int children, CategoricalResults crResult) { super(crResult, children); this.catAtt = catAtt; this.leftBranch = null; } public NodeCCat(int catAtt, Set<Integer> left, CategoricalResults crResult) { super(crResult, 2); this.catAtt = catAtt; this.leftBranch = new int[left.size()]; int pos = 0; for(int i : left) leftBranch[pos++] = i; Arrays.sort(leftBranch); } public NodeCCat(NodeCCat toClone) { super(toClone); this.catAtt = toClone.catAtt; if(toClone.leftBranch != null) this.leftBranch = Arrays.copyOf(toClone.leftBranch, toClone.leftBranch.length); } @Override public int getPath(DataPoint dp) { int[] catVals = dp.getCategoricalValues(); if (leftBranch == null) return catVals[catAtt]; else { if (Arrays.binarySearch(leftBranch, catVals[catAtt]) < 0) return 1; else return 0; } } @Override public TreeNodeVisitor clone() { return new NodeCCat(this); } @Override public Collection<Integer> featuresUsed() { IntList used = new IntList(1); used.add(catAtt+numNumericFeatures); return used; } } /** * Node for classification that splits on a numeric feature */ private static class NodeCNum extends NodeC { private static final long serialVersionUID = 3967180517059509869L; private int numerAtt; private double threshold; public NodeCNum(int numerAtt, double threshold, CategoricalResults crResult) { super(crResult, 2); this.numerAtt = numerAtt; this.threshold = threshold; } public NodeCNum(NodeCNum toClone) { super(toClone); this.numerAtt = toClone.numerAtt; this.threshold = toClone.threshold; } @Override public int getPath(DataPoint dp) { double val = dp.getNumericalValues().get(numerAtt); if( val <= threshold) return 0; else return 1; } @Override public TreeNodeVisitor clone() { return new NodeCNum(this); } @Override public Collection<Integer> featuresUsed() { IntList used = new IntList(1); used.add(numerAtt); return used; } } /** * Base node for classification */ private static class NodeC extends NodeBase { private static final long serialVersionUID = -3977497656918695759L; private CategoricalResults crResult; /** * Creates a new leaf node * @param crResult the results to return */ public NodeC(CategoricalResults crResult) { super(); this.crResult = crResult; children = null; } /** * Creates a new node with children that start out with null (path disabled) * @param crResult the results to return * @param children the number of children this node has */ public NodeC(CategoricalResults crResult, int children) { super(children); this.crResult = crResult; } public NodeC(NodeC toClone) { super(toClone); this.crResult = toClone.crResult.clone(); } @Override public CategoricalResults localClassify(DataPoint dp) { return crResult; } @Override public int getPath(DataPoint dp) { return -1; } @Override public TreeNodeVisitor clone() { return new NodeC(this); } @Override public Collection<Integer> featuresUsed() { return Collections.EMPTY_SET; } } /** * Base node for regression and classification */ private static abstract class NodeBase extends TreeNodeVisitor { private static final long serialVersionUID = 6783491817922690901L; protected TreeNodeVisitor[] children; public NodeBase() { } public NodeBase(int children) { this.children = new TreeNodeVisitor[children]; } public NodeBase(NodeBase toClone) { if(toClone.children != null) { children = new TreeNodeVisitor[toClone.children.length]; for(int i = 0; i < toClone.children.length; i++) if(toClone.children[i] != null) children[i] = toClone.children[i].clone(); } } @Override public int childrenCount() { return children.length; } @Override public boolean isLeaf() { if(children == null) return true; for(int i = 0; i < children.length; i++) if(children[i] != null) return false; return true; } @Override public TreeNodeVisitor getChild(int child) { if(child < 0 || child > childrenCount()) return null; return children[child]; } @Override public void disablePath(int child) { if(!isLeaf()) children[child] = null; } @Override public boolean isPathDisabled(int child) { if(isLeaf()) return true; return children[child] == null; } } /** * Base node for regression */ private static class NodeR extends NodeBase { private static final long serialVersionUID = -2461046505444129890L; private double result; /** * Creates a new leaf node * @param result the result to return */ public NodeR(double result) { super(); this.result = result; } /** * Creates a new node with children that start out with null (path disabled) * @param crResult the results to return * @param children the number of children this node has */ public NodeR(double result, int children) { super(children); this.result = result; } public NodeR(NodeR toClone) { super(toClone); this.result = toClone.result; } @Override public double localRegress(DataPoint dp) { return result; } @Override public int getPath(DataPoint dp) { return -1; } @Override public TreeNodeVisitor clone() { return new NodeR(this); } @Override public Collection<Integer> featuresUsed() { return Collections.EMPTY_SET; } } /** * Base node for regression that splits no a numeric feature */ private static class NodeRNum extends NodeR { private static final long serialVersionUID = -6775472771777960211L; private int numerAtt; private double threshold; public NodeRNum(int numerAtt, double threshold, double result) { super(result, 2); this.numerAtt = numerAtt; this.threshold = threshold; } public NodeRNum(NodeRNum toClone) { super(toClone); this.numerAtt = toClone.numerAtt; this.threshold = toClone.threshold; } @Override public int getPath(DataPoint dp) { double val = dp.getNumericalValues().get(numerAtt); if( val <= threshold) return 0; else return 1; } @Override public TreeNodeVisitor clone() { return new NodeRNum(this); } @Override public Collection<Integer> featuresUsed() { IntList used = new IntList(1); used.add(numerAtt); return used; } } private class NodeRCat extends NodeR { private static final long serialVersionUID = 5868393594474661054L; /** * Categorical attribute to split on */ private int catAtt; /** * The cat values that go to the left branch, or null if no binary cats are being used */ private int[] leftBranch; public NodeRCat(int catAtt, int children, double result) { super(result, children); this.catAtt = catAtt; this.leftBranch = null; } public NodeRCat(int catAtt, Set<Integer> left, double result) { super(result, 2); this.catAtt = catAtt; this.leftBranch = new int[left.size()]; int pos = 0; for(int i : left) leftBranch[pos++] = i; Arrays.sort(leftBranch); } public NodeRCat(NodeRCat toClone) { super(toClone); this.catAtt = toClone.catAtt; if(toClone.leftBranch != null) this.leftBranch = Arrays.copyOf(toClone.leftBranch, toClone.leftBranch.length); } @Override public int getPath(DataPoint dp) { int[] catVals = dp.getCategoricalValues(); if (leftBranch == null) return catVals[catAtt]; else { if (Arrays.binarySearch(leftBranch, catVals[catAtt]) < 0) return 1; else return 0; } } @Override public Collection<Integer> featuresUsed() { IntList used = new IntList(1); used.add(catAtt+numNumericFeatures); return used; } @Override public TreeNodeVisitor clone() { return new NodeRCat(this); } } }

713 F.2d 1040 36 UCC Rep.Serv. 1569 Nancy L. BLY, Administratrix of Wayne A. Bly, deceased, Appellee,v.OTIS ELEVATOR COMPANY, a New Jersey Corporation, Appellant. No. 82-1430. United States Court of Appeals,Fourth Circuit. Argued Jan. 13, 1983.Decided Aug. 4, 1983. Jean-Pierre Garnier, Falls Church, Va., for appellant. Stephen D. Annand, Alexandria, Va. (Cohen & Annand, P.C., Charles Shepard Cox, Jr., Alexandria, Va., on brief), for appellee. Before WINTER, Chief Judge, PHILLIPS, Circuit Judge, and BUTZNER, Senior Circuit Judge. JAMES DICKSON PHILLIPS, Circuit Judge: 1 Otis Elevator Co. appeals from a final judgment in the district court, entered on a general jury verdict, holding it liable on a breach of warranty theory for personal injuries suffered by plaintiff's decedent. Because there was error in the jury instructions regarding the manufacturer's duty to warn as an element of the warranty of merchantability, we vacate the judgment and remand for a new trial. 2 * Plaintiff's decedent, Wayne Bly, was injured while operating a lift truck at the Avtex Fibers Plant in Virginia where he was employed. The truck was equipped in the front with a lift mechanism used to transport large canisters throughout the factory, and in the rear with a shin-high guard designed to provide protection from objects encountered while the truck was travelling in reverse. Operators of this and other lift trucks owned by Avtex stand on a small platform at the rear of the truck with their torso facing forward to maneuver the steering and other controls. Because the large canisters impair visibility to the front, the trucks are driven in reverse around the plant, necessitating that operators look over their shoulders in the direction of travel. 3 The particular truck Bly was driving when he sustained injury was manufactured by Baker Raulang Co., which sold the truck to the predecessor-in-interest of Avtex in 1944. Baker Raulang last exercised control over the truck in 1948, when it modified the truck's lift mechanism. 4 The parties agree that Bly was injured in September 1979 when he backed his truck without looking behind him and ran into a canister perched on the front-end lift of another truck. Because that canister was raised off the floor, the shin-high guard on Bly's truck slipped underneath, and Bly was crushed between the canister and the controls of his truck. 5 Bly's mother, as administratrix of his estate, then instituted the present wrongful death action against Otis Elevator Co.--the successor-in-interest to Baker Raulang--seeking relief on the basis of negligence and breach of warranty for alleged defects in the lift truck. The district court granted Otis's motion for directed verdict on the negligence count, holding as a matter of law that Bly's contributory negligence barred recovery. Accordingly, the case went to the jury solely on a theory of breach of the implied warranty of merchantability. 6 On appeal from the judgment entered on a jury verdict for the plaintiff, Otis contends there was insufficient evidence to support a verdict on this theory, and that the trial court's jury instructions erroneously included a charge on the manufacturer's duty to warn. We sketch briefly the contours of an action for breach of warranty, and then turn to the merits of these claims. II 7 Under Virginia law, which the parties agree controls resolution of this diversity action, manufacturers and sellers of defective products can be held liable on theories of negligence and breach of the implied warranty of merchantability. See Featherall v. Firestone Tire & Rubber Co., 219 Va. 949, 961-64, 252 S.E.2d 358, 367 (1979); Logan v. Montgomery Ward & Co., 216 Va. 425, 428, 219 S.E.2d 685, 687 (1975); see also Va.Code §§ 8.2-314, 8.2-318 (1965). A personal injury or wrongful death claimant may recover for breach of warranty under Virginia law by establishing "(1) that the goods were unreasonably dangerous either for the use to which they would ordinarily be put or for some other reasonably foreseeable purpose, and (2) that the unreasonably dangerous condition existed when the goods left the defendant's hands." Logan v. Montgomery Ward & Co., 216 Va. at 428, 219 S.E.2d at 687. A product can be "unreasonably dangerous" if defective in assembly or manufacture, see, e.g., Matthews v. Ford Motor Co., 479 F.2d 399, 400 (4th Cir.1973), if imprudently designed, see, e.g., Dreisonstok v. Volkswagenwerk, A.G., 489 F.2d 1066, 1071 (4th Cir.1974), or if not accompanied by adequate warnings about its hazardous properties, see, e.g., Spruill v. Boyle-Midway, Inc., 308 F.2d 79, 85-86 (4th Cir.1962). III 8 We dispose quickly of Otis's contention that the evidence adduced on trial was insufficient to support the jury verdict.1 9 The primary theory of plaintiff's breach of warranty claim was that the lift truck as designed was defective and "unreasonably dangerous" because its shin-high guards provided inadequate protection to the operator in rear-end collisions. In support of this theory, plaintiff produced substantial expert and documentary evidence which if accepted by the jury indicated that, by the 1940's when this truck was manufactured and sold, industry and government circles had recognized the need for waist-high guards on such lift trucks as a reasonable means of ensuring greater safety. Cf. Dreisonstok v. Volkswagenwerk, A.G., 489 F.2d 1066, 1073 (4th Cir.1974) ("if an article can be made safer and the hazard of harm may be mitigated 'by an alternative design or device at no substantial increase in price,' then the manufacturer has a duty to adopt such a design"). In light of this evidence suggesting that waist-high guards reflected prevailing industry views on the appropriate level of operator protection at the time this lift truck was manufactured and sold, coupled with expert testimony opining that the truck was, in this sense, defective in design, a jury could certainly conclude that a lesser standard of protection was "unreasonable" and a breach of the warranty of merchantability that proximately caused Bly's death.2 See Foster v. Ford Motor Co., 616 F.2d 1304, 1311 (5th Cir.1980); Weakley v. Fischbach & Moore, Inc., 515 F.2d 1260, 1268-69 (5th Cir.1975). 10 A jury verdict that unquestionably rested upon this theory of defective design would accordingly have been unassailable. That this was the basis of the verdict is not, however, certain, and in consequence the matter does not end here. IV 11 In addition to her theory of defective design, plaintiff sought to prove at trial that Otis, and its corporate predecessor, had breached the warranty of merchantability by failing to warn of the hazards of operating the lift truck. She asserted that the machine should have been accompanied, when sold in 1944 and refitted in 1948, by instructions and warnings concerning its potential dangers.3 Moreover, plaintiff introduced evidence that Otis had been notified in 1977 of a similar accident involving the same truck at the Avtex plant, in which an operator named Harry Wilson had been killed,4 to demonstrate Otis's further obligation at that later date to warn Avtex about the lift truck. 12 The trial court charged the jury on plaintiff's theory that "in the original design in '44, '48, and after the learning of the '77 death of Wilson ... it was necessary that the defendant give a warning in order for the forklift not to be unreasonably dangerous or cause an unreasonable risk of injury to users of the lift." The court summarized the factual elements that could give rise to this putative duty to warn, and then observed: "It was not a continuing duty to warn. This duty ended in '44 and at the latest '48. However, you may find, provided the conditions for a requirement of such a warning existed, a renewal of that obligation after notice of the Wilson death."5 13 We disagree with the general contention advanced by Otis on appeal that the trial court, by instructing on the duty to warn, necessarily erred by injecting negligence principles into a case being tried solely on a theory of breach of warranty. A manufacturer or seller that fails to warn of the dangers of its products can be held liable, in appropriate circumstances, either on a theory of negligence, see Featherall v. Firestone Tire & Rubber Co., 216 Va. 949, 961-64, 252 S.E.2d 358, 367 (1979); Restatement (Second) of Torts § 388 (1965), or on a theory of implied warranty, see Reid v. Eckerds Drugs, Inc., 40 N.C.App. 476, 482-83, 253 S.E.2d 344, 348-49 (N.C.Ct.App.), review denied, 297 N.C. 612, 257 S.E.2d 219 (1979), or strict liability in tort,6 see Greiner v. Volkswagenwerk Aktiengeselleschaft, 540 F.2d 85, 91 (3d Cir.1976); Martinez v. Dixie Carriers, Inc., 529 F.2d 457, 465 (5th Cir.1976); Jackson v. Coast Paint & Lacquer Co., 499 F.2d 809, 812 (9th Cir.1974); Restatement (Second) of Torts § 402A comments h, j (1965). A manufacturer may breach its implied warranty of merchantability by failing to warn or instruct concerning dangerous propensities or characteristics of a product even if that product is flawless in design and manufacture. See Reid v. Eckerds Drugs, Inc., 40 N.C.App. at 482-83, 253 S.E.2d at 348-49; cf. Davis v. Wyeth Laboratories, 399 F.2d 121, 128 (9th Cir.1968) (failure to warn gives rise to strict liability in tort, even where no manufacturing or design flaw). In essence, the product is rendered "unreasonably dangerous" by the absence of proper warnings or instructions. See Barber v. General Electric Co., 648 F.2d 1272, 1277 (10th Cir.1981); Karjala v. Johns-Manville Products Corp., 523 F.2d 155, 158 (8th Cir.1975); Alman Brothers Farms & Feed Mill, Inc. v. Diamond Laboratories, Inc., 437 F.2d 1295, 1303 (5th Cir.1971). 14 This duty to warn under a theory of implied warranty or strict liability in tort is obviously similar, in many respects, to the manufacturer's duty to warn under a negligence theory.7 But it differs in critical aspects important to the resolution of this appeal. 15 First, as the district court properly charged the jury, the duty to warn under an implied warranty theory focuses upon whether the lack of warning renders the product unreasonably dangerous; in contrast, a manufacturer will be liable in negligence for a failure to warn if its conduct is unreasonable.8 See Jackson v. Coast Paint & Lacquer Co., 499 F.2d 809, 812 (9th Cir.1974); Woodill v. Parke Davis & Co., 79 Ill.2d 26, 35, 37 Ill.Dec. 304, 308, 402 N.E.2d 194, 198 (1980); 2 L. Frumer & M. Friedman, Products Liability § 16A[f][vi] (1982).9 Second, under a negligence theory the duty to warn is continuous and is not interrupted by manufacture or sale of the product, see Large v. Bucyrus-Erie Co., 707 F.2d 94 (4th Cir.1983) (Winter, C.J., dissenting); W. Prosser, Law of Torts § 96, at 647 (4th ed. 1971), whereas the duty to warn under a theory of strict liability exists only at the time the product leaves the manufacturer's control, see Barber v. General Electric Co., 648 F.2d 1272, 1277 (10th Cir.1981); Logan v. Montgomery Ward & Co., 216 Va. 425, 428, 219 S.E.2d 685, 687 (1975); 1 R. Hursh & H. Bailey, American Law of Products Liability § 4.13, at 674 (2d ed. 1974).10 This distinction clearly reflects the emphasis in strict liability upon the danger of the product rather than the conduct of a manufacturer: if a product is not rendered unreasonably dangerous by the absence of warnings when it leaves the manufacturer's control, it cannot at some later date "become" unreasonably dangerous due to the lack of warnings. 16 In light of this latter distinction, the trial court erred in instructing the jury that it could find a "renewal" of the duty to warn after Otis received notice in 1977 of the death of Wilson in circumstances similar to the accident suffered by plaintiff's decedent. Under the breach of warranty theory by which this case was submitted to the jury, the manufacturer's duty to warn, as relevant to determining whether the lift truck was "unreasonably dangerous" when manufactured and sold, could have arisen only in 1944 and 1948 when Baker Raulang had control over the truck. By instructing the jury that the duty to warn could be revived by subsequent notice of injuries, the district court introduced elements of a negligence cause of action for failure to warn not applicable where the controlling theory is breach of warranty. This instruction made possible a verdict based on an erroneous theory of liability. Accordingly, it fatally taints the general jury verdict, which cannot stand. V 17 For the foregoing reasons, the judgment of the district court is vacated, and the case remanded for further proceedings consistent with this opinion. 18 VACATED AND REMANDED. BUTZNER, Senior Circuit Judge, dissenting: 19 I agree with the court's exposition of the law. I disagree about application of the law to the jury instructions. 20 The only proof of the forklift's unfitness was its defective design which made it unreasonably dangerous when the manufacturer sold it. 21 The court correctly told the jury that in order for the plaintiff to recover she had to show that the forklift was unfit when it was made available for the buyer's use in 1944 or 1948. The instructions explained that the word "unfit" meant "reasonably dangerous or a condition that subjects the users of the forklift to an unreasonable risk of injury." The court emphasized the importance of the time for determining whether the forklift was unfit by telling the jury that they should consider the state of technology and art on the subject "as it was in the years 1944 and at the latest '48 because they are the crucial years when this product was in the--last in the possession of the defendant." 22 The instructions on warning must be considered in the context of the entire charge. They cannot be divorced from the following paragraph of the charge: 23 In order to find an obligation to warn, however, you must first find that there was an unfit condition to warn about and that a warning was necessary to prevent the forklift from being unreasonably dangerous or subjecting its users to an unreasonable risk of injury. 24 Thus, the jury could not base its verdict solely on a lack of warning in 1977. Under the instructions of the court, the jury first had to find "an unfit condition to warn about." The only "unfit condition" was defective design which the court had told the jury must be determined by the state of the art in the 1940's. 25 Parenthetically, if the manufacturer had fulfilled its warranty by giving adequate warning when it sold the forklift in the 1940's, the court's instruction about warning in 1977 would have injected error in the case. But this is not the situation, for the evidence is undisputed that the manufacturer gave no warning when it sold the defectively designed machine. 26 In sum, considering the charge as a whole, I believe that the instructions about warning did not mislead the jury. In order for the plaintiff to prevail, the jury was required to find that the forklift was unfit when it left the manufacturer's hands. Because the evidence was sufficient to support such a finding, I would affirm the judgment entered on the verdict of the jury. 1 Plaintiff's cause of action for breach of warranty is not time-barred by the four-year contracts statute of limitations, which ran from the tender of delivery in 1944 or 1948; actions under Virginia law grounded in a personal injury-products liability factual pattern are governed by the torts statute, which runs from the time of injury. See Tyler v. R.R. Street & Co., 322 F.Supp. 541, 543 (E.D.Va.1971) 2 In 1954, Avtex sold the lift truck in issue, and repurchased it in 1963. Thereafter, the truck was used continuously by Avtex until Bly's accident in 1979 Otis contends that this sale and repurchase should insulate it from liability for breach of warranty, because the truck was no longer "in the stream of commerce" initiated by the manufacturer. We find no merit to this argument, given that Virginia has expressly abolished any vertical privity requirement in actions for breach of warranty. See Va.Code § 8.2-318 (1965). If the defect existed at the time of sale, see Logan v. Montgomery Ward & Co., 216 Va. 425, 428, 219 S.E.2d 685, 687 (1975), and if the product was not substantially changed after the time of sale, cf. Restatement (Second) of Torts § 402A(1)(b) (1965), it is of no moment that the product has changed hands in the interval after leaving the manufacturer's control. 3 At oral argument, for instance, counsel for plaintiff suggested that operators of the lift truck should have been instructed to face their torsos to the rear of the truck, operating the controls by reaching behind them 4 More precisely, evidence introduced at trial indicated that Otis was notified in 1977 of a suit brought by Wilson's estate alleging, inter alia, that the shin-high platform guards on the lift truck were inadequate and defective, and had caused Wilson's death 5 In total compass, the duty-to-warn charge, including that portion quoted by the dissent, was as follows: In addition to defective, unfit design, the plaintiff asserts that originally in '44 through '48, and again after the defendant was notified of the death of Wilson while Wilson was operating this forklift ... it was necessary that the defendant give a warning in order for the forklift not to be unreasonably dangerous or cause an unreasonable risk of injury to users of the lift. * * * Where, however, any such risk would be apparent to any person that the manufacturer reasonably should foresee might use the product there is no duty to warn. * * * The duty to warn if it existed was during the times, the years when the forklift was made available to Avtex. It was not a continuing duty to warn. This duty ended in '44 and at the latest '48. However, you may find provided the conditions for a requirement of such a warning existed, a renewal of that obligation after notice of the Wilson death. In order to find an obligation to warn, however, you must first find that there was an unfit condition to warn about and that a warning was necessary to prevent the forklift from being unreasonably dangerous or subjecting its users to an unreasonable risk of injury. You've heard evidence of the Wilson death and the Wilson lawsuit filed by his estate against this same defendant, the Otis Elevator Company. This testimony is solely for the purpose of showing that the defendant had notice of a possible defect in the design of the lift insofar as not having any waist-high guards and is to be considered by you in determining whether that notice renewed or gave rise to any obligation on the part of Otis to warn in order to fulfill its warranty obligation. During its deliberations, the jury requested and was given supplemental instructions on the duty to warn. In these the court reiterated the substance of its prior instructions, including: "The duty to warn, if it existed, was during ... the forties, when the forklift was made available to Avtex. It was not a continuing duty thereafter. However, you may find, provided the conditions for the requirement of such a warning existed, a renewal of that obligation after the notice of the Wilson death." 6 Virginia has not in terms adopted strict liability in tort, see Restatement (Second) of Torts § 402A (1965), as yet another and separate theory of products liability, but it is well established that warranty liability under Virginia law for personal injuries caused by defective products is the functional equivalent of strict tort under the Restatement formulation. See Matthews v. Ford Motor Co., 479 F.2d 399, 401 n. 2 (4th Cir.1973); Chestnut v. Ford Motor Co., 445 F.2d 967, 968 & n. 1 (4th Cir.1971); Speidel, The Virginia "Anti-Privity" Statute: Strict Products Liability Under the Uniform Commercial Code, 51 Va.L.Rev. 804, 822, 828-29 & n. 61 (1965). See generally Basko v. Sterling Drug, Inc., 416 F.2d 417, 427-28 (2d Cir.1969); J. White & R. Summers, Uniform Commercial Code § 9-7, at 355-56 (2d ed. 1980). Accordingly, for purposes of analysis, we draw upon cases decided under the rubric of strict liability in tort to gauge the elements of Virginia's implied warranty of merchantability 7 We disagree with plaintiff's contention that Gardner v. Q.H.S., Inc., 448 F.2d 238 (4th Cir.1971), equated negligence and implied warranty principles with respect to the duty to warn. After discussing the contours of a cause of action for negligent failure to warn, the court noted that "[t]he same is true with respect to a cause of action for breach of an implied warranty under the Uniform Commercial Code." Id. at 242. Contrary to plaintiff's suggestion, however, this reference was to the relaxation of privity requirements and the need to prove the "inherent danger" of a product, not to the elements of the duty to warn. Cf. id. at 243 ("the separate theories of negligence and breach of warranty [for failure to provide adequate warning of hidden danger] are not always coextensive") 8 As many courts and commentators have noted, see, e.g., Werner v. Upjohn Co., 628 F.2d 848, 858 (4th Cir.1980), cert. denied, 449 U.S. 1080, 101 S.Ct. 862, 66 L.Ed.2d 804 (1981); Chestnut v. Ford Motor Co., 445 F.2d 967, 968 & n. 1 (4th Cir.1971); Kidwell, The Duty to Warn: A Description of the Model of Decision, 53 Tex.L.Rev. 1375, 1377-78 (1975); Noel, Products Defective Because of Inadequate Directions or Warnings, 23 Sw.L.J. 256, 267 (1969), the scope of liability imposed under a breach-of-warranty theory in many if not most duty-to-warn cases will be identical to that imposed under a negligence theory. Yet there may well be instances in which a manufacturer has "reason to know" that its product is "likely to be dangerous"--triggering liability under the Restatement (Second) of Torts § 388 (1977)--but its failure to provide a warning does not render the product "unreasonably dangerous" under a theory of implied warranty or strict liability in tort 9 See also Keeton, Products Liability--Design Hazards and the Meaning of Defect, 10 Cum.L.Rev. 293, 315 n. 87 (1979); Wade, On Product "Design Defects" and Their Actionability, 33 Vand.L.Rev. 551, 553 (1980) 10 But see Comment, Products Liability: Post-Sale Warnings, 1978 Ariz.St.L.J. 49, 62 (suggesting that seller may have duty to warn after sale under strict liability theory)

Leaf Gas Exchange and Chlorophyll a Fluorescence in Maize Leaves Infected with Stenocarpella macrospora. This study investigated the effect of macrospora leaf spot (MLS), caused by Stenocarpella macrospora, on photosynthetic gas exchange parameters and chlorophyll a fluorescence parameters determined in leaves of plants from two maize cultivars ('ECVSCS155' and 'HIB 32R48H') susceptible and highly susceptible, respectively, to S. macrospora. MLS severity was significantly lower in the leaves of plants from ECVSCS155 relative to the leaves of plants from HIB 32R48H. In both cultivars, net CO2 assimilation rate, stomatal conductance, and transpiration rate significantly decreased, while the internal to ambient CO2 concentration ratio increased in inoculated plants relative to noninoculated plants. The initial fluorescence and nonphotochemical quenching significantly increased in inoculated plants of ECVSCS155 and HIB 32R48H, respectively, relative to noninoculated plants. The maximum fluorescence, maximum PSII quantum efficiency, coefficient for photochemical quenching, and electron transport rate significantly decreased in inoculated plants relative to noninoculated plants. For both cultivars, concentrations of total chlorophyll (Chl) (a+b) and carotenoids and the Chl a/b ratio significantly decreased in inoculated plants relative to noninoculated plants. In conclusion, the results from the present study demonstrate, for the first time, that photosynthesis in the leaves of maize plants is dramatically affected during the infection process of S. macrospora, and impacts are primarily associated with limitations of a diffusive and biochemical nature.

The group in a press release made available to Vanguard Live said that the mammoth crowd that showed at the airport Friday to welcome Ekweremadu are rented. “We wish to disassociate ourselves from the rented crowd the received Ike Ekweremadu at Enugu airport,” The statement read. According to the Council, the crowd were staged by an Enugu lawmaker from Ezeagu local government area (LGA) of the state. “A desperate Enugu lawmaker from Ezeagu Lga who’s a stooge of Ekweremadu in Enugu State house of assembly, went to Onitsha main market and Obete Enugu market to persuade touts and beggers to masquerade as Ohanaeze youths,” “he was betrayed by lack of communication skills exhibited by a 60 years old man parading himself as a Youth leader”. The Council, however, said it disapproved of the attack on Ekeweremadu in Germany and vowed to sanction the lawmaker involved in the organisation of the false crowd.

TEMPO: a contemporary model for police education and training about mental illness. Given the increasing number of interactions between police and people with mental illnesses (PMI), there has been widespread interest in the development of education for police about how best to interact with PMI. This paper reflects the review of current practice in a variety of jurisdictions across Canada as well as in the United States (U.S.), the United Kingdom (U.K.) and Australia; it proposes a comprehensive model of learning based on the literature that addresses not only the content in the narrow sense but also the importance of broader contextual knowledge and understanding in developing effective education and training. Embedded in the principles articulated in the Mental Health Strategy for Canada, the TEMPO (Training and Education about Mental illness for Police Organizations) model is a multilevel learning strategy for Canadian police personnel. Learning objectives and key principles are articulated in order to ensure the model is applicable to a wide range of police agencies and individual jurisdictional needs. In addition to providing a firm basis of factual knowledge for police personnel, the resultant model embraces a human rights/anti-stigma philosophy, provides for a range of education appropriate to diverse police audiences, emphasizes a systems approach to police/mental health liaison activities and addresses issues related to the delivery and implementation of police education and training.

Q: What are the consequences of having many files in a directory in an ext4 filesystem? My task is to store a list of JSONs on disk (without using any database) and I have these options: Store them in a single, large file. Store them in separate files, keyed by their IDs. Personally I prefer the second option since it allows direct addressing any JSON by their ID without ever having to touch any other JSONs. However, there are almost 0.1 to 1 million JSON entries and I'm afraid of the possible negative consequences on the underlying filesystem (ext4 in my case): Will this go over filesystem limits about the number of files (either in a directory or in a whole filesystem)? Will this cause a slowdown while retrieving a specific ID? To be more specific, I believe the list of files under a directory are maintained by the directory's inode structure, but I'm not sure what data structures (list or map) it uses to keep the file list. Is there any performance gain in the lookup if I use a hierarchy of directories? For example, put 0123456789.json into root/01/0123456789.json instead of root/0123456789.json? A: Having 1 million files in a single directory would slow things down, but so would parsing an aggregate JSON with 1 million entries. Your best bet is indeed to use hashed directories, but you probably want to go two levels deep rather than just one. Namely, put 0123456789.json in root/0/01/0123456789.json, and 987654321.json in root/9/98/987654321.json.

Q: Day of week in GWT / Java Date api How do I get day of week for a given Date in GWT on client side? I know I can get it from Date object usin getDay() but I am not sure if this is the best way. Is there any alternative? Is it acceptable & reliable to use deprecated methods of java Date api? A: Try DateTimeFormat format = DateTimeFormat.getFormat("c"); // try with "E" pattern also String dayOfWeek = format.format(new Date()); System.out.println(dayOfWeek); For more patterns have a look at DateTimeFormat output 0 for sunday

1998 ARRS President's Award. The potential of in vivo vascular tissue engineering for the treatment of vascular thrombosis: a preliminary report. American Roentgen Ray Society. Current gene therapy and tissue engineering protocols suffer from a number of inherent limitations. In this study, we examine the feasibility of a new approach for the treatment of vascular thrombosis: in vivo tissue engineering. Rabbit femoral veins were transfected in situ with either a previously characterized adenoviral-construct-expressing tissue plasminogen activator or a viral (adenoviral-construct-expressing beta-galactosidase) or nonviral (buffer) control and used as cross sections (n = 3). Treated veins were then harvested and grafted into the ipsilateral common femoral artery as an interposition vein graft. A potent stimulus for thrombus formation was then introduced into the recipient artery downstream of the graft. Six days later, the rabbits were sacrificed, and the grafts and downstream arteries were harvested. Vessel segments were then examined for thrombus according to defined anatomic zones. Transfection efficiency and presence of smooth muscle cells in the vein graft were also evaluated. The engineered vein graft showed a significant reduction in thrombus formation within both the graft and the downstream artery relative to nonviral (buffer) and viral (adenoviral-Rous sarcoma virus beta-galactosidase [Adv/RSV-betagal]) controls. Underlying endothelial cell transfection efficiency of 90% was observed in viral controls (Adv/RSV-betagal). A 2.4-fold increase in smooth muscle alpha-actin positive cells in the engineered vein graft was seen compared with nonviral (phosphate-buffered saline) controls. A 10-fold increase in smooth muscle alpha-actin-positive cells in the engineered vein graft relative to viral (Adv/RSV-betagal) controls was also observed. In vivo tissue engineering is a new paradigm in molecular medicine that is a viable alternative to conventional gene therapy and tissue engineering for the treatment of vascular thrombosis.

Department of Justice Authority to Provide “Protective Custody” for Defectors W hile any com ponent o f the D epartm ent o f Justice may contract w ith the D epartm ent of state to perform the latter's security functions, the D epartm ent o f State is not a uthor­ ized to provide protective custody to defectors w ho are neither leading figures in, nor direct representatives of, their governm ent. T he A ttorney G eneral has authority under the Im m igration and Nationality A ct to prevent departure o f an alien defector w ho is being repatriated under duress and might, in a particular case, have discretionary authority to provide some sort o f protective custody for that defector. U nder § 235(b) o f the Im m igration and N ationality A ct, the Im m igration and N aturaliza­ tion Service has authority to detain a defector w ho is deportable or excludable, until such time as he is granted political asylum. If a defector is assaulted,-harassed, specifically threatened, or abducted, so as to bring into play one of several potentially applicable federal crim inal statutes, federal law enforce­ ment agencies may be authorized to play a role in his protection. T he Secretary o f State may designate any defector an official guest in order to make it a federal offense to assault, harass, intim idate, coerce, imprison, threaten, kidnap, o r kill the defector. January 17, 1980 M EM ORANDUM OPIN IO N FOR TH E ACTIN G ASSOCIA TE ATTORNEY G EN ER A L This responds to your inquiry regarding the authority of Department of Justice agencies to protect aliens who have defected to the United States. You ask us to assume: (1) That the defector is not an obvious source of intelligence information; (2) That the defector is within the United States and at or near an office of the Federal Bureau of Investigations (FBI) or the Immigration and Naturalization Service (INS); (3) That the defector is seeking political asylum in the United States; (4) That neither the Department of State nor any other govern­ ment agency has “firm information” that the defector is threatened with either forced repatriation or bodily harm; and 348  (5) That “the circumstances of the defection are such that a reasonable person might wish to take security precautions.” You ask whether, in such a case, any component of the Department of Justice would have authority to fulfill a request made by the Depart­ ment of State to provide “protective custody” for the defector. You do not define “protective custody.” We shall assume that it does not involve taking any action against the defector’s will, and that the defector consents to any arrangement made for his protection. We shall also assume that it involves at least protecting the defector against the possibility of physical attack. For the reasons stated below, we believe that no component of the Department of Justice has authority even to protect defectors against the possibility of physical attack in all cases of the sort you describe, although certain agencies may have authority to provide protection against the danger of physical attack, and perhaps a form of protective custody as well, in some cases. 1. U.S. Marshals Service Acting Under Agreement With the Department of State Under 31 U.S.C. § 686(a), “[a]ny executive department . . . or any bureau or office thereof . . . may place orders with any other such department, establishment, bureau, or office, for . . . work, or services, of any kind that such requisitioned Federal agency may be in a position to supply or equipped to render . . . .” This provision would authorize any component of the Department of Justice to contract with the Department of State to perform the latter’s security functions. Since the Marshals Service administers the federal witness protection program, 18 U.S.C. prec. § 3481, 28 C.F.R. § 0.111(c), it is the agency most clearly “in a position to” protect defectors. Thus, if the Department of State can itself provide protective custody for defectors, the Marshals Serv­ ice can also do so under an agreement with it.1 In most cases of the sort you describe, however, the Department of State lacks the authority even to protect defectors against the possibil­ ity of a physical attack. Under 22 U.S.C. § 2666, qualified Department of State security officers “are authorized to carry firearms for the purpose of protecting heads of foreign states, official representatives of foreign governments, and other distinguished visitors to the United States . . . and members of the immediate families of any such 1 The M arshals Service currently has an agreem ent w ith the D epartm ent o f State to “ provide m anpow er and equipm ent as determ ined by the M arshals Service, in o rd e r to augm ent the State D epartm ent's capacity to carry out its protective functions in the most secure m anner.” "subject to m anpow er availability and norm al mission requirem ents.” This agreem ent provides that the D e p art­ ment o f State is to reim burse the M arshals Service for its w ork. W hether the requisitioning agency must reimburse the agency providing the service depends on the term s o f the agencies* respective authorization and appropriations statutes. Sec 13 Com p. G en. 234 (1934): 34 Com p. G en. 42 (1954). 349  persons,” 2 No other statute gives the Department of State explicit au­ thority to protect anyone. It might be suggested that visiting athletes and artists, for example, are “distinguished foreign visitors” and perhaps “official representatives” of their governments; if they defect, § 2666 might authorize their protection. Tw o arguments militate against this interpretation, however. First, the original version of § 2666, in effect until 1975, authorized security officers of the Department of State to protect “heads of for­ eign states, high officials of foreign governments and other distin­ guished visitors to the United States . . . and official representatives of foreign governments and of the United States attending international conferences, or performing special missions.” Pub. L. No. 84-104, 69 Stat. 188 (1955). There is no indication in the legislative history that the 1975 rewording was intended to alter the scope of the statute. See S. Rep. No. 337, 94th Cong., 1st Sess. 22 (1975). Because the term “distin­ guished visitors” was linked with “high officials of foreign govern­ ments” in the original version of § 2666 and even more clearly in its legislative history, see, e.g., H.R. Rep. No. 468, 84th Cong., 1st Sess. 1 (1955), “distinguished visitors” must, we believe, be limited to leading political, diplomatic, and military figures. We doubt it can be extended to include all prominent foreign visitors who might happen to defect while in the United States. The original version of § 2666 also suggests that the “official representatives” protected are those “attending inter­ national conferences, or performing special missions.” Again, Congress seemed to have in mind official conferences concerned with political, military, or diplomatic matters; one of the justifications for the bill was the need “to guarantee the safety from compromise of the vast amount of highly classified material needed at an international conference.” S. Rep. No. 552, 84th Cong., 1st Sess. 2 (1955). Congress may have intended to expand this category somewhat by omitting the reference to international conferences and special missions, but there is no reason to believe that “official representatives” includes persons other than those acting directly on behalf of their respective governments. The second argument reinforces this conclusion. In 1972 Congress amended several statutes to make it a federal crime to assault, threaten, harass, kidnap, or kill “official guests.” Pub. L. No. 92-539, §§ 101-301, 86 Stat. 1070 (1972), amending 18 U.S.C. §§ 112, 1116, 1201. Congress created this category of “official guests” because it wanted federal criminal laws to “operate to protect the rights of visiting artists, aca­ demic and scientific groups, and other groups and individuals who ought not be beyond the pale of Federal concern.” S. Rep. No. 1105, 92d Cong., 2d Sess. 9 (1972). Congress thought that such visitors would 2 U nder 22 U.S.C. § 2666, D epartm ent o f State security officers are also authorized to protect “ the Secretary o f State, the D eputy Secretary o f State, official representatives o f the U nited States G ov ern m en t, and m em bers o f the im m ediate families o f any such persons.'* 350 otherwise receive no federal protection against such offenses, see, e.g., id. at 7; yet at the time, the predecessor of § 2666 had been in effect for 17 years. When Congress amended §2666 in 1975, it did not include “official guests” in the new version of the statute; it retained the term “official representatives.” This again suggests that Congress did not wish to authorize Department of State security officers to protect even such prominent foreign visitors as athletes, artists, and academics.3 For these reasons, we seriously doubt that the Department of State has authority to request the Marshals Service to protect defectors who are neither leading figures in, nor direct representatives of, their gov­ ernments. Moreover, it is unlikely that an “official representative” would retain his status if the country he purported to represent at­ tempted to strip him of it. The Marshals Service would, of course, be able to protect “distinguished foreign visitors” who defect 4—presum­ ably a small proportion of the cases we are considering here. 2. FBI Authority Under 28 U.S.C. §553 (1), (3), the FBI is empowered “to detect and prosecute crimes against the United States” and “to conduct such other investigations regarding official matters under the control of the D e­ partment of Justice and the Department of State as may be directed by the Attorney General.” W hatever authority these provisions may give to protect potential victims of federal crimes against whom a specific threat has been made, we believe that they do not authorize the FBI to protect defectors in the circumstances we are considering here. Com­ pare 28 U.S.C. §553 with 18 U.S.C. §3481 note (specifically authorizing the Attorney General to “provide for the security o f ’ government witnesses who testify against alleged participants in organized crim e.)5 3 22 U.S.C. § 2667 em pow ers D epartm ent o f State security officers "engaged in the perform ance o f the duties prescribed in section 2666" to "arrest w ithout w arrant and deliver into custody any person violating section . . . 112 o f title 18 in their presence o r if they have reasonable grounds to believe that the person to be arrested has com m itted o r is com m itting such a violation.” See also 22 C .F .R . §2.1. U nder 18 U.S.C. § 112, it is a crim e to assault, harass, intimidate, coerce, threaten, o r similarly harm foreign officials, internationally p rotected persons, o r official guests. F or reasons w e give in section 5, infra, we believe that §2667 may enable D epartm ent o f State security officers to provide some aid to defectors w ho have been specifically threatened o r harm ed. But for tw o reasons, §2667 cannot be read to authorize D epartm ent o f State security officers to protect "official guests," o r others w ithin the scope o f § 112, if they have not been specifically threatened. First, security officers w hose mission was to provide such protection w ould be "engaged in the perform ance o f . . . duties" not enum erated in § 2666. Second, we doubt that the au th o rity to enforce a statute by arresting violators implies the authority to protect persons w hen no specific threat has been made, especially w hen another statute expressly authorizes the protection o f a smaller class o f persons. 4 Indeed, the M arshals Service may already have this pow er under the existing agreem ent, sue note I supra. 5 T h e D irector o f C entral Intelligence, the A tto rn ey G eneral, and the C om m issioner o f Im m igration and N aturalization can authorize certain aliens to enter the U nited States, notw ithstanding o ther immigration laws, if their entry "is in the interest o f national security o r essential to the furtherance of the national intelligence mission." SO U.S.C. §403h. Pursuant to this authority, the National Security Council and the D irecto r o f C entral Intelligence have established a program for dealing w ith defectors w ho are valuable to intelligence agencies. T he FBI plays a role in this program , but the program plainly omits authority for the FBI o r any o th er agency to house o r otherw ise maintain defectors o f the sort you describe. This, too, suggests the FBI has no role in providing protective custody in the circum stances w e are considering here. 351  3. The Attorney General’s Authority To Enforce § 215 of the Immigration and Nationality Act Under 8 U.S.C. § 1103(a), the Attorney General is “charged with the administration and enforcement of [the Immigration and Nationality Act] and all other laws relating to the immigration and naturalization of aliens.” Ordinarily, he carries out this responsibility through the Immi­ gration and Naturalization Service. Section 215(a)(1) of the Act pro­ vides: Unless otherwise ordered by the President, it shall be unlawful . . . for any alien to depart from . . . the United States except under such reasonable rules, regula­ tions, and orders, and subject to such limitations and ex­ ceptions as the President may prescribe. 8 U.S.C. § 1185(a)(1). It appears to us that current regulations would not, in general, authorize the Attorney General to prevent the depar­ ture of a defector in the circumstances you describe.6 But we believe that § 215(a)(1) would authorize regulations prohibiting the departure of, for example, an alien defector who was being repatriated either under duress or in circumstances that cast doubt on the ability of the United States to protect defectors.7 If a regulation were issued that c T h e regulations, 22 C .F .R . §46.2, provide that “ [n]o alien shall depart, or attem pt to depart, from the U nited States if his d ep artu re w ould be prejudicial to the interests o f the United States under the provisions o f [22 C .F .R .] §46.3.*' Section 46.3 then specifies the categories o f aliens w hose departure “ shall be deem ed prejudicial to the interest o f the U nited States." N one o f these categories will apply to all defectors o f the kind you describe, and few o f the categories are likely to apply to any. Section 46.3(g), for exam ple, prohibits the d ep artu re o f “ (a]ny alien w ho is needed in the United States as a w itness in . . . any criminal case under investigation o r pending in a court in the United States." An investigation o f a possible violation o f som e state o r federal crim inal statute, see section 5 infra, might be w arran ted in some cases o f the kind you describe and the d efector might be needed as a witness at that investigation. But if. as you specified, there is no “ firm inform ation” that the defector is likely to be abducted o r physically harm ed, a crim inal investigation will generally not be w arranted. Section 46.3(h) prohibits the dep artu re o f “ (a]ny alien w h o is needed in the United States in connection w ith any investigation o r proceeding being, o r soon to be. conducted by any official executive, legislative, o r judicial agency in the U nited States o r by any governm ental com m ittee, board, bureau, commission, o r body in the U nited Stales, w h eth er national, state, o r local.*' 22 C .F .R . § 46.3(h). This provision might appear to allow a defector to be kept in the U nited States, if. for exam ple, a governm ent body planned to ask him form ally about his reception by A m erican officials or about relatives or assets rem aining in the nation from w hich from he defected and possible diplom atic action concerning them. But w e doubt that this provision w ould be construed to reach cases in w hich the formal inquiry is a pretext and the true “ prejudice to the interests o f the U nited States’* stems not from the alien's failure to appear at the inquiry but from the m anner o r circum stances in w hich he departed. Invoking 22 C .F .R . §46.3(k) w ould present the same problem. It effectively prohibits the departure o f an alien w hose case “ involves circum stances o f a [character] sim ilar” to the o ther categories under §46.3. W hile not all o f these categories involve, for exam ple, national security or national defense, see, e.g., 22 C .F .R . §46.3 (0. (g). (h). they all do involve, at the least, aliens w hose personal characteristics— their know ledge, intentions, o r legal liabilities—make their d eparture prejudicial to the U nited States. N one involves an alien w ho does not wish to depart; none involves an alien w hose personal ch aracteristics are unim portant but w ho w ould depart in a m anner or under circum stances w hich reflect unfavorably on the United States. F o r these reasons, we believe that new regulations should be issued if the A ttorney G eneral is to exercise his p o w er under § 215(a)(1) to prevent the departure of defectors in the circum stances you mention. 7 By its terms, § 2 1 5(a) grants the President full pow er to regulate the departure o f aliens, requiring only that the regulations be reasonable. T h e legislative history o f § 215(a) show s that C ongress C o ntinue d 352 effectively prohibited the departure of a defector in the circumstances you describe, we believe the § 215(a)(1) might, in a particular case, authorize the Attorney General to provide some form of protective custody for that defector. Nothing in § 215(a)(1) suggests that the Attorney General must me­ chanically refrain from acting until a defector whose departure he is authorized to prevent is boarding an airplane. Implicit in the Attorney General’s duty to enforce the Immigration and Nationality Act is the authority to use all reasonable and necessary means to see that it is enforced. See, e.g., United States v. Krapf 285 F.2d 647, 650 (3rd Cir. 1961); United States v. Jones, 204 F.2d 745, 754 (7th Cir. 1953); United States v. Kelly, 55 F.2d 67 (2d Cir. 1932). In addition, law enforcement authorities customarily have great discretion to decide how to enforce the law. Thus, the Attorney General may determine in a particular case that in order to prevent a defector from departing he must, for exam­ ple, keep the defector under surveillance so that he can act quickly to prevent a departure or abduction. For similar reasons, the Attorney General would, we believe, be entitled to screen a defector’s contacts with other people or to guard the defector in order to prevent attempts to coerce the defector to leave.8 These steps would appear to be the kind of protective custody you have in, mind. They would, we believe, be authorized if they were part of a good faith effort to enforce § 215(a)(1) in light of its underlying policies. Indeed, the structure of § 215 suggests that the Attorney General has unusually broad discretion to decide which measures are necessary to prevent violations of that section. Section 215(a)(1) declares that it is “unlawful” for certain aliens to leave the United States but prescribes no penalties for violations. Those penalties, which applied both to aliens who illegally entered or departed the United States and to American citizens who attempted to enter or depart without passports, see Immi­ gration and Nationality Act, Pub. L. No. 82-414, ch. 477, § 215(a)(1), 66 Stat. 190 (1952) (prior to 1978 amendment), were repealed by Con­ gress in 1978. Pub. L. No. 95-426, § 707(d), 92 Stat. 993. The legislative history of the repeal suggests that while Congress did not wish to “obstruct” or penalize the travel of American citizens, it intended to leave intact the President’s authority to regulate the entry or departure iniended the President 10 have “ broad and com prehensive power.*' “ w ide discretion and w ide a u th o r­ ity o f action.*' H.R. Rep. No. 485, 65th Cong., 2d Sess. 2-3 (1918) (accom panying Act of M ay 22, 1918. Pub. L. No. 65-154. ch. 81, § 1(a), 40 Slat. 559. w hich § 2 1 5(a) essentially reenacted. See H.R. Rep. No. 1365. 82d Cong.. 2d Sess. (1952)). T here is no reason to believe that C ongress did not intend the President to use this pow er to pursue the im portant humanitarian and foreign policy aims that w ould be served by preventing the departure o f aliens w ho do not wish to leave. Indeed. C ongress envisioned the President using his authority as a “ counterstroke*' against the “ propaganda” efforts of "hostile nations." H.R. Rep. No. 485. 65lh Cong.. 2d Sess. 3 (1918). MIn this connection we em phasize o u r assum ption that the defector consents to the steps the A ttorney G eneral is taking to protect him. Ii is not at all clear that the A ttorney G eneral can legally isolate a defector in this w ay w ithout his consent. Also, we assume that the A ttorney G eneral will com ply w ith any international obligations the U nited States has to permit contacts w ith defectors. 353 of aliens under § 215(a)(1). See 124 Cong. Rec. 15770 (May 31, 1978) (remarks of Rep. Eilberg). Moreover, nothing in the language of § 215(a)(1) suggests that it is intended to be merely admonitory. Com­ pare 8 U.S.C. § 1185(a) with 36 U.S.C. § 175 (flag code); see Holmes v. Wallace, 407 F. Supp. 493, 494-97 (M.D. Ala. 1976). The primary purpose of § 215(a)(1), then, must be to authorize preventive action, either administrative or judicial, against aliens who are about to depart illegally. Several other sections of the Immigration and Nationality Act give great discretion to the administrators charged with their enforce­ ment, thus suggesting that Congress envisioned administrative not judi­ cial action to enforce § 215(a)(1); in addition, as we have said, the Attorney General is specifically charged with enforcing the Act, 8 U.S.C. § 1103(a). Since prevention is the only means of enforcing § 215(a)(1), and the Attorney General is primarily responsible for en­ forcing it, one may reasonably infer that the Attorney General can act more vigorously to prevent violations of § 215(a)(1) than he might act in preventing violations of statutes with more diverse enforcement mechanisms. This further supports the conclusion that in some cases § 215(a)(1), by implication, authorizes the Attorney General to provide defectors whose departure he can prevent with a form of protective custody. Since Congress has not explicitly authorized such protective custody of defectors, however, compare 8 U.S.C. § 1185(a)(1) with 18 U.S.C. prec. § 3481, we would advise that the Department take steps to inform the appropriations committees of the Senate and House that we regard § 215(a)(1) as authority to do so in isolated instances and on a tempo­ rary basis in connection with the enforcement of § 215(a)(1). 4. Delaying the Grant of Political Asylum Until an alien is granted political asylum, the Immigration and Natu­ ralization Service has authority to detain him if he fits either of two categories. We believe it is reasonable to assume that a defector who is detained can be adequately protected. Under § 235(b) of the Immigra­ tion and Nationality Act, 8 U.S.C. § 1225(b): Every alien [with exceptions not relevant here] who may not appear . . . at the port of arrival to be clearly and beyond a doubt entitled to land shall be detained for further inquiry to be conducted by a special inquiry officer. See generally 8 C.F.R. § 235.3(b). If an alien has legally entered the country, § 235(b) cannot authorize his detention. But while attempting to defect, an alien may render himself technically deportable—perhaps by violating a condition of his visa—or may be about to render himself 354  deportable. Section 242(a), 8 U.S.C. § 1252(a), would then apply: Pending a determination of deportability in the case of any alien . . . such alien may, upon warrant of the A ttor­ ney General, be arrested and taken into custody. Any such alien taken into custody may, in the discretion of the Attorney General and pending such final determination of deportability, (1) be continued in custody, or (2) be re­ leased under bond . . . or (3) be released on conditional parole. But such bond or parole . . . may be revoked at any time by the Attorney General, in his discretion, and the alien may be returned to custody under the warrant which initiated the proceedings against him and detained until final determination of his deportability.[9] The Supreme Court has rejected the view that § 242(a) authorizes the Attorney General to detain an alien only if the alien’s detention is necessary to secure his appearance at a deportation hearing. See Carlson v. London, 342 U.S. 524, 534, 541 n.35 (1952). The Court has suggested that an alien may be detained pending deportation proceedings when^ ever the Attorney General has a “reasonable apprehension” that releas­ ing the alien will injure the national interest, see id. at 538, 542, and has not required that the “reasonable apprehension” be supported with specific threats or facts; broad generalizations suffice. See id. at 541, 544. Moreover, as the Court has acknowledged, the legislative history of § 242(a) makes plain Congress’ intention to vest the Attorney G en­ eral with considerable discretion in deciding which aliens to detain. See id. at 540-41.10 Since the Attorney General can reasonably conclude that the national interest would be injured if a defector were severely harassed or forcibly repatriated, we believe that in the cases we are considering here § 242(a) would authorize the detention of a deportable defector who consented11 to be detained. Since most aliens who have been granted political asylum will not be deportable or excludable, it appears that the Immigration and Natural­ ization Service has authority to detain a defector only until he is granted asylum. A defector who is entering the country is likely to submit his application for asylum to an immigration judge, “who shall consider that application in connection with an exclusion hearing. . . 44 Fed. Reg. 21253, 21258 (1979). A defector who is already in the United States will probably submit his application to the district direc­ tor. Id. In that case, regulations provide that: The applicant shall appear in person before an immigra­ tion officer prior to adjudication of the application. . . . 9 C urrent regulations require that dep o rtatio n proceedings be form ally initiated before an alien is detained under § 242(a). 8 C .F .R . § 242.2(a). 10 T he G eneral C ounsel's office o f the Im m igration and N aturalization Service informs us that adm inistrative interpretations o f § 242(a) essentially follow the Suprem e C ourt's. ** See note 12 infra. 355  The district director shall request the views of the De­ partment of State before making his decision unless in his opinion the application is clearly meritorious or clearly lacking in substance. The district director may approve or deny the application in the exercise of discretion. 8 C.F.R. § 108.2. An exclusion hearing is potentially an elaborate affair, see 8 C.F.R. § 236.2, and creates opportunity for delay. The district director, and the Department of State where it plays a role, might in the normal course also contribute to delay. Nothing in the Immigration and Nationality Act prohibits an immigration judge or district director, in managing his docket, from giving priority to other cases over one which both parties are willing to delay. If the defector consents,12 then, and if he is otherwise lawfully in custody, the Immigration and Naturalization Service might delay action on his application for asylum and keep him in custody until any danger to him subsides and until, in due course, his request for asylum is granted. This approach appears to authorize protective custody for some of the defectors your memoran­ dum describes. 5. The Federal Law Enforcement Role if a Defector Is Assaulted or Threatened The Department of Justice has authority to protect defectors of the kind you describe only in the circumstances we have discussed. You should be aware, however, that once a defector is assaulted, harassed, specifically threatened, or abducted, federal law enforcement agencies may be authorized to play a role. Specifically, we believe, for reasons stated below, that the Secretary of State may designate a defector an “official guest” and in that way give federal law enforcement agencies clear jurisdiction over any assaults, harassment, threats, and similar offenses against the defector, without regard to the interstate character of the offense or to any of the other usual bases for federal law enforcement jurisdiction. This conclusion may be important to you in dealing with defections in the future. As we noted earlier, several federal statutes make it a crime to injure 12 If a d efecto r does not consent, he will be able to invoke portion o f § 242(a) itself to gain relief: A ny c o u rt o f com petent jurisdiction shall have authority to review or revise any d eterm ination o f the A tto rn ey G eneral concerning detention, release on bond, or parole pending final decision o f d eportabilily upon a conclusive show ing in habeas' c orpus proceedings that the A tto rn ey G eneral is not proceeding w ith such reasonable dispatch as may be w arran ted by the particular facts and circum stances in the case of any alien to determ ine deportability. 8 U.S.C. § 1252(a). He may also be able to raise serious constitutional questions about his continued detention. See Stack v. Boyle. 342 U.S. I (1952); compare Carlson v. London. 342 U.S. 524 (1951). with Barcnhlalt v. United Stales. 360 U.S. 109. 128 (1959). 1 356 “official guests” of the United States in these ways. For example, 18 U.S.C § 112 provides: (a) Whoever assaults, strikes, wounds, imprisons, or offers violence to a foreign official, official guest, or internationally protected person or makes any other violent attack upon the person or liberty of such person, or . . . makes a violent attack upon his official premises, private accommodation, or means of transport or attempts to commit any of the foregoing shall be fined not more than $5,000 or imprisoned not more than three years, or both. (b) Whoever willfully— (1) intimidates, coerces, threatens, or harasses a foreign official or an official guest or . . . (2) attempts to intimidate, coerce, threaten, or harass a foreign official or an official guest . . . * * * * * shall be fined not more than $500 or imprisoned not more than six months, or both. Other statutes make it a federal offense unlawfully to kill or attempt to kill an official guest, id. § 1116(a), to kidnap an official guest, id. § 1201(a)(4), or to threaten to assault, kidnap, or kill an official guest, whether or not in connection with an extortionate demand, id. § 878 (a), (b). For purposes of applying these statutes, an official guest is defined as “a citizen or national of a foreign country present in the United States as an official guest of the Government of the United States pursuant to designation as such by the Secretary of State.” 18 U.S.C. § 1116(b)(6).13 We believe that the Secretary of State can designate a defector as an official guest solely in order to bring him within the coverage of these criminal statutes, thus enabling federal law enforce­ ment agencies14 to act against anyone who assaults, threatens, harasses, coerces, kidnaps, or otherwise similarly injures a defector. As we have noted, Congress created the category of official guests because it wanted federal criminal law to “operate to protect the rights of visiting artists, academic and scientific groups,” and similar groups and individuals. S. Rep. No. 1105, 92d Cong., 2d Sess. 9 (1972). Certain aspects of the legislative history suggest that Congress did not intend to permit foreign visitors to be classified as official guests simply because they were threatened. For example, in suggesting to Congress the 13 T he Secretary o f State has delegated his authority to designate official guests to the D eputy U nder Secretary o f State for M anagement. 22 C .F .R . §2.4. 14 As we have said the FBI has general authority “ to detect and prosecute crim es against the United States" and to conduct certain o th er investigations. 28 U.S.C. § 553 (1), (3). D epartm ent of State security officers are specifically authorized, “ w hile engaged in the perform ance o f the duties prescribed" by statute, see pp. 2-4 supra, “ to arrest w ithout w arrant and deliver into custody any person violating section . . . 112 o f title 18 in their presence o r if they have reasonable grounds to believe that the person to be arrested has com m itted o r is com m itting such a violation.'* O ther law enforcem ent agencies have some authority to arrest persons they reasonably believe t<S have com m it­ ted felonies. See. e.g.. 18 U.S.C. § 3056(a) (Secret Service). 357 language that became the definition of “official guest,” the then Secre­ tary of State said, “This will allow me to designate individuals or groups of individuals who are here for important international sports or other events. . . . This would accord protection to foreign nationals who visit the United States for such special reasons as to compete in international sports events.” Id. at 15-16. In general, Congress focused on threats to visitors which were, at least in part, the result of the visitors’ special role in activities of interest to both their country and ours. Congress was also concerned with the implicit obligation we have to their respective countries to protect such visitors. If these were the bases of Congress’ decision to make it a crime to assault or threaten “official guests,” that category cannot be extended to reach ordinary visitors who are threatened only because they have defected.15 The legislative history, however, contains no clear references to reciprocity, or to the fear that Americans will be inadequately pro­ tected abroad; this suggests that Congress may have been concerned less with international obligations than with our international reputa­ tion. That reputation would be injured if a defector were attacked or threatened by the nation from which he defected. Moreover, while the legislative history does not refer to the danger that defectors might be forcibly repatriated, Congress clearly had in mind politically motivated threats and acts against foreign visitors; the killing of Israeli athletes at Munich in 1972 was repeatedly cited as an example of the sort of crime which would have to be left entirely to the states if federal criminal laws were not extended to official guests. See, e.g., id. at 9, 15. And nothing in the statutes or their legislative history makes an exception for politically motivated violence or coercion by the nation of which the guest is a citizen. Finally, Congress carefully considered the issues of federalism in­ volved in creating a category of “official guests” and allowing the federal government, in addition to the states, to punish certain crimes against them. For example, the sponsor of the provision including “official guests” in the several federal criminal statutes gave, as his principal reason, “State governments simply cannot cope alone with crimes involving international politics and diplomacy.” Id. at 9. In language we have already quoted, the Senate Committee noted that the protection would extend generally to “groups and individuals who 16 We do not believe that the Secretary of State must designate a visitor an official guest before he enters the country. The statutory definition arguably requires that an official guest be “present in the United States . . . pursuant to designation,” suggesting that a person who is present in the United States on his own initiative cannot qualify as an official guest. This language is not, however, unequivocal; it does not specify that an official guest must have entered the country pursuant to a designation. The phrase “pursuant to designation as such by the Secretary of State” may, we believe, be read simply to modify “official guest,” describing how one attains that status. Moreover, the legislative history indicates that the category of "official guest” was created precisely in order to provide a federal role in enforcing laws making it illegal to assault, harass, or kidnap foreign nationals visiting the United States. There seems to be little reason to insist that the Secretary must foresee, before the visitors enter the country, that they will be threatened. 358 ought not be beyond the pale of Federal concern.” Id. This emphasis on federalism suggests that the defining characteristic of official guests is their importance to foreign policy and related concerns of the federal government; &ie treatment of defectors is at least as important to foreign policy as the treatment of visiting artists and athletes. In addi­ tion, if there is a possibility that a defector will be harassed or coerced by the nation from which he has defected, the federal government is likely to be involved in negotiations and diplomatic maneuvers which must be coordinated with law enforcement efforts undertaken on the defector’s behalf. For these reasons, we believe that the Secretary of State can designate any defector an official guest in order to make it a federal offense to assault, harass, intimidate, coerce, imprison, threaten, kidnap, or kill the defector. L a r r y L . S im m s Deputy Assistant Attorney General Office o f Legal Counsel 359 

Gary Neiwand Gary Neiwand (born 4 September 1966) is an Australian retired track cyclist. He is a former world champion, who also won four Olympic medals during his career. Biography Early life and career Neiwand was born in Melbourne, Victoria, on 4 September 1966, the son of Ronald Charles Neiwand. Neiwand represented his country at cycling for more than a decade. He was Commonwealth Games sprint champion in 1986 and won his first Olympic medal at the 1988 Summer Olympics in Seoul, winning bronze in the sprint event. He repeated his success at the Commonwealth Games in 1990, taking the gold in the sprint. Neiwand won his first World Championship medal in 1991 in Stuttgart, winning the bronze in the sprint. He then went on to become the World Champion in both the keirin and sprint in 1993. 1994 saw Neiwand complete a hat-trick, winning his successive third sprint gold medal at the Commonwealth Games. After cycling Neiwand spiralled into depression after missing his goal of a gold medal in the Keirin at the 2000 Summer Olympics. He began drinking heavily, his marriage fell apart and he gained weight, ballooning to 116 kg. In 2006 Neiwand was sentenced to eighteen months in prison for breaching a court order preventing him from harassing a former girlfriend. Neiwand began his sentence at Port Phillip before being moved to Beechworth. He was released on probation after serving nine months in jail. He has since rebuilt his relationship with his children, Malcolm and Courtney, and ex-wife, Cathy. Neiwand joined the Sunrise Foundation in 2007, an organisation which, through the high schools, aims to demystify and tackle depression in the community. It is run by former North Melbourne and Sydney footballer, Wayne Schwass. On 5 December 2011 Niewand pleaded guilty to two charges of wilful and obscene exposure in Melbourne, after having been charged on two separate occasions in January and May 2011 for exposing himself to women while masturbating in his car. These offences occurred after Niewand had been released on a court order in December 2010 on unrelated charges, where he had been ordered to attend Forensicare and counselling about his behaviour. The court heard that having attended counselling over the previous six months Niewand had made significant progress, had his drinking problems under control, and was concerned and upset about his offending and his predicament. The magistrate regarded a report from the Forensicare counsellor as "promising", adjourning the case until March 2012 when he was sentenced to four months jail, wholly suspended for two years. Awards and honours Neiwand received the Australian Sports Medal on 14 July 2000 and the Centenary Medal on 1 January 2001. Major results 1986 1st Sprint, Commonwealth Games 3rd Sprint, Goodwill Games, Moscow 1988 3rd Sprint Summer Olympics 1990 1st Sprint, Commonwealth Games 1991 3rd Sprint, UCI Track Cycling World Championships 1992 2nd Sprint, Summer Olympics 1st Sprint, Copenhagen 1993 2nd Sprint, GP de Paris 1st Keirin, UCI Track Cycling World Championships 1st Sprint, UCI Track Cycling World Championships 1994 1st Sprint, Commonwealth Games 1996 2nd Sprint, Busto Garolfo 2nd Keirin, UCI Track Cycling World Championships 1st Team Sprint, UCI Track Cycling World Championships (with Darryn William Hill & Shane Kelly) 1999 2nd Team Sprint, Frisco 2000 3rd Sprint, Summer Olympics 2nd Keirin, Summer Olympics References External links Category:1966 births Category:Living people Category:Australian Institute of Sport cyclists Category:Australian male cyclists Category:Cyclists at the 1988 Summer Olympics Category:Cyclists at the 1992 Summer Olympics Category:Cyclists at the 1996 Summer Olympics Category:Cyclists at the 2000 Summer Olympics Category:Olympic cyclists of Australia Category:Olympic bronze medalists for Australia Category:Olympic silver medalists for Australia Category:Cyclists at the 1986 Commonwealth Games Category:Cyclists at the 1990 Commonwealth Games Category:Cyclists at the 1994 Commonwealth Games Category:Sportspeople from Melbourne Category:Olympic medalists in cycling Category:UCI Track Cycling World Champions (men) Category:Medalists at the 2000 Summer Olympics Category:Medalists at the 1992 Summer Olympics Category:Medalists at the 1988 Summer Olympics Category:Commonwealth Games gold medallists for Australia Category:Commonwealth Games medallists in cycling Category:Australian track cyclists

Tuesday, 9 October 2012 Tank! SWMBO was off having a girly night out with her sisters this last weekend, leaving me all to my own devices. What was a lonely boy to do? Play with toys of course! I decided to catch up with the exploits of the US infantry platoon from 'Buying the Farm' (blimey, was that over a year ago? I need to game more often!). Both sides rolled up the same mission, which meant it defaulted to an 'escalating conflict' mission for both sides, effectively a straight meeting engagement. I particularly wanted to try out the revised vehicle rules from the beta version of FAD 5 (which can be found through the yahoo group, here), so both forces had some armour to play with. The scenario was that Lt Robinson and the boys had given the job of scouting ahead to find the enemy, mounted up in half tracks and jeeps, with an M8 Greyhound and Sherman in support. Unfortunately they are about to run into a small German group heading the other way with intention of setting up an armour ambush. Kar 98s have the 'bolt action' rule (only roll 1D6 for fire effect) and the Stugs are 'slow firing' (cannot move and fire) The forces aren't exactly balanced, but to be honest I always like the 'story' side of gaming, so I go with what seems an appropriate force for an interesting game. Aerial Recon of the battlefield The US forces approached form the 'south', the Germans the 'north'. The US approached in two columns, the M8, HQ, 2nd and 3rd squad advancing up the road to the hamlet; with the Sherman and 1st squad advancing up past the farm building towards the centre. The Germans sent the armoured car and Stugs up the centre, with the infantry skirting the farm on their right flank. The US enter the battlefield German armoured column advances The US right flank advanced rapidly up to the hamlet, depositing the platoon HQ and bazooka team in one of the ruins where they could cover the cross roads with fire. Meanwhile the Sherman nosed into a hedgerow for cover and, as its escorting infantry de-bussed and took to the hedgerows for cover, took a pot-shot at the advancing Stugs, but with more enthusiasm than accuracy! Fortunately the return fire was little better. The two armoured cars both raced to contest the crossroads, with the M8 getting the first shot off, writing off the '222, only to get popped by one of the Stugs, which had veered off into cover in the farm, while the other engaged in a ill-fated gunnery duel with the Sherman (who's crew had clearly calmed down and worked out what they were doing) and the halftracks advanced down the right flank. Assault gun hide n seek Aware that their best chance of success was to take out the Sherman, giving them superiority in armour, the remaining Stug risked an advance to the crossroads, placing trust in luck and armour, while the PAK38 equipped half-track advanced to the farm building where it could set up for a flank shot. Sadly for them, the Sherman crew were now on a roll, brewing up the unfortunate assault gun. Worse was to come, as the bazooka crew found themselves with a possible shot at the half track and made the most of it, convincingly decimating it and its passengers, despite it being hull down. Being the only ones left, the remaining German infantry, which had started advancing through the woods on the US flank, decided that discretion was the better part of valour against a combined arms force in good cover and fell back, ending the game. Forwards! Or maybe not... Burnt out vehicles at the crossroads US Infantry 'encourage' the German withdrawl The final state of play (in colour, no less!) All in all, a very interesting game. With armour on the table, the infantry pretty much put their heads down until one side had superiority, perhaps not an unrealistic result! The only infantry unit to open fire was the bazooka, with spectacular results. Armoured combat in FAD5 is incredibly simple and pretty deadly between matched opponents. Roll to hit, roll to penetrate and, if you do, it's likely a kill. No record keeping of damage points or endlessly shooting different bits off the same tank here! For me, this seems entirely appropriate for armour clashing in a game that's pretty much occuring in small arms range. The Stugs really struggled with a having to act offensively in this game because of the 'slow firing' rule but, again, I think this felt right and I think I'll keep it. It'll be interesting to see how they fare in a more defensive role. I do think that some using opportunity fire will be needed for these and AT guns to keep them from being too vulnerable to more conventional tanks. I did find that the vehicles fair zipped around the table. Was it too much or was this appropriate? Not sure, Jury's out on that one for now (I could always swap units from inches to cm if needs be). The temptation to nip about is probably what led to both armoured cars getting themselves exposed and in so much trouble. It certainly proves why recon units were not supposed to go engaging 'true' armour!

Significant down-regulation of the plasminogen activator inhibitor 1 mRNA in pancreatic cancer. We examined mRNA expression of the urokinase-type plasminogen activator (uPA), its receptor (uPAR), and the plasminogen activator inhibitor 1 (PAI-1) in a panel of adenocarcinomas of the pancreas (PC) and cancers of the papilla of Vater (CPV). Expression profiles were compared with paired uninvolved normal tissues to define a possible differential role of these genes in tumorigenesis of both tumor types. Urokinase-type plasminogen activator, uPAR, and PAI-1 mRNA expression was analyzed by real-time quantitative reverse-transcriptase polymerase chain reaction (TaqMan) in 25 PC, 7 CPV, and in the paired uninvolved normal tissues. Uninvolved normal tissue probes from PC and CPV showed similar mRNA expression profiles of uPA, uPAR, and PAI-1. Whereas expression levels of uPA (P = 0.81) and uPAR (P = 0.75) were not statistically significant different between tumor and paired normal tissues, PAI-1 levels were significantly down-regulated in tumor compared with paired normal tissue samples (Wilcoxon test; P < 0.006). No differences in mRNA expression of uPA, uPAR, and PAI-1 between PC and CPV were observed. Expression levels of the 3 genes were not associated with tumor stage, grading, or survival. Increased mRNA expression of uPA and uPAR could not be detected in PC and CPV; however, PAI-1 mRNA expression levels are significantly down-regulated in PC, which might lead to higher activity levels of uPA components. Our data are merely hypothesis generating and should be validated in larger translational studies.

Pae Gil-su Pae Gil-su (, also written Pae Kil-su, born March 4, 1972) is a North Korean gymnast. He won the gold medal for the pommel horse at the 1992 Summer Olympics (tied with Vitaly Scherbo). And he won the gold medal at the 27th, 28th and 32nd World Gymnastics Championships. Pae attended Pyongyang Sinri Primary School and the Korean Physical Education College. External links Competitive results at gymn-forum.net Category:1972 births Category:Living people Category:North Korean male artistic gymnasts Category:Olympic gymnasts of North Korea Category:Gymnasts at the 1992 Summer Olympics Category:Gymnasts at the 1996 Summer Olympics Category:Gymnasts at the 2000 Summer Olympics Category:Olympic gold medalists for North Korea Category:World champion gymnasts Category:Medalists at the World Artistic Gymnastics Championships Category:Olympic medalists in gymnastics Category:Medalists at the 1992 Summer Olympics Category:Asian Games medalists in gymnastics Category:Gymnasts at the 1990 Asian Games Category:Gymnasts at the 1998 Asian Games Category:Sportspeople from Pyongyang Category:People's Athletes Category:Asian Games gold medalists for North Korea Category:Asian Games bronze medalists for North Korea Category:Medalists at the 1990 Asian Games Category:Medalists at the 1998 Asian Games

Q: Spring bean reference not working I have the following bean: package com.test; @Component public class Sample{ String modified = null; @Value("${url}") private String url; public Sample(){ System.out.println(url ); if(baseUrl.equals(""){ throw new RuntimeException("missing"); } else{ modified = "test"+url; } } } I have added: <context:annotation-config /> <context:property-placeholder location="classpath:test.properties"/> & <context:component-scan base-package="com.test"/> and trying to access above "modified" field as below <bean id="url" class="java.lang.String"> <constructor-arg value="#{sample.modified}" /> </bean> in my application context. But I keep getting the following error: Field or property 'sample' cannot be found on object of type 'org.springframework.beans.factory.config.BeanExpressionContext' Not sure why i get this error? A: When Spring creates the object it uses the default constructor. It can't set the property until after it constructs it. Instead of what you have, try this to see if the value is being set. @PostConstruct public void init(){ System.out.println(url ); if(baseUrl.equals(""){ throw new RuntimeException("missing"); } }

A Sustainable Business Network Model for Southern Africa The relevance of the Southern African Development Community (SADC) regional integration process emanates in part from the view that there is a trade-off between regional integration and integration with the global economic system. Whilst integration with the global economy could provide the impetus for economic growth, regional integration could provide the required protection from the ills of globalization. This argument is based upon the belief system that regionalism could be used as an instrument to create a new equilibrium that balances the protection of the vulnerable and the interests of the particular population living in that environment against the integrative, technological dynamics associated with globalization. In this regard the SADC can typically be classified as a vulnerable region as it is dealing with a marriage of third and first world realities characterized by the existence of cultural colonialism accompanied with perceptions of higher and lower cultures, severely distorted levels of inequality, destructive development policies, extreme poverty and demands for immediate relief for the sake of survival. The Problem Statement If Southern Africa intends to succeed, a new look is required of the region’s social capital capabilities using fresh and creative business strategies, which will build on mutual foundations of trust. This implies amongst others a need to develop strong partnerships between all the stakeholder groups that can be unified into a strong wish power supported by a congruent business approach. The Proposed Business Network Solution The new business development strategy in Southern Africa should emphasize the endogenous business sector, of which the informal small business sector is the most important factor. It is estimated that the majority of Southern Africa’s working population operates in the informal economic sector without legislation, insurance, employee protection or rules. It is therefore important to integrate the informal small businesses into the formal economic sector as it may serve as the springboard to effective enterprise development in the SADC. The second step is to ensure that all small businesses are well embedded in open business network relationships in order to enable them to search and obtain resources that they do not possess. By doing this an opportunity-driven approach is adopted that may lead to competitive advantages for businesses for it creates the opportunity to take advantage of the knowledge and problem-solving abilities of all actors belonging to their supply and demand networks directly or indirectly. However, to ensure sustainable business development an epistemic small business community is regarded as a pre-requisite. It requires that the small business community network should share a set of normative and principle-belief system, which provide a value-based rationale for the business and social actions of all the actors. Business framework conditions should also be crafted that could provide a raison d’etre and an agreement of the common purpose that the small business environment pursues. The outcome of the network implementation process is that every business in essence becomes a networked organization or an extended enterprise positioned to compete in the global market. It is envisaged that the implementation of the network approach should be accompanied by some social interventions, as small business owners in the SADC do not operate in a normal market economy. One option for example is to enter into sub-contracting relationships with large companies to produce sub-components of final products. Jan Grundling is the Head of the Centre for Entrepreneurship at Tshwane University of Technology, South Africa. He has published extensively in various national and international Journals. He has also contributed more than 45 papers at national and international conferences. Young Entrepreneurship Are you a young Entrepreneur? If so, we encourage you to check out our young Entrepreneurship resource center, full of the articles and information you need to help you generate ideas and start your own business as a young Entrepreneur. Or find more information at the Kauffman Foundation for Entrepreneurship or the Entrepreneurship Resources at Entrepreneurship.org

Towel Drills Back in 2009! Love the Cubs? Have an opinion? Want to express it in a forum read by dozens of Cubs fans on a bi-annual basis? Then we're looking for you. Hate getting paid to do stuff? Live in your parent's basement? Think you know more than everyone else, so much so that you create your own forum to prove it? Then we're also looking for you. Towel Drills started last year as a way for a few friends to share their Cubs rants with even more friends. Somewhere along the line, you, Cubs Fan, started paying attention. Before long, Towel Drills was bigger than Jebus. And now we're asking for your help to continue to grow TD into the Blogawesomeosphere. TD is looking to take on a few additional writers for the upcoming season. We need people who can commit to posting on a regular basis and who have access to post throughout the day. We're specifically seeking at least one "news man" who can post breaking Cubs news the segundo it drops. Freedom to also write about what interests you in relation to the Cubs we can offer. Any other form of payment, monetary, seshual, or other, we cannot (but only b/c the seshual thing is illegal). If you're interested in trying your hand at blogging, hit us up at toweldrills@gmail.com. How you "apply" is up to you. Send us some writing, some ideas, an 8x10 glossy of Ronnie Woo Woo (please, don't), send bags of cash (please, do)-your call. We're looking forward to adding some talent to the staff and to continue making TD the best gosh darn TD it can be! RECENT POSTS C.C. ya later. Looks like the deal that has been rumored since Sunday is in fact going to go down. C.C. Sabathia has been traded to the Milwaukee Brewers for stud 1B/OF prospect Matt LaPorta along with minor league pitchers Zach Johnson and Rob Bryson and possibly 3B Taylor Green. Looks like it's time for the Cubs to seriously start looking in the rearview mirror at the Brewers. The Brewers started off rough and have had some key injuries to Yvonni Gallardo and, for a short time, Ben Sheets. On May 19th the Brewers were 20-24 and in last place in the NL Central. Since then, the Brew Crew is 29-15 and have climbed within 3.5 games of the 1st place Cubs, tied for 2nd in the division with the Cardinals. The Brewers will get 2 starts out of Sabathia before the All-Star game and he should be a key cog in their run over the entirety of the 2nd half. This race could well come down to the Cubs/Brewers again this year, with their roles reversed from 2007. Long-term, I'm happy in a way that the Brewers did make this move. LaPorta is, by all indications, a stud in the making. And the other players traded seem at the very least to have the potential to turn into good major leaguers. Trading that amount of talent for a Sabathia rental (no way the Brewers pop to re-sign him next year), along with the fact that Ben Sheets is likely going to sign a free agent deal elsewhere for next year as well, and it's clear that the Brewers are going for it hard this year. The price they'll pay is in the next few years...good for the Cubs. Kudos to the Brewers for landing the player who is far and away likely the best player available this year. Sabathia instantly makes the Brewers a much scarier team-and the thought of Sabathia pitching in the "JV League" might give me nightmares tonight. The question that now remains: how will Jim Hendry and the Cubs counter? Comment (1) MAV July 7, 2008 at 7:51 AM That's some serious brass the Brew-Crew traded away here. The Cubbies don't even have that equivalent in their farm system.. That's a scary proposition knowing we'll be seeing plenty of C.C. and Sheets down the stretch. And of course Fielder is the new cubbie killer. Looking forward to a battle royale! Let's get Alfie back on the field, and let's get it on!

Start Here My Readers Just when you thought that you were cruising down easy street, this thing called life had different plans for you. Sometimes the plans were unannounced, and at other times, you knew they were coming. When this life moment arrived, were you knocked off of your feet, questioning how to stand back up, or did you hold your ground, overcoming the obstacles and challenges that lie ahead? In the past, have you been challenged with with adversity? Are adverse conditions looming on your horizon? Or, are you like most of us and have adversity you are dealing with today? If you answered yes to these questions, then the follow up question is: Do you want to create opportunities for yourself that you have yet to uncover? If you’ve answered yes to this questions, then this is a place for you. What You’ll Find Here Through my articles, I have shared ideas with thousands of readers about understanding and implementing strategies to work through some of life’s most difficult situations. The people who read my content have lost a jobs, struggle financially, and deal with addiction(s). My readers suffer from emotional, mental and physical illnesses. And some people, they just seem to be down on their luck. However, there are methods for overcoming the mountain in front of you and getting to the other side. In my writing share a mixture of practical insights, easy to implement strategies, and honest opinions that can help move you from where you stand today, to where you want to be tomorrow. Additionally, I share real life examples of everyday people – JUST like you and me – that have defied all odds and have beaten or are beating adversity. What’s Next? If your life is about to change, is in the middle of changing, or has already changed, then I would suggest checking out a few of the articles I have written in the past. Interested in learning how to manage complex change – start here. Are you searching for work opportunities that align with your passion and purpose – check this out. Would you like to be better than just ‘mediocre’ – you’ll love this one. After you’ve done this and are ready for the next step, then I suggest subscribing to my weekly articles. I promise I won’t clutter the inbox – one article, once a week. It could be the catalyst that gets you moving forward to living life just how you imagined.

Intestinal Flora Modification of Arthritis Pattern in Spondyloarthropathy. The reactive form of spondyloarthropathy appears inducible by exposure to agents of infectious diarrhea, but do those organisms represent the tip of the iceberg, as indicated by renewed interest in gastrointestinal flora? Prevalence of spondyloarthropathy (20% of chimpanzees [Pan] and 28% of gorillas) is independent of subspecies and species, respectively. However, there are major differences in arthritis patterns, a characteristic shared with humans. Do patterns of arthritis correlate with gastrointestinal flora? Could such associated modifications be in the form of disease induction or represent protective effectors (at least against the extent of peripheral arthritis)? The skeletons of 2 chimpanzee subspecies (79 Pan troglodytes troglodytes and 26 Pan troglodytes schweinfurthii) and 2 gorilla species (99 Gorilla gorilla and 38 Gorilla beringei) adults were examined, and arthritis pattern noted. Feces of Eastern (P. schweinfurthii and G. beringei) and Western (great apes collected in their normal ranges) apes were assessed for 16S rRNA c and its character. Patterns of arthritis recognized on examination of skeletons showed geographic variation in skeletal distribution. East African apes (P. troglodytes schweinfurthii and G. beringei) had pauciarticular arthritis and frequent sacroiliac disease, whereas West African apes (P. troglodytes troglodytes and G. gorilla) had polyarticular peripheral joint disease with minimal sacroiliac involvement. DNA evidence revealed that Corynebactericeae were prominently represented in great apes with polyarticular disease, whereas Dietzia and Bifidobacterium exposure correlated with reduced peripheral joint arthritis distribution. Suggestions of a protective effect (in this case, limiting extent of peripheral arthritis, but not the disease itself) offered by these organisms are well represented by documented effects in other diseases (eg, tuberculosis) in the zoologic record. Perhaps it is this disease-modifying character that reduces the extent of the peripheral erosive disease, while increasing propensity to axial (sacroiliac) disease. A potential role for probiotic organisms in management of arthritis in humans is suggested, as has been documented for tuberculosis, gastrointestinal disorders, and food allergies.

Details Description TSHttpTxnErrorBodySet() sets HttpSM::t_state.internal_msg_buffer without freeing any old contents in there. There can be an error message in that if you have a request with a bad hostname and you let the transaction get past DNS lookup. Instead it should free the contents, or there should be another field that it sets and nothing else does. Correct me if I'm wrong but with this patch it's still leaking mimetype. See HttpSM::setup_internal_transfer line 5416 which frees internal_msg_buffer_type...it's expected that mimetype was malloced since clearly it's being freed. So that means there is still a memory leak in TSHttpTxnErrorBodySet(). Brian Geffon added a comment - 07/Dec/11 21:40 - edited Correct me if I'm wrong but with this patch it's still leaking mimetype. See HttpSM::setup_internal_transfer line 5416 which frees internal_msg_buffer_type...it's expected that mimetype was malloced since clearly it's being freed. So that means there is still a memory leak in TSHttpTxnErrorBodySet().

Q: Reactivex Observable blocking UI thread So Im using a MVVM pattern with Rx, the thing is when I run the operation that is suppose to run in a background thread my UI thread gets blocked, this is my viewmodel: class dataViewModel (application: Application): AndroidViewModel(application) { val dataRepository = DataRepository.getInstance(application.applicationContext) val listData = MutableLiveData<List<Data>>() var isLoading = MutableLiveData<Boolean>() fun loadData(){ isLoading.value = true dataRepository.getData().subscribeOn(Schedulers.newThread()) .observeOn(AndroidSchedulers.mainThread()) .subscribeWith(object: DisposableObserver<List<Data>>(){ override fun onComplete() { //Update UI isLoading.value = false } override fun onNext(retrievedData: List<data>) { listData.value = retrievedData } override fun onError(e: Throwable) { //Handle error } }) } } And this is the repository method: fun getData(): Observable<List<Data>> { var dataList = query.search //this query may take a while and blocks the UI thread return Observable.just(dataList) } I'm I missing something? A: Your query.search happens in the UI thread, becuase you're calling it before your Rx operators will be involved. You should change getData method to fun getData(): Observable<List<Data>> { return Observable.fromCallable { query.search() } } In this case the query.search will be called in whatever thread is defined in this Rx chain (using subscribeOn).

$1000 FAQ Below, is a list of frequently asked questions. We tried to gather a list of some of the most common questions, and hopefully these help you to answer your questions. Is this 100% open-source code script? Our application is 101% open-source and ready to customize from back-end to front-end You are fully allowed to extend, customize or redesign only for your business only. How long does it take to setup after I purchase? Once an order has been submitted and paid, you will receive an e-mail to download the application. Usually, it takes 5 to 6 hours during business hours. Are there any hidden charges? No, there are not any hidden charges. You will receive fully functional booking application. Can I request custom changes? Yes! you can. We do provide custom web development at very affordable price. Please arrange your requirements document and open ticket from customization's area. Which hosting should I use for this? The application is smart and it supports all kind of hosting. You can even run this booking engine with minimum server specs and no additional things are required. How can I get support and advice? Get in touch with our support team through the Contact us page or send us a ticket through our ticket support system using your client area. Is that possible to use the application on more than one domain? Yes it will cost you $100 for each additional domain license. Please be advised this is only possible if you want to use for your own business Do you offer refund? Sorry, our application is open-source and it can be copied after sale. Keep in mind that we are not able to offer refund. Please take full overview on our demo website before proceed to order. So, we can avoid any dispute in future. Updates are free for lifetime to version however you need to buy only if there is entire new version. for example when you purchase the code and the current version is v7 you can keep using it for lifetime with it's fixes updates builts and support until new release but when there is a new upgrade for example like v7.1 and still you wanted to use this latest version v7.1 then you have to purchase the code once again.

[Sacroiliac joint dysfunction with groin pain after an operation for lumbar spinal disorder. A case report]. A 75-year-old male presented with groin pain after an operation to treat lumbar spondylolisthesis (L5). Groin tenderness was localized to the medial border of the anterior superior iliac spine (ASIS). Radiographical and physical examination raised the suspicion of sacroiliac joint (SIJ) dysfunction. Injection of a painkiller into the SIJ relieved symptoms, including groin tenderness. Symptoms improved gradually, and finally disappeared after five SIJ injections. Groin pain has been reported as a referred symptom of SIJ dysfunction in 9.3-23% of patients. Prior to the patient undergoing surgery to treat lumbar spondylolisthesis, SIJ dysfunction had not been noted on physical examination. Long periods spent in the abnormal posture due to lumbar spondylolisthesis induced SIJ stress. After the operation, an improvement in daily activity actually increased stress on the SIJ, resulting in SIJ dysfunction. Certain pathologies, including SIJ dysfunction, should be considered as residual symptoms after operations for lumbar spinal diseases.

Gas prices, firefighters and volunteers. It's a recipe for legislative success, right? Not when you are in the dog days of summer in a general election year. Nonetheless, Sen. Bob Casey and Rep. Jason Altmire, both Democrats, hailed a bill they introduced Wednesday that touches on all three. The two lawmakers want to lessen the pain of high gas prices on volunteer firefighters by offering federal grants to help them recoup the costs. Casey laid it on in particularly dramatic terms to get his point across at a news conference today, where he was flanked by six volunteer fire fighters from Pennsylvania. "In the dark of night and the middle of winter ... we want to make sure that that fire truck has enough gasoline to get to that fire," Casey said. "This is serious business for the people of Pennsylvania and the people across America." One firefighter, Bill Carlisle of Fairview Township in York County, said his department's fuel costs have more than doubled to $25,000 this year, meaning that many more bake sales are needed to raise cash. Casey said he has discussed the bill -- which he estimated would cost $50 million annually -- with Senate leadership and he is hoping to include the measure in any energy package that comes up this year.

werrrd Month: May 2015 One of the many reasons people self-harm: to feel like they have control in this one aspect of their life. Chuck Palahnuik took this idea and wrote a novel. It’s called Invisible Monsters. If you want to read this book on your own, you should do that.. I will now continue with spoilers. The novel’s main character is Shannon McFarland. We learn about her life as it is currently and as it was previously from her point of view. The story is told in a nonlinear manner, where various events are included when Shannon believes it’s necessary… sort of a flow of conscious kind of knit… but sometimes, things aren’t obviously tied together. We know very little at first. We know Shannon’s jaw has been shot off her face thus making it impossible for her to talk except in gurgling sounds. We quickly learn Shannon was a model, and, obviously, can no longer be a model. We learn Shannon does not really care much for her family.. there’s some strain since her brother Shane died. Shannon has a few important people in her life she is constantly talking about. There’s Evie, her best friend prior to her jaw-accident, Manus, her boyfriend prior to the jaw-accident, and Brandy, her best friend post accident. Every one of these people are the most superficial people you could ever meet. I mean, Manus sleeps with men just so he can feel like he’s still got it… Brandy is all about keeping up appearances… and we don’t know much about Evie except she’s a model and encourages Shannon to go for hand modeling and to reconstruct her jaw so that she can continue modeling… because all that matters is your appearance. Anyway… the opening scene is Shannon standing next to a dying Brandy and Evie standing with a smoking gun. The entire book is all the pieces put together so the reader and learn about what lead to this moment. In as few words as possible: Brandy is Shane, after 90% of MTF surgery. Shane pretended to have died and changed his body because he doesn’t want to be a girl and this is the worst thing he can do to himself… self-mutilation. Manus is a detective who sexually assaulted Shane and caused Shane to feel like he deserved less than good. Evie slept with Manus while Shannon laid in the hospital bed. And… Shannon shot herself in the jaw. Yes. This crazy woman shot herself in the jaw because she didn’t like her life as a pretty model anymore. She wanted to experience life as an ugly person. The grass is always greener. At the end, when she’s standing there watching her brother die because of her… her brother who actually suffered. Nooooow Shannon’s sorry. She’s sorry she mutilated herself for no reason except she was bored. In my opinion, Shannon is the most self-centered of them all. WHO THE EFF WOULD DO THAT BECAUSE THEY ARE THAT BORED? For realz. You can’t move to another country and teach english? You can’t just eat a lot and become fat? You can’t travel around the world? Learn to weld? Have a baby? It makes complete sense to not only destroy your life, but destroy the lives of those around you. COMPLETE SENSE. Then, you have the nerve to give Shane your identity and decide that makes up for everything you’ve done.. WHO THE EFF ARE YOU to play God? It’s not okay that you play Evie when all she did was sleep with your boyfriend. What IS that compared to burning down two houses? It’s not okay that you drag Manus around and drug him when all he did was sleep with Evie, assault your brother, and try to kill you…. oh wait… he deserved worse, you’re right. But, what about poor innocent Brandy? The woman who took you in and tried to build you up after your accident. She didn’t deserve for you to hate her. To plan her demise. Shane walked away from the family because you didn’t open the window for him. Brandy took that bullet because you didn’t stop it. You hate Shane because he took your attention. REALLY?! I’m not sure if we are supposed to feel bad for Shannon. I can’t tell if she’s supposed to be the hero. All I know is that I would never forgive her. I wouldn’t want her in my life. There aren’t any heroes in this book. Everyone is a villain. I picked up Invisible Monsters because I’ve read Survivor and I’ve watched Fight Club. Both of these are pretty good in my opinion. I.M. is not horrible, but it seems to me that Palahnuik only knows how to write apathetic-sounding characters who end up in a whirl-wind situation of their own doing. Not that C.P. is a bad writer. Not at all. But… maybe… he’s only good with one type of story. That’s cool. It’s like how Paul Rudd is always casted as the funny guy or how Scarlett Johansson is always casted as the boobier woman. What did you think about Invisible Monsters? Do you interpret it differently? Maybe you can convince me to have a more positive opinion. However, you might be a person who has reached the graduate school level. You probably did everything right and you don’t need my help. Cool. But, you need to know how to properly choose your adviser. This is critical. This is the person you are going to work closely with for the next 3-4 years of your life. This person will shape how you think and what kind of researcher you will be. They are your *insert field of study here* mother/father/parent. You need to do proper investigation: 1) Determine a professor’s research interests. If you don’t care even a little about what someone is doing research in, you’ll hate your time and you will lack motivation and you will suffer. Three years doesn’t seem like a long time, but it drags on and on. Remember in High School when you would watch the clock during the last hour of the day… yeah, the last three years will be like that. One way to get a really good idea of what a professor does: read his/her/their papers. You may not be able to understand everything, but if you think “oh man, this is cool!” or “that is really interesting, I want to know more” or even “this goes over my head, I love the fact this person is a genius”… then you will want to take the next step with this professor. 2) Sit in on your POI. First, it’s best if you’ve had a class with your POI or if you’ve attended a research presentation (which you should begin attending after your first semester) they’ve given. Then, you know how this person teaches or how this person presents. This helps in that you’ll know how they’ll advise/guide you. You can determine whether their style works for you that way. 3) Meet with the professor of interest (POI). However, if you haven’t been so lucky… a meeting with your POI will have to do. Tell them you are interested in research with them. Ask them questions like: what basic background material would I need? or, what classes should I take? or, are there papers/books not given in class that I can read to do what you do? are you open to answering questions I may have? how often will you expect me to meet with you? what are other responsibilities you expect? etc. Problems with this method: people always behave better before you get to know them. It’s later down the line that you will learn how wrong this person is for you. The next item addresses how to avoid this. 3) Talk to students who are studying under or who have studied under your POI. Students who have already graduated will most likely be more honest about their experience. Students currently working with the POI may not be completely honest in fear; plus, it’s not very professional to spread gossip. Despite this, though, you can tell by the answer (or lack thereof) what kind of professor he/she/they are. A professor worthy of praise will be praised. A professor who is hard in a disrespectful way will either have silent students or students who will complain behind closed doors. 4) REGs are an excellent way to judge a POI. Most graduate programs will provide funding for summer research projects with professors. Working with your POI for a summer is a great way to determine whether you will continue to work well together. Note: Everyone has their annoying quirks. You need to determine whether you can look beyond the POI’s treatment of you in order to obtain your degree. Will this professor have you present at conferences and thus provide you networking relationships? Will this professor help you submit several papers? Will this professor be the one to help you pass your defense? These are questions you can determine by asking the POI-students. The compromise is determined by you. Recommendation: None of these things are worth giving up your dignity and self-respect. Even if you don’t have any of the benefits, by walking away, you’ll at least have your self-worth. I speak from my failure to do so.. Note: Everything I say is based on my own college/graduate school experience. To help others from making my mistakes. Youth just don’t know. They are listening to adults. Adults who still think a college education is the end-all-be-all. It’s not true anymore. For first generation students, you’ll go to college thinking: this is it. After college I can be a self-sufficient adult, and I can start my own family. I’ll have a career by time I am 23, so I’ll have children by time I’m 25. It’s later than my own parents, but it’s okay because I’ll still be 45 when my own child goes to college. blah. blah. blah. Graduating college is the biggest accomplishment first generation students will EVER think to strive toward. At least.. that is how it was for me. I didn’t know anything. My mom didn’t know anything. My grandparents, my aunts and uncles, all the people I looked up to in my family didn’t know anything. Teachers and guidance counselors never talk about anything except getting into college. No one ever talks about what happens after the acceptance.. what happens in college.. what happens when you graduate.. students are so uninformed, it’s not even funny. The best thing you can ever be is informed.. Many, many, many students are going to college now. So many. Just think about it. Wikipedia (a source you aren’t allowed to use in ANYTHING, but it’s extremely useful) says there were 21 MILLION students in college in the US as of 2012.. There are probably at least 1 million more.. and that is only in the US. If all these people are getting degrees, what is the point? That IS the point, isn’t it? In order to get into a good college, it’s stressed for students to have good GPAs and to participate in clubs and to join Honor Societies and to volunteer and to diversify and to blah blah blah. I’m in awe of the youth who do all of this AND have to deal with bullsh** in their personal lives. YOU are amazing individuals. It doesn’t end when you get in though. You’re told that if you want a good job you have to keep doing all those things. Except, there are more distractions when you are away from home for the first time.. and so many students suffer because they’ve made the wrong decisions regarding these exploits. It’s more difficult to keep that GPA and participate in those clubs or to volunteer.. but, you can’t get that job without ALL these things. In fact, you have to do more than what you’re told. You need INTERNSHIPS and WORK EXPERIENCE. A college education isn’t enough anymore. So many companies are looking for people who have 3+ years of experience. Even if they are willing to hire a new graduate, they are looking for qualities that come from experience. That means, you need to find out what these skills are and earn them in your 4 years of college. It’s a lot. I know… but you are competing with 21 MILLION students.. and many of those students are adults going back to get their education after the 3+ years of experience. I don’t know about you, but I NEVER get picked for dodge ball when I’m a player and I ALWAYS choose the person who knows what they are doing when I’m the captain. When you don’t have these extra tidbits under your belt, you’ll have a hard time finding a job. That’s when you’ll contemplate graduate school. You’re thinking: I went to college like I was supposed to.. I did everything I was told.. and it still wasn’t enough.. I’m still working this shi**y retail job serving these shi**y people.. I can’t do this the rest of my life.. What can I do? And you think graduate school is the place to be. But it’s not. In fact, it’s worse than getting a Bachelor’s degree. I list reasons you shouldn’t go in another post. You need more. All the extracurricular activities (volunteer work, honor societies, internships, good GPA) that made you look good before still stand.. but now you need to submit/publish papers, present at conferences, teach classes, participate in research groups.. In other words, gaining your professional experience while earning your degree.. all the while pretending you know what you are saying and doing.. all the while feeling as though you are an imposter.. People without a higher education stress that higher education is the way to go because they see everyone around them getting raises and promotions. It’s not true. What you need to succeed is BOTH. You need professional experience in your field of interest. You need a degree to show those mother-effers that you are worthy. You don’t need just one middle finger, you need both middle fingers straight up as a statement that you did more than everything right. You went above and beyond. Please share advice in the comments for the youth deciding upon undergraduate school or graduate school. As a recent Doctor of Philosophy, it’s my job to inform youth (high school seniors and undergraduate seniors) what higher education involves. This is the first of a series of blogs I will post. Five reasons you shouldn’t go to graduate school: 1) If you are doing it because you don’t know what you want to do. Graduate school is a full blown commitment. You are dedicating your LIFE to what you study. You are Michelangelo on your back for weeks painting that Sistine Chapel ceiling. You need to LOVE what you do in order to withstand the lack of blood flow to your fingers. If you don’t love what you do, you will stop caring about everything around you and you will have wasted 5-6 years of your life instead of finding that something you care about. 2) If you are going under the impression a graduate degree will get you a better/higher-paying job. I will post more about this in another blog.. but, this is the WRONG reason to get a higher education. There is absolutely NO guarantee you’ll get a better job or a higher-paying job or any job for that matter. You go to graduate school to learn. Because you love to learn. Because you love to help others learn. WHY THE EFF ELSE WOULD YOU GO?! 3) If you don’t get funding. BIGGEST MISTAKE is to go to graduate school without funding. If 1) or 2) are true… and you don’t get funding. You’re the biggest idiot ever (here’s looking at myself).. because if you don’t love what you do but you are thousands of loans in debt… you’ll feel trapped and will never be able to switch careers. Or, if you can’t find a job to pay those loans back, you’ll be in debt forever. 4) If you don’t know EXACTLY what to study. You’ll have the greatest success in graduate school if you know exactly what kind of research you want to do. You can take the correct classes and find the best adviser for your research. If you go in blindly, you’ll end up taking the wrong classes and doing someone else’s research. You’ll write your thesis on a topic you probably don’t care about, and that’ll be the experience that determines the jobs for which you’ll be eligible… trapping you in an area of research you’ll NEVER CARE ABOUT. 5) If you’re going because all the cool kids are going. We’ve had it drilled into our heads: go to college. you won’t get a job if you don’t go to college. In a sense, this is true (I will write about this more later). However, it’s also not so true because now we’re getting it drilled: go to graduate school, you’ll get a BETTER job if you go to graduate school. Now, millions of kids are getting bachelors degrees and this devalues higher degrees. In addition, half of those kids will try graduate school, making master’s degrees a dime a dozen and PhDs aren’t as rare as they used to be. There is NOTHING wrong with a trade like electrician or mechanic. Not everyone is meant to have a full blown college education. Many people are better with their hands than with book learning. Please don’t go because it’s what everyone else is doing. What you should do: Take time off from school. Find something you want to do. If that something requires you to go back to school, then so be it. If it doesn’t, then you just saved yourself time, money, and pain. Travel if possible. If not, work jobs that humble you, that open your mind, that help other people. Of course… what you should do and what is possible for us to do are two different things, unfortunately.

[Failures of knee joint prostheses. An analysis of knee prostheses and component revisions, 1980-1987]. Total knee arthroplasty has become more common in the last 20 years as a result of the continuous improvement in prosthetic design and operative techniques. Therefore, there have also been more failures. To show the reasons for these failures and the difficulties and results of revision surgery we studied the 48 uni- or tricompartmental revisions of total knee arthroplasties we performed in our institute between 1980 and 1987. Twelve to 94 months (mean 40.4) after revision we found 66% good or very good results. The main reason for failure was a poor surgical technique, which we found in 69% in unicompartmental, and in 28% in tricompartmental prostheses. Revision surgery needs a high level of practical knowledge and flexibility. The main problems consist of correction of the alignment and the filling of bone defects. Thus, we reconstructed ten large defects at the tibial plateau, which resulted in one very good, seven good and two moderate results.

Paul Anderson (UK writer) Paul Anderson (born 1964) is a British writer specialising in computer science and technology. He writes speculative work on the future of technology and was awarded the EPSRC's inaugural Computer Science Writer of the Year prize in 2007. Until 2010 he was the Technical Editor of Jisc's TechWatch series of white papers, which analyse technology trends and predict their potential future impact on the UK's higher education environment. In May 2012, Taylor & Francis/CRC Press published his book Web 2.0 and Beyond: Technologies and Principles, which takes an interdisciplinary approach to understanding how the Web has come to be the way it is and how it might develop in the future. External links Paul Anderson's blog, Techlunch CRC Press: Web 2.0 and Beyond. References Category:1964 births Category:Living people Category:British technology writers

Matt Reid (tennis) Matt Reid (born 17 July 1990) is an Australian professional tennis player. His highest ATP singles ranking is world No. 183, which he reached on 3 February 2014. His career high in doubles was at world No. 60 set on 11 September 2017. Tennis career As a junior, he was runner-up in the Wimbledon Boys Doubles in 2008 partnering Bernard Tomic. He reached the second round of the Australian Open Men's Doubles in 2013 partnering Samuel Groth. In 2013, Matt Reid made his Grand Slam main draw debut at Wimbledon after booking his spot through qualifying. He lost in round one in straight sets. In February 2014, Reid won the Burnie Challenger, which increased his ranking to a career high of 183. ATP career finals Doubles: 1 (1 runner-up) Challenger and Futures finals Singles: 16 (7–9) Doubles: 50 (27–23) Performance timelines Singles Doubles Current through to the 2019 Australian Open. References External links Category:1990 births Category:Living people Category:Australian male tennis players Category:Sportspeople from Sydney Category:Tennis people from New South Wales

Joseph S. Nye, Jr. a former US assistant secretary of defense and chairman of the US National Intelligence Council, is University Professor at Harvard University and a member of the World Economic Forum Global Agenda Council on the Future of Government. He is the author, most recently, of Is the Ame…read more Comments Please dude, be honest for a change. Obama has a proven track record, yeah, proven to be BAD. Sure his supporters are going to do what he does, blame everybody but himself. "Oh, he had to adapt to two ongoing wars", the war in Iraq was basically over since the SURGE, which he said wouldn't work when he was in the Senate, actually DID work. Also, Bush went to Iraq with the BLESSINGS of both houses of Congress, including Obama who was in the Senate at the time. He promised to pull soldiers out of both places if elected, so saying he had to adapt to two wars was HIS DECISION, not anybody else's. Republican's didn't force him to keep GITMO open either, and he triumphantly signed his executive order to close Gitmo in a year, yet it still remains to this day. Whose he gonna blame that one on? Mitt Romney scares me as a RINO candidate, because if he DOES move back to the middle, it will possibly bring on an independent challenger who is much more conservative, and might jeopardize a conservative win and Obama will win in true Clinton fashion, without a majority of the vote. We will have to wait and see. I do know that America has been KICKED awake by the HUGE grab for power that Obama and his regime have decided to take on by "recess appointing" people while congress is not in recess, and nothing has been done about it by RINO republicans. He has decided that he is not going to enforce DOMA, or the Immigration laws of this country. He is going to ignore congress' "contempt of congress" charge against Eric Holder for witholding documents pertinent to their investigation of his program that ended up with many dead innocent mexicans, and 2 border patrol agents with guns supplied by Holder to the Mexican drug cartel for the purposes, apparently, of proclaiming that our gun laws were not tough enough, but it blew up in Holder's face. Face it, Obama's Presidency is INEPT. He passed a law that he SWORE wasn't a tax. The Congress voted on it as "Not a tax", but it was a tax, and now, despite Obama saying he would not raise taxes on anybody making 250K or less (which is a lie now), he will raise his hand triumphantly, and go down in history as the man who imposed the largest tax increase on ALL American's in the history of our Union. As bad as Romney "COULD BE", we know Obama is BAD AS HE IS now, and no pointing fingers at everybody else is going to save his sorry arse. I just pray Romney keeps his conservative message, because it worked for Reagan and it will work for him. Obama makes Carter look like a school girl in comparison, so we need someone who will stick to his conservative values and will finally do what is RIGHT for America, and save us from bankruptcy. Something he is far more qualified to do than Obama, who has never held a job as a CEO or done anything to run a company, realizing what it takes to do so, and turn a profit to stay in business. The Government doesn't have to make a profit, they can just print more money to make up for it, which devalues our dollar more and more. It's time for regime change. Read more Wow, this is what passes for political science?! So, tell us, Dr. Nye, based on your psycho-emotional analysis of their lives, which one, Obama or Romney, has better emotional intelligence? It isn't politicians switching their positions after elections that makes a sham of democracy it is the illusion of any choice in the first place. American capitalism is secure when two candidates wholly and entirely devoted to maintaining it run for office, which is to say, uh, in every election. Elections are rituals designed to give the populace the illusion that they are making some sort of choice and have some sort of say in the process. True a President Gore may not have invaded Iraq, but, being an oilman, he would have been no less anxious to open Iraqi oil to more outside corporate control. Similarly, neither Obama nor Romney favor any basic change when it comes to the status quo on energy policy or anything else I can think of. Health care? Well, Obama is a socialist for corporations so he wants to give private health insurance companies tax dollars while Romney wants people to exclusively pay premiums to them. Take any other issue, you'll find the same. Vote for nobody as only nobody can promise fundamental changes that will wrest control from corporations. Read more Robert Skidelsky on why the right economic policies cannot work without the right public expectations. Project Syndicate provides readers with original, engaging, and thought-provoking commentaries by global leaders and thinkers. By offering incisive perspectives from those who are shaping the world's economics, politics, science, and culture, Project Syndicate has created an unrivaled global venue for informed public debate.

Intravenous Chlorothiazide Versus Enteral Metolazone to Augment Loop Diuretic Therapy in the Intensive Care Unit. In cases of loop diuretic resistance in the intensive care unit (ICU), recommendations for a specific second-line thiazide agent are lacking. To compare the effects of intravenous chlorothiazide (CTZ) and enteral metolazone (MET) on urine output (UOP) when added to furosemide monotherapy therapy in critically ill adults. This was a retrospective cohort study conducted in the medical, surgical, and cardiothoracic ICUs of a quaternary medical center. The primary outcome was change in UOP induced by the study interventions compared with furosemide alone. Secondary outcomes included onset of diuresis, eventual need for hemodialysis, and incidence of adverse events. A total of 122 patients (58 in CTZ, 64 in MET) were included. When added to furosemide monotherapy, CTZ induced a greater change in UOP at 24 hours compared with MET (2405 vs 1646 mL, respectively; P = 0.01). CTZ also caused a more rapid dieresis: 1463 mL total UOP in the first 6 hours compared with 796 mL in the MET group ( P < 0.01). There were no differences found regarding ICU length of stay, need for renal replacement therapy, or survival to discharge. The CTZ arm required more potassium supplementation to maintain normokalemia (median 100 vs 57 mEq in MET; P = 0.02) and carried a higher cost (mean $97 vs $8, P < 0.01). Both CTZ and MET induced significant increases in UOP. CTZ induced a greater and more rapid change and was associated with higher cost and greater need for potassium replacement. Randomized controlled trials are needed to establish whether a preferable thiazide diuretic exists in this setting.

Manufacturing cells such as lithium-ion secondary cells is a complex process characterized both by the chemical interactions that must necessarily occur between the involved substances as well as the mechanical challenges in effectively and reliably assembling several components into a small housing in a sequence that often cannot be reversed. That is, cells are typically fully assembled and thereafter conditioned in a formation process involving electrical charging/discharging as well as heat treatment, before any realistic opportunity exists to quality check the cell. In addition, if a defective cell is discovered at such a late stage in the process, the cell must normally be discarded (e.g., sent to recycling) because there is no practical way to correct the defect at that point. As such, the formation process is an important part of cell manufacturing.

A waveguide filter is an electronic filter that is constructed with waveguide technology. Waveguides are typically hollow metal tubes inside which an electromagnetic wave may be transmitted. Filters are devices used to allow signals at some frequencies to pass (e.g., the passband), while others are rejected (e.g., the stopband). Filters are a basic component of electronic engineering circuits and have numerous applications. These include selection of signals and reduction of noise. Waveguide filters are most useful in the microwave band of frequencies, where they are a convenient size and have low loss. Examples of microwave filter use are found in satellite communications, telephone networks, and television broadcasting, for example. When employed as filters, air cavity waveguide filters have the ability to handle high power and low loss at a fixed frequency. To serve systems with multiple channels, several cavity filters are integrated with switches into a switched filter bank. With the addition of each channel however, the size increases, the cost increases and performance is lowered. These are three of the key performance distracters to air cavity waveguides. Another conventional waveguide filter is a Hititte tunable filter formed as a monolithic microwave integrated circuit (MMIC). This is a single MMIC with multiple tunable filter channels. While compact, these filters have very poor insertion loss (e.g., −30 to −8 dB) making them unusable for most filter bank applications.

There is an image forming apparatus which forms a visible image (toner image) on an image carrier. The image forming apparatus has various motors, various electric circuits, and a heater (hereinafter, referred to as a heating component). The heating component of the image forming apparatus shares apart of an image forming operation when a current is applied. The amount of heat generated from the heating component of the image forming apparatus varies in accordance with an operation load. The amount of heat generated from the heating component of the image forming apparatus in an image forming mode becomes greater than that in a standby mode and a sleep mode of the image forming apparatus. In the image forming mode, the larger the number of continuous prints is, the higher the temperature of the heating component of the image forming apparatus is. Each heating component has an allowable temperature for operating normally. Components other than the heating component of the image forming apparatus also have an allowable temperature based on heat resistance of the components or dimensional stability of the components. The image forming apparatus has cooling fans in order to use each component within an allowable temperature range. The cooling fans include an air blowing fan which supplies low-temperature air to the inside of the apparatus, and an air discharge fan which discharges heated air from the apparatus. The air blowing fan blows air toward the heating component. In the image forming apparatus in the related art, the cooling fans are turned on and off in each operation mode. In the sleep mode, all of the cooling fans are stopped. In the standby mode, the cooling fan excluding the air discharge fan is stopped. In the image forming mode, all of the cooling fans are driven. For example, in the image forming mode, a CPU drives the cooling fans even in the temperature environment in which there is room for the allowable temperature when starting an operation or the like. Each of the cooling fans is designed such that the temperature thereof does not exceed the allowable temperature of each component even if the amount of heat generated from each heating component becomes maximum. For this reason, power consumption is increased due to the rotation of the cooling fans particularly in the image forming mode. Furthermore, noise is increased due to the rotation of the cooling fans.

Memory cards, such as Secure Digital (SD) cards, are used with host devices, such as personal computers and other electronic devices, to store and/or retrieve data. Typically, the memory card fits into a slot in the host device or in a memory card reader attached to the host device. In some host devices, the slot is as long as the memory card, and a mechanism is used to lock-in and eject the memory card. For example, an SD card has a notch on one of its side surfaces, and a detent or other biasing mechanism in the host device can engage the notch when the SD card is completely inserted into the slot. In operation, a user pushes the memory card into the slot until the entirety of the memory card is inside the host device, at which point, the memory card “clicks” into a locking position. When the user removes pressure from the memory card, a spring mechanism inside the host device pushes the memory card slightly toward the user, so a small portion of the memory card is exposed from the host device. To eject the memory card from the host device, the user pushes the exposed portion of the memory card further into the slot, which causes the detent mechanism to disengage from the notch on the memory card and causes a spring inside the host device to eject the memory card from the slot. Other host devices do not have a spring-loaded locking-and-ejection mechanism. For example, some host devices are designed to have a super-compact, ultra-slim form factor without sufficient room for the spring-loaded locking-and-ejection mechanism. As a result, the memory slot in such host devices is designed with a shorter depth, so when the memory card is fully inserted into the slot, a significant portion of the memory card is exposed from the host device. This allows the user to easily grasp the exposed portion of the memory card with his fingers and pull the memory card out of the slot. While an exposed memory card is suitable in many environments, an exposed memory card may not be aesthetically pleasing in some environments, especially with super-compact, ultra-slim form-factor host devices, where an exposed memory card may detract from the host device's sleek appearance and style. The exposed memory card may also prevent the host device from fitting into a form-fitting carrying case.

Q: Do these mean the same: "your friends, boss" .vs. "your friends, your boss"? Do these have the same meaning? And are they both grammatically correct? your friends, your lover, your boss, your dog. your friends, lover, boss, dog. I don't want to put "your" after every comma. A: Yes, they are. The difference would be what is emphasized, the second is just a list, the first stresses that these things are really yours. You might do that if you were in an argument about the ownership of the items, but it's not typical.

<?php /** * CodeIgniter * * An open source application development framework for PHP * * This content is released under the MIT License (MIT) * * Copyright (c) 2014-2019 British Columbia Institute of Technology * Copyright (c) 2019-2020 CodeIgniter Foundation * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to deal * in the Software without restriction, including without limitation the rights * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell * copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in * all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN * THE SOFTWARE. * * @package CodeIgniter * @author CodeIgniter Dev Team * @copyright 2019-2020 CodeIgniter Foundation * @license https://opensource.org/licenses/MIT MIT License * @link https://codeigniter.com * @since Version 4.0.0 * @filesource */ namespace CodeIgniter\HTTP; use Config\UserAgents; /** * Abstraction for an HTTP user agent */ class UserAgent { /** * Current user-agent * * @var string */ protected $agent = ''; /** * Flag for if the user-agent belongs to a browser * * @var boolean */ protected $isBrowser = false; /** * Flag for if the user-agent is a robot * * @var boolean */ protected $isRobot = false; /** * Flag for if the user-agent is a mobile browser * * @var boolean */ protected $isMobile = false; /** * Holds the config file contents. * * @var \Config\UserAgents */ protected $config; /** * Current user-agent platform * * @var string */ protected $platform = ''; /** * Current user-agent browser * * @var string */ protected $browser = ''; /** * Current user-agent version * * @var string */ protected $version = ''; /** * Current user-agent mobile name * * @var string */ protected $mobile = ''; /** * Current user-agent robot name * * @var string */ protected $robot = ''; /** * HTTP Referer * * @var mixed */ protected $referrer; //-------------------------------------------------------------------- /** * Constructor * * Sets the User Agent and runs the compilation routine * * @param null|\Config\UserAgents $config */ public function __construct(UserAgents $config = null) { $this->config = $config ?? new UserAgents(); if (isset($_SERVER['HTTP_USER_AGENT'])) { $this->agent = trim($_SERVER['HTTP_USER_AGENT']); $this->compileData(); } } //-------------------------------------------------------------------- /** * Is Browser * * @param string $key * * @return boolean */ public function isBrowser(string $key = null): bool { if (! $this->isBrowser) { return false; } // No need to be specific, it's a browser if ($key === null) { return true; } // Check for a specific browser return (isset($this->config->browsers[$key]) && $this->browser === $this->config->browsers[$key]); } //-------------------------------------------------------------------- /** * Is Robot * * @param string $key * * @return boolean */ public function isRobot(string $key = null): bool { if (! $this->isRobot) { return false; } // No need to be specific, it's a robot if ($key === null) { return true; } // Check for a specific robot return (isset($this->config->robots[$key]) && $this->robot === $this->config->robots[$key]); } //-------------------------------------------------------------------- /** * Is Mobile * * @param string $key * * @return boolean */ public function isMobile(string $key = null): bool { if (! $this->isMobile) { return false; } // No need to be specific, it's a mobile if ($key === null) { return true; } // Check for a specific robot return (isset($this->config->mobiles[$key]) && $this->mobile === $this->config->mobiles[$key]); } //-------------------------------------------------------------------- /** * Is this a referral from another site? * * @return boolean */ public function isReferral(): bool { if (! isset($this->referrer)) { if (empty($_SERVER['HTTP_REFERER'])) { $this->referrer = false; } else { $referer_host = @parse_url($_SERVER['HTTP_REFERER'], PHP_URL_HOST); $own_host = parse_url(\base_url(), PHP_URL_HOST); $this->referrer = ($referer_host && $referer_host !== $own_host); } } return $this->referrer; } //-------------------------------------------------------------------- /** * Agent String * * @return string */ public function getAgentString(): string { return $this->agent; } //-------------------------------------------------------------------- /** * Get Platform * * @return string */ public function getPlatform(): string { return $this->platform; } //-------------------------------------------------------------------- /** * Get Browser Name * * @return string */ public function getBrowser(): string { return $this->browser; } //-------------------------------------------------------------------- /** * Get the Browser Version * * @return string */ public function getVersion(): string { return $this->version; } //-------------------------------------------------------------------- /** * Get The Robot Name * * @return string */ public function getRobot(): string { return $this->robot; } //-------------------------------------------------------------------- /** * Get the Mobile Device * * @return string */ public function getMobile(): string { return $this->mobile; } //-------------------------------------------------------------------- /** * Get the referrer * * @return string */ public function getReferrer(): string { return empty($_SERVER['HTTP_REFERER']) ? '' : trim($_SERVER['HTTP_REFERER']); } //-------------------------------------------------------------------- /** * Parse a custom user-agent string * * @param string $string * * @return void */ public function parse(string $string) { // Reset values $this->isBrowser = false; $this->isRobot = false; $this->isMobile = false; $this->browser = ''; $this->version = ''; $this->mobile = ''; $this->robot = ''; // Set the new user-agent string and parse it, unless empty $this->agent = $string; if (! empty($string)) { $this->compileData(); } } //-------------------------------------------------------------------- /** * Compile the User Agent Data * * @return void */ protected function compileData() { $this->setPlatform(); foreach (['setRobot', 'setBrowser', 'setMobile'] as $function) { if ($this->$function() === true) { break; } } } //-------------------------------------------------------------------- /** * Set the Platform * * @return boolean */ protected function setPlatform(): bool { if (is_array($this->config->platforms) && $this->config->platforms) { foreach ($this->config->platforms as $key => $val) { if (preg_match('|' . preg_quote($key) . '|i', $this->agent)) { $this->platform = $val; return true; } } } $this->platform = 'Unknown Platform'; return false; } //-------------------------------------------------------------------- /** * Set the Browser * * @return boolean */ protected function setBrowser(): bool { if (is_array($this->config->browsers) && $this->config->browsers) { foreach ($this->config->browsers as $key => $val) { if (preg_match('|' . $key . '.*?([0-9\.]+)|i', $this->agent, $match)) { $this->isBrowser = true; $this->version = $match[1]; $this->browser = $val; $this->setMobile(); return true; } } } return false; } //-------------------------------------------------------------------- /** * Set the Robot * * @return boolean */ protected function setRobot(): bool { if (is_array($this->config->robots) && $this->config->robots) { foreach ($this->config->robots as $key => $val) { if (preg_match('|' . preg_quote($key) . '|i', $this->agent)) { $this->isRobot = true; $this->robot = $val; $this->setMobile(); return true; } } } return false; } //-------------------------------------------------------------------- /** * Set the Mobile Device * * @return boolean */ protected function setMobile(): bool { if (is_array($this->config->mobiles) && $this->config->mobiles) { foreach ($this->config->mobiles as $key => $val) { if (false !== (stripos($this->agent, $key))) { $this->isMobile = true; $this->mobile = $val; return true; } } } return false; } //-------------------------------------------------------------------- /** * Outputs the original Agent String when cast as a string. * * @return string */ public function __toString(): string { return $this->getAgentString(); } }

Show HN: Icon Tryer Outer – easily try out icons on a mobile device - DanielDe https://www.icontryerouter.com/ ====== ent101 Very useful! do you think you might add support for device simulation? i.e. show how the icon will look on a variety of devices in the browser... ~~~ DanielDe Thanks! That's a good idea, but I don't have plans to do that right now. The intention of the project is to let you try out icons on an actual device, as opposed to in a simulator.

Postnatal shifts in ischemic tolerance and cell survival signaling in murine myocardium. The immature heart is known to be resistant to ischemia-reperfusion (I/R) injury; however, key proteins engaged in phospho-dependent signaling pathways crucial to cell survival are not yet defined. Our goal was to determine the postnatal changes in myocardial tolerance to I/R, including baseline expression of key proteins governing I/R tolerance and their phosphorylation during I/R. Hearts from male C57Bl/6 mice (neonates, 2, 4, 8, and 12 wk of age, n = 6/group) were assayed for survival signaling/effectors [Akt, p38MAPK, glycogen synthase kinase-3β (GSK-3β), heat shock protein 27 (HSP27), connexin-43, hypoxia-inducible factor-1α (HIF-1α), and caveolin-3] and regulators of apoptosis (Bax and Bcl-2) and autophagy (LC3B, Parkin, and Beclin1). The effect of I/R on ventricular function was measured in isolated perfused hearts from immature (4 wk) and adult (12 wk) mice. The neonatal myocardium exhibits a large pool of inactive Akt; high phospho-activation of p38MAPK, HSP27 and connexin-43; phospho-inhibition of GSK-3β; and high expression of caveolin-3, HIF-1α, LC3B, Beclin1, Bax, and Bcl-2. Immature hearts sustained less dysfunction and infarction following I/R than adults. Emergence of I/R intolerance in adult vs. immature hearts was associated with complex proteomic changes: decreased expression of Akt, Bax, and Bcl-2; increased GSK-3β, connexin-43, HIF-1α, LC3B, and Bax:Bcl-2; enhanced postischemic HIF-1α, caveolin-3, Bax, and Bcl-2; and greater postischemic GSK-3β and HSP27 phosphorylation. Neonatal myocardial stress resistance reflects high expression of prosurvival and autophagy proteins and apoptotic regulators. Notably, there is high phosphorylation of GSK-3β, p38MAPK, and HSP27 and low phosphorylation of Akt (high Akt "reserve"). Subsequent maturation-related reductions in I/R tolerance are associated with reductions in Akt, Bcl-2, LC3B, and Beclin1, despite increased expression and reduced phospho-inhibition of GSK-3β.

Certs We went to Lub­bock, Tx and had our units and doors test­ed at Texas Tech Uni­ver­si­ty. There are sev­er­al places to have test­ing done; how­ev­er, Texas Tech is the lead­ing rec­og­nized author­i­ty on wind sci­ence engi­neer­ing and test­ing. The engi­neer of record, Lar­ry Tan­ner P.E., at our test­ing asked if we were going to sell our door as a sep­a­rate item. He said, “Our door was the best he had seen,” “no one is mak­ing any­thing like this.” We informed him that this was our inten­tion. Check out our Pic­ture Gallery and under the tool bar “D.I.Y.” our door is exact­ly like our stan­dard Gen 3 shel­ter door; how­ev­er we build each one to fit your needs — any size open­ing you have — and the price is only $2260.00. We should men­tion trade asso­ci­a­tions and like groups. While they do have a pur­pose, at this time we have cho­sen to not par­tic­i­pate in these groups (and spend $$ with them). Our pol­i­cy and busi­ness eth­ic is sim­ple: fol­low the basic prin­ci­ples of life and be fair in all our deal­ings. the cus­tomer is Boss. We need to sat­is­fy the cus­tomer and treat every­one “like we want to be treat­ed.” We will be glad to pro­vide a list of cus­tomers we have done work for as ref­er­ences of our per­for­mance. We are con­stant­ly try­ing to improve our prod­uct and give our cus­tomers the best val­ue for their trust in us— this is our focus.

Come and spend your holiday with us at The Orchards Holiday Park. Winner of The AA, Campsite of the year for the SE for 2015. The Orchards has stunning countryside views, 5 star facilities and friendly staff. These are some of the reasons our visitors return year after year. We have been awarded 18, gold David Bellamy awards for conservation, we are the only Best of British park on the Isle of Wight and we are affiliated to The Caravan Club. The West Wight has walking,cycling,nature reserves, sandy beaches, rugged coastlines, historical buildings and pretty villages to explore. We have 64 Holiday Caravans,all with,toilet, shower,fridge,heater,colour TV DVD and microwave, most with double glazing and central heating.There are 160 marked allocated pitches aranged in 3 paddocks.Some terraced, some hardstanding and some grass.Our facilities include a heated, Indoor pool and poolside coffee shop, open all season, outdoor heated pool, (open late May to early Sept), shop, takeaway, TV room pool table and table tennis, play areas and dog walking area.The impressive new Facilities Centre includes some of the best touring facilities in the country, with a large washing up, laundry and ironing area, family bathroom, family shower room ladies vanity area, walkers drying room,campers kitchen area, full accessible facilities and baby changing room. All with the luxury of underfloor heating.Online booking available on own website. WiFi available at a small cost. The NO 7 bus stops at the entrance to the site every hour going to Yarmouth and Newport, where you can catch connecting buses. Book as a Package Deal (includes Ferry) or on a Non-Packaged Basis. Wi-Fi is available Nearby Attractions Nearby attractions include, Carisbrooke Castle, Chessil Pottery (where you can make your own pot) and the sandy beaches of Compton and Colwell bay are all within 5 miles of the park. Newtown Creek, the only National Wildlife Centre on the island, is a must for wildlife lovers. There are also footpaths and cycle routes from the park. Nearby are, The Needles, Alum Bay coloured sands, Osborne House,Cowes, Blackgang Chine, Dinosaur Isle ,Isle of Wight Pearl. Yarmouth, our nearest small town, is one of the most picturesque and oldest on the Island,With individual shops cafes and art galleries, a busy harbour overlooked by a castle, and a wooden pier ideal for watching boats on The Solent. Freshwater with a large Co-op is only 10 minutes away. Directions From East Cowes (Southampton) and Fishbourne (Portsmouth)• Drive to Newport and follow signs to Yarmouth on A3054• After approximately 6 miles pass through Shalfleet village• After a further half mile turn left at the Horse & Groom Inn• Follow signs to Newbridge• Entrance to The Orchards is opposite the village Post Office Further directions can be found on or website. Special Offers and Discounts for UKCampsite.co.uk Visitors Standard Offers And Packages Expiry: Never http://www.orchards-holiday-park.co.uk/touring-and-camping/tariff/ Please note all offers must be mentioned at the time of booking. Subject to availability. Usually only one offer can be used per booking, please check with site for further details. Nearest English Heritage Visitor Attractions Watch a Video The facilities shown on this website are a guide only, and you should check with the campsite direct before making any booking. The UKCampsite.co.uk will not be held responsible for any errors or omissions Like most websites this one uses cookies. Some may have been set already. You can read our cookie policy here. To accept our cookies and hide this click here, or if you continue to use the site, we'll assume you're happy to accept them anyway

Empirical evidence for resource-rational anchoring and adjustment. People's estimates of numerical quantities are systematically biased towards their initial guess. This anchoring bias is usually interpreted as sign of human irrationality, but it has recently been suggested that the anchoring bias instead results from people's rational use of their finite time and limited cognitive resources. If this were true, then adjustment should decrease with the relative cost of time. To test this hypothesis, we designed a new numerical estimation paradigm that controls people's knowledge and varies the cost of time and error independently while allowing people to invest as much or as little time and effort into refining their estimate as they wish. Two experiments confirmed the prediction that adjustment decreases with time cost but increases with error cost regardless of whether the anchor was self-generated or provided. These results support the hypothesis that people rationally adapt their number of adjustments to achieve a near-optimal speed-accuracy tradeoff. This suggests that the anchoring bias might be a signature of the rational use of finite time and limited cognitive resources rather than a sign of human irrationality.

The 6'5 Freshman Guard from Conyers, GA., was orginally denied fall admittance at Louisville. However, he eventually met the NCAA eligibity requirments and joined the Louisville Cardinals Basketball Team Wednesday. Who is Kevin Ware and How will he help the Cardinals in 2011-2012 Basketball season? Read and find out.

Venus Versus Virus Venus Versus Virus Sumire is a girl who has had the ability to see ghosts since she was a child. One night when she stays in school until late, she finds a broach on the ground. She tries to pick it up and accidentally pricks her finger on it, and a monster appears and attacks her. Suddenly, another girl clad in gothic lolita clothing arrives bearing a gun and kills the demon. Once the ordeal is over with, the girl gives Sumire a flier for a gothic lolita clothing store, Venus Vanguard. When Sumire comes to the store, the girl, whose name is Luchia, she explains to her that because she pricked her finger on Luchia's broach, Sumire now has the power to see ghosts much more clearly, as well as evil spirits called Viruses. Now, in exchange for saving her life, Luchia demands that Sumire will play bait for the Viruses and help her out in her job.

#ifdef __cplusplus extern "C" { #endif void nms_cuda_compute(int* keep_out, int *num_out, float* boxes_host, int boxes_num, int boxes_dim, float nms_overlap_thresh); #ifdef __cplusplus } #endif

Q: A loop suddenly ceases to iterate? I have this php code: <?php $numarray = trim($_GET['num']); $i = strlen($numarray); $result = ""; $numneedarray = array( 90 => 'ninety', 80 => 'eighty', 70 => 'seventy', 60 => 'sixty', 50 => 'fifty', 40 => 'forty', 30 => 'thirty', 20 => 'twenty', 19 => 'nineteen', 18 => 'eighteen', 17 => 'seventeen', 16 => 'sixteen', 15 => 'fifteen', 14 => 'fourteen', 13 => 'thirteen', 12 => 'twelve', 11 => 'eleven', 10 => 'ten', 9 => 'nine', 8 => 'eight', 7 => 'seven', 6 => 'six', 5 => 'five', 4 => 'four', 3 => 'three', 2 => 'two', 1 => 'one' ); for ($v = 1; $v <= $i; $v++) { if ($i > 10) { exit("Has to be 10-digit or less."); } if ($i == 3) { $result .= ", " . $numneedarray[$numarray[strlen($numarray) - 3]] . " hundred"; if ($numarray % 100 == 0) { echo "Hi95"; break; } } elseif ($i == 2) { if ($numarray[$v] == 1) { $othernum = $numarray[strlen($numarray) - 1]; $othernum2 = $numarray[strlen($numarray) - 2] * 10; $othernum3 = $othernum2 + $othernum; if (strlen($numarray) > 2) { $result .= " and "; } $result .= $numneedarray[$othernum3]; echo "Hi107"; } else { $othernum = $numarray[strlen($numarray) - 2] * 10; $othernum2 = $numarray[strlen($numarray) - 1]; $result .= " and " . $numneedarray[$othernum] . " " . $numneedarray[$othernum2]; echo "Hi112"; } } if ($i == 10) { $digit = substr($numarray, 0, 1); $result .= $numneedarray[$digit] . " billion"; if ($numarray % 1000000000 == 0) { break; } } elseif ($i == 9) { $number = substr($numarray, 1, 3); $digit1 = substr($number, 0, 1); $digit1con = $numneedarray[$digit1] . " hundred"; $digit2 = substr($number, 1, 1); $noneed = false; if ($digit2 != 1) { $digit2con = $numneedarray[$digit2 * 10]; } else { $digit23 = substr($number, 1, 2); $digit23con = $numneedarray[$digit23]; $noneed = true; } $digit3 = substr($number, -1); $digit3con = $numneedarray[$digit3]; if ($noneed == true) { $result .= ", " . $digit1con . " and " . $digit23con . " million"; } else { $result .= ", " . $digit1con . " and " . $digit2con . " " . $digit3con . " million"; } if ($numarray % 100000000 == 0) { echo "Hi"; break; } } elseif ($i == 6) { $number = substr($numarray, 4, 3); $digit1 = substr($number, 0, 1); $digit1con = $numneedarray[$digit1] . " hundred"; $digit2 = substr($number, 1, 1); $noneed = false; if ($digit2 != 1) { $digit2con = $numneedarray[$digit2 * 10]; } else { $digit23 = substr($number, 1, 2); $digit23con = $numneedarray[$digit23]; $noneed = true; } $digit3 = substr($number, -1); $digit3con = $numneedarray[$digit3]; if ($noneed == true) { $result .= ", " . $digit1con . " and " . $digit23con . " thousand"; } else { $result .= ", " . $digit1con . " and " . $digit2con . " " . $digit3con . " thousand"; } if ($numarray % 100000 == 0) { echo "Hi89"; break; } } echo $i; $i = $i - 1; } if (strlen($numarray) == 1) { echo $numneedarray[$numarray]; } echo $result; ?> The num value is equal to 1234567890. When I refresh the page, the value of $i only goes from 10 -> 9 -> 8 -> 7 -> 6 and then suddenly stops. Why would the loop stop running? A: You try to check each digit: for ($v = 1; $v <= $i; $v++) { But then you decrease $i in the loop: $i = $i - 1; That's why you see less loops than you expected.

Toxicity, repellency, and effects of acetamiprid on western subterranean termite (Isoptera: Rhinotermitidae). The insecticidial and biological activity of the cyano-substituted neonicotinoid acetamiprid was determined against the western subterranean termite, Reticulitermes hesperus Banks (Isoptera: Rhinotermitidae). Acetamiprid was very active against termites in topical applications, with an LD50 = 0.02 ng per termite. Even though acetamiprid was extremely toxic in topical applications, deposits > or =50 ppm on sand were required to consistently provide >90% kill of termites within 7 d after a 1-h exposure. Termites were quickly affected by brief exposures to sand treated with 1 ppm acetamiprid and within 1 h, their locomotion was dramatically impaired. Acetamiprid was transferred from donors to recipients only when donors were held on deposits > or =50 ppm for 1 h. Deposits even as low as 1 ppm were repellent and termites failed to tunnel into treated sand, and there was no significant mortality. Exposure to acetamiprid impaired locomotion of termites as did other slow-acting neonicotinoids, such as imidacloprid. Acetamiprid was repellent at all concentrations tested, acting like type I pyrethroid treatments in soil. A new subcategory of type III soil termiticides is proposed that incorporates the sublethal and delayed effects observed in neonicotinoid insecticides, and repellency at certain concentrations.

[Nasal glioma--case report]. Benign congenital lesions resulting from the deficient regression of neuroglial tissue in normal embryonic development are called gliomas. They are usually located in the nasofrontal region and are diagnosed in the postnatal period. They are included in the differential diagnosis of nasofrontal midline masses. We present a case of the 9-year-old boy treated in the ENT Department of Wroclaw Medical University with the initial diagnosis of nasal polyposis or tumour which appeared to be glioma in the postoperative histological evaluation. The tumour was removed endoscopically. We present magnetic resonance imaging of the case together with the review of the literature concerning gliomas. We conclude that each doctor has to be aware of possible neoplasm even in the youngest groups of the patients.

Metribuzin resistance in photoautotrophic Chenopodium rubrum cell cultures. Characterization of double and triple mutations in the psbA gene. Sequence analysis of eight metribuzin-resistant mutants of photoautotrophic Chenopodium rubrum cell cultures revealed new mutations in the psbA gene coding for the 32 kDa herbicide binding protein. Mutants were found to possess either two or three changes in the amino acid sequence of the D1-protein between positions 219 and 272.

Kernel-level attacks or malicious programs, such as rootkits, that compromise the kernel of an operating system are one of the most important concerns in systems security at present. These attacks can run at the same privilege level as the kernel and can thus modify kernel-level code or sensitive data to hide various malicious activities, to change operating system behavior, or even to take complete control of the system. Kernel-level security tools can be crippled and made ineffective by these attacks, which can run, access, and modify these security tools. A large body of research has adopted virtual machine (VM) monitor technology in an effort to mitigate such attacks. A higher privileged hypervisor outside of a virtual machine in which the kernel runs can enforce memory protections and preemptively intercept events throughout the operating system environment. A major reason for adopting virtualization is to isolate security tools from an untrusted VM by moving those security tools to a separate, trusted, secure VM, and then using introspection to monitor and protect the untrusted VM from inside the trusted VM. Approaches that passively monitor various security properties, by periodically looking inside the untrusted VM for evidence of suspicious activity, have been proposed, but passive monitoring can only detect remnants of an already successful attack. Active monitoring from outside of the untrusted VM, in contrast, has the advantage of detecting attacks earlier and preventing certain attacks from succeeding. Active monitoring from outside of an untrusted VM can be achieved by placing secure hooks inside the untrusted VM, to intercept various events and invoke the security tool residing in a separate secure VM. Because the secure VM is isolated from the untrusted VM, so as to prohibit tampering, switching between the VMs occurs through the hypervisor. But the large overhead for switching between the untrusted VM, the hypervisor, and the secure VM makes this approach suitable only for actively monitoring a few events that occur less frequently during system execution. Thus, with previous systems and methods, in-VM monitoring provides an inadequate level of security, while monitoring from outside a VM is feasible only when limiting the number and type of events that can be actively monitored, and.

Granuloma faciale: distribution of the lesions and review of the literature. Granuloma faciale (GF) is an uncommon inflammatory condition characterized by reddish brown papules and plaques that usually involve the facial area. Extrafacial lesions are rare. Histologically, the lesions are marked by leukocytoclastic vasculitis and extensive fibrin deposition. There are a variety of treatment options available for GF.

Tyn Head Court Tyn Head Court, also known as Wethered Court, is a historic home located at Dover, Kent County, Delaware. It dates to about 1740, and is a two-story, three bay, stuccoed brick dwelling with a gambrel roof. The roof has three shed-roofed dormers. It was added to the National Register of Historic Places in 1973. References Category:Houses on the National Register of Historic Places in Delaware Category:Houses completed in 1740 Category:Houses in Dover, Delaware Category:National Register of Historic Places in Dover, Delaware

Tennessee is among 28 states sharing a total $950,694 in grant funding for projects to combat white-nose syndrome, a U.S. Fish and Wildlife Service news release states. The disease is estimated to have killed nearly 6 million bats in North America. "These grants provide essential support to our state partners in responding to this disease," said Dr. Jeremy Coleman, the service's national white-nose syndrome coordinator. "The research, monitoring, and actions made possible by these grants have yielded valuable results and insights for our national response" to the disease. Funding for the grants was provided through the Endangered Species Recovery program. Proposals were received from 28 states requesting $1,042,938. All eligible requests were given at least partial awards, ranging from about $7,000 to $47,500, for the total of $950,694. Tennessee is projected to receive $47,500, but the amounts are under final review.

The Revenue Men The Revenue Men is a British television series, produced and transmitted by the BBC between 1967 and 1968. The series dealt with cases handled by the Investigation Branch of Customs and Excise such as the illegal import of goods, illegal immigration and business transactions amongst travellers. The Revenue Men was produced by Gerard Glaister. The series lasted for three series and 39 episodes in total. In spite of this fact, all of the episodes were later wiped, with no episodes extant in the BBC archives as of 2009. External links Category:BBC television dramas Category:Lost BBC episodes Category:1960s British drama television series Category:1967 British television series debuts Category:1968 British television series endings Category:British crime television series

This project is aimed towards understanding the mechanisms which mediate and regulate Ca-2+ signaling in salivary gland cells. Neurotransmitter stimulation of fluid secretion in salivary glands is mediated via a biphasic elevation in cytosolic [Ca-2+]; an initial transient increase due to internal release and a latter sustained increase due to Ca-2+ influx. Sustained fluid secretion is directly dependent upon the sustained elevation of [Ca-2+] and thus on Ca-2+ influx. Recently, our efforts have been focused on the Ca-2+ influx mechanism in salivary gland cells, which appears to be a mediated via store-operated Ca-2+ entry (SOCE) that is ubiquitously present in many other non-excitable cells. The molecular mechanism(s) of this influx has not yet been determined in any cell type. Recently, the transient receptor potential (TRPC) family of ion channel proteins have been proposed as molecular components of the store-operated Ca-2+ influx channel (SOCC). However, the physiological function(s) of the presently identified TRPCs has not yet been fully established. By expressing TRPC1 in vivo in rat SMG by using an adenovirus encoding hTrp1 (AdHA-hTrp1) and in salivary gland cell lines, we had previously reported that TRPC1 is involved in the regulation of store-operated calcium influx in salivary gland cells. In the past year our major effort has continued to be towards characterizing SOCE and identifying the role of TRPC channels in agonist-stimulated calcium entry in salivary gland cells. Consistent with our previous studies, we have now reported that caveolin has a critical role in SOCE by regulating the plasma membrane assembly of TRPC1 channels. Mutation in the caveolin-binding domain of TRPC1 disrupted its plasma membrane localization and exerted a dominant negative effect on SOCE. Further, we have reported that TRPC3 interacts with SNARE proteins and that this interaction is involved in membrane trafficking of TRPC3. We have shown that TRPC3 undergoes constitutive and regulated trafficking mechanisms. Importantly, agonist-stimulated PIP2 hydrolysis, increases exocytotic insertion of TRPC3 into the plasma membrane and this contributes towards the increase in calcium entry seen in stimulated agonist-cells. Our previous studies have been largely carried out with HSG cells in which we have demonstrated conclusively that TRPC1 is the primary SOCE component. Towards characterizing SOCE components in other salivary gland cells, we have now studied SOCE in HSY cells. Our results demonstrate that distinct store-operated Ca2+ channels are present in different cells. In HSG cells, SOCE is associated with a relatively Ca2+ selective current whereas HSY cells display a non-selective cation channel. Both of these channels are distinct from the CRAC channel in RBL cells, the components of which are presently unknown. Although the physiological relevance of these different SOCE channels is not presently clear, we have now examined the molecular components of the channel in HSY cells. Interestingly, while the HSG channel appears to primarily depend on TRPC1, the HSY channel appears to be formed by the coassembly of TRPC1 and TRPC3. Further, TRPC1-TRPC3 interactions are mediated via their N-terminal domains. Expression of the NTRPC1 disrupts SOCE in HSY cells. Thus, we propose that TRCs can assemble as homomers or heteromers to form SOC and that the channel properties are defined by the specific TRPC components that are involved. To assess the physiological relevance of TRPC channels we have examined their routing in polarized epithelial cells. We have established stable TRPC expression in MDCK and salivary epithelial (SMIE) cells, both of which form high resistance monolayers when cultured on Transwell filters. We have obsevred that TRPCs have distinct cellular localization. TRPC3 is apically localized, TRPC1, TRPC5, and TRPC2 are basaly localized while TRPC6 is found in both apical and basal regions of the cell. Further, endogenous TRPC3, TRPC1, and TRPC6 were found at the same locale. We have also studied the regulation of the TRP channels in these cellular regions. Consistent with previous reports, Ca2+ signaling proteins were also predominantly localized in the apical region of these cells. Further, TRPC3 was assembled in a complex with TRPC6, but not TRPC1, and key Ca2+ signaling proteins like IP3R, G-proteins, and PLC. Importantly, we showed that TRPC3/TRPC6 channels can mediate apical calcium uptake and transepithelial calcium transport in polarized epithelial cells. These data demonstrate a novel role for TRPC3/TRPC6 channels; i.e. agonist-stimulated apical calcium uptake. Consistent with this, we detected localization of TRPC3 and TRPC6 in the apical regions of salivary gland and kidney ducts. Studies are ongoing to determine how these apically localized channels are regulated and what is their physiological function. In the coming fiscal year we will continue our studies along these directions. A major focus will directed towards determining novel TRPC interacting proteins to help us to understand their function and regulation. We will also continue to study the trafficking of TRPC channels and the mechanisms involved in thier assembly and multimerization. Stduies will also be directed identifying specific SOC currents generated by different TRPC combinations.