[00:47] when I log onto another account's desktop, my music player goes silent. But if I'm running a noisy game, it is still heard. How can I make the music always play? [12:24] hi all [12:26] I was trying to install edubuntu classroom server to connect some clients as x-terminal, the installation went ok, but i cant configure the clients. i cant find the ltsf.conf file. am using the latest edubuntu. [12:29] zzaza: u have to create that file [12:29] it should be here: /var/lib/tftpboot/ltsp/i386/lts.conf [12:30] and the first line should be [default] [12:31] nubae: i didt creat that file, i last tried edubuntu version 7.10 all was striaght forward [12:32] i did check in /var/lib/tftpboot but there was no lts.conf file [12:33] I know... [12:33] u have to create it [14:00] ahhh damn Nick Kendall.... === ogra_ is now known as ogra [16:16] Morning all [16:16] hi sbalneav [16:17] morning sbalneav [16:17] I'm having a problem with pulseaudio... [16:17] I get a connection refused when it tries to connect to the pulse server [16:18] ah wait... forget it, just fixed it... [16:18] What was it?> [16:19] needed echo "default_driver=pulse" in .libao [16:19] wierd though, thats new for intrepid [16:19] cause hardy worked fine [18:06] hi Lns [18:06] hey nubae [18:07] after adding something to anacron in cron.hourly, does one need to reload cron? [18:07] nubae: Not if it works like regular old cron [18:07] the man page says itself that one of its features is that you don't need to reload it after editing configs [18:08] its weird, I've got a script that works fine if i execute it, and its in /etc/cron.hourly, but doesnt seem to get exectued [18:08] by cron [18:08] nubae: did you chmod +x it ? [18:09] yep [18:09] err [18:09] yeah it would if yo ucan yousefl ;) [18:09] ick..monday morning fingers [18:09] nubae: did you reload anacron or no? [18:09] I did that too, without much success :-) [18:10] it works if I make an entry with crontab -e, but I thought the point of cron.hourly is u drop the script in there to exectue every hour [18:10] it is [18:10] hmmm... then what am I missing [18:10] nubae: what does the script do? [18:11] awks users out of passwd, shadow and groups and puts into another file [18:12] ah [18:12] ah? u know whats wrong? [18:12] nubae: have you verified other scripts run in cron.hourly? [18:12] no, not yet ;) [18:12] there arent any [18:12] have you verified crontab has a stanza for cron.hourly runs? [18:13] and there is no /var/spool/anacron/cron.hourly [18:13] which I guess answers your question [18:13] hrmm... /me is an old style cron user [18:14] where everything is in /etc/crontab [18:14] well guess Im gonna have to do the same, I just thought this would be neater [18:14] sure, i mean it should work [18:14] is it stock anacron in hardy? [18:15] 17 * * * * root cd / && run-parts --report /etc/cron.hourly --- that should execute anacron [18:15] oh ok we're talking about different beasts here [18:15] yep [18:15] but it doesn! [18:15] nubae: did you try running "run-parts --report /etc/cron.hourly" manually on a shell, and/or examining logs? [18:15] (as root) [18:16] no thats an idea, let me try that [18:17] yep works [18:17] sigh... this makes no sense [18:19] hrm [18:19] look in your cron logs [18:21] nubae: FYI AFAIK there's anacron, then cron. they're 2 different things, and both installed by default [18:21] make sure you're running the right one [18:21] anacron does cron.hourly, cron.daily, etc as I understand it [18:22] and cron does /etc/crontab [18:22] or am I missing something? [18:23] nubae: anacron uses /etc/anacrontab [18:23] well doh no wonder... there is no hourly in there :-) [18:23] how silly is that [18:23] ;) [18:24] its supposed to replace cron but has no cron.hourly [18:24] at least got to the bottom of it, thanks Lns [18:24] np [18:24] yeah what's weird is that i have a different line in /etc/crontab for hourly [18:25] with cd / && in the front? [18:25] yeah [18:25] yeah took that out =) [18:25] and why does both crontab and anacrontab have day/week/month stanzas? [18:27] it should be one or the other [18:27] its very confusing [18:27] yeah... [18:27] I really thought cron.hourly was run by cron, but its not, its run by anacron and isnt in the anacrontab [18:28] well it looks like it's supposed to, with "17 * * * * root cd / && run-parts --report /etc/cron.hourly" in crontab [18:29] yes but without the corresponding line in cron.hourly in anacron, its not being spooled it looks like [18:30] nubae: looks like it works for me... Oct 6 10:17:01 Fibonacci /USR/SBIN/CRON[15189]: (root) CMD ( cd / && run-parts --report /etc/cron.hourly) [18:31] I have nothing in cron.hourly though so i haven't tested [18:31] yeah its annoying cause its working for me too [18:31] nubae: are you using a sane shell descriptor at the top of your script? [18:31] yes [18:32] gtonna test with another script [18:32] hrm..actually running it manually didn't work for me [18:32] oh wait [18:33] hmmmmmmmmm [18:34] yeah it's not working for me either (running run-parts manually) [18:34] so crontab pretends to do its job [18:35] yeah [18:35] and even reports to syslog [18:35] cron executes cron.hourly, but then it stops... so I guess cron.hourly triggers anacron? [18:35] is that possible? [18:35] not sure [18:35] because its only hourly not working [18:35] its just a dir...i can't see how [18:35] tried with a real simple test... touch itworks.txt [18:36] nada... [18:36] yeah me too [18:36] only if you i wonder why crontab has a different style execution for .hourly [18:36] taking out the test anacron [18:38] yeah what does that do, test anacron? [18:38] tests to see if anacron exists [18:38] but to me it almost looks like run-parts is broken [18:39] run-parts --list /etc/cron.hourly/ doesn't result in any scripts being returned, although i do have one in there and it's chmod +x'd [18:39] well, lets try a daily one... by setting the clock to a minute after or so... [18:42] no work either [18:42] ure right run-parts must be broken [18:42] that means all the scripts that are in there are not being executed [18:42] lol [18:42] thats a pretty serious bug [18:42] well i'd hope others would notice too [18:42] we could be doing something wrong [18:44] yeah like what? ever document Ive read and every wiki points to simply putting the script in .hourly or .daily or whatever, + there are scripts in there already and they're not being executed [18:44] nubae: http://ubuntuforums.org/showthread.php?p=5677105 [18:44] its a mystery [18:45] that would mean that by default everyones crons are broken [18:45] i sure would hope that isn't the case [18:45] thats *buntu, not just edu (i'm using ubuntu) [18:46] me too [18:47] so how do we verify that, I guess if a reinstall fixes it, then that is verification [18:48] guess so.. ? [18:49] sudo apt-get reinstall cron anacron --- that does not work [18:50] --reinstall [18:50] yeah the www doc is wrong [18:52] nah that doesnt work [18:53] nubae: apt-get --reinstall install cron anacron [18:53] yeah I know, I mean that didnt fix the problem [18:55] ah [18:56] not sure...very eerie though [18:57] its run parts thats broken, cause putting into crontab normally works [18:58] hold on a sec [18:58] just dont want to put a bug report up, unless I'm absolutely certain :D [18:59] AH! [18:59] you have an extension ( .sh ?) on your script, no? [18:59] run-parts works when you remove all script extensions [19:00] no [19:00] :-) [19:00] no .sh, I read about that [19:01] well that fixed it for me =) [19:01] are you using /bin/sh as your shell in the script? [19:01] have you 'set -e' in it at the top? [19:01] wait Ive still got the cd / && missing, let me put it back [19:02] set -e ? [19:03] i believe that sets the environment [19:03] ok, lets see if that does it for me [19:04] yep :-) [19:04] thanks [19:04] shit, that was right at the top of the run-parts manpage too [19:04] that only took half a day [19:04] If neither the --lsbsysinit option nor the --regex option is given then the names must consist entirely of upper [19:04] and lower case letters, digits, underscores, and hyphens. [19:05] so the 'set -e' did it for you, eh? [19:05] yep [19:05] damn anal cron ;) [19:05] nowhere was that mentioned [19:06] think I'll add an entry to my blog about this wonderful cron experience [19:06] lol..nice [19:06] god, feel like I was banging my head against a wall [19:07] cron (moreso run-parts) should REALLY be more forgiving [19:08] of course i'm sure there's a very good reason for its behavior, I'm just not seeing it [19:08] I had #!bin/sh there [19:08] jsut not the env [19:08] gonna check the other scripts [19:09] default ones normally have it - that's where i got it from i think [19:09] well apt has it disabled [19:09] "set -e" terminates the shell if any command returns a non-zero status [19:10] check yours... cron.daily/apt [19:17] yep [19:17] doesn't have it [19:17] (commented out) [19:18] so wtf? [19:19] not sure..maybe not every script requires it [19:20] yeah, guess I should just be glad it works now :D [19:20] yeah but understanding it would be nice too =) [21:01] SO... regarding https://bugs.launchpad.net/bugs/269188 - the people in #ubuntu-mozillateam seem to think that getting this patch into the beta Intrepid is even a longshot. Any thoughts, since it affects *buntu LTSP setups on a critical level? [21:01] Launchpad bug 269188 in firefox "Extreme slowness, "Firefox is already running" error for >3 users launching Firefox in LTSP environment" [Undecided,Incomplete] [21:03] Lns: The school I was working at before also mentioned extreme firefox slowness [21:03] nubae1: I don't doubt it [21:05] It seems like it would be critical to any Ubuntu LTSP setups with > 3 thin-clients (and nobody at the server feeding it mouse/keyboard entropy for everyone else) [21:05] I'm just really scared that people aren't even going to look as far as Launchpad when installing *buntu LTSP and will just walk away since Firefox is broken [21:05] And if it's really THAT hard to get a patch pushed into Hardy (even in -proposed)... what's the use? [21:06] I'll add my 2 cents to the bug list and shout a bit ;-) [21:06] Lns: Any reason given? [21:06] heh...thanks nubae1 it'll be nice to have someone else's whining in there ;) [21:07] sbalneav: 'fta' in #ubuntu-mozillateam says "we could take it in nspr, but it's a bit late in this cycle.. and we have the de-sonamization still under discussion blocking both nspr and nss" [21:08] I don't know what "nspr", "nss" or "de-sonamization" means... [21:08] hehe === nubae1 is now known as Nubae [21:09] Nubae: Section on "hardware", talking about hubs, switches, etc... [21:09] You want them converted to sect3's, or maybe a bulletlist. [21:09] Oh, btw, gimme 2 seconds, I'll push up my branch [21:11] ok... whatever u think looks best, I set them up as sect3s, but obviously openoffice killed that [21:14] Crumb. Can I delete a branch on launchpad? [21:14] * sbalneav thinks he'll never get used to LP [21:22] * Lns watches http://www.youtube.com/watch?v=VyEl3w7SFK4 [21:24] heh, I was watching that the other day [21:25] I opted for printouts in the end [21:28] ;) [21:32] Lns: does your triage fix the issue totally? [21:32] Nubae: I don't know, but I'm assuming so. I haven't tested it (I don't know how to build packages/PPA it for use in my environment) [21:32] That's why i'm learning about MOTU... [21:33] I know that through all my troubleshooting, it was determined that /dev/random was sucking up all the entropy, especially when creating new user profiles in Firefox, and /dev/urandom would be a better choice since it doesn't block requests when entropy runs out [21:33] I have a hack-ish patch that would work for testing [21:34] but not the same patch..it would, however, prove that using /dev/urandom would fix the bug [21:35] \o/ [21:35] got it [21:35] Nubae: OK, check out lp:~sbalneav/ltsp/ltsp-docs-sbalneav [21:36] There's a LOT more work needing to be done, I'll work on it tonight, but at least I've got all of it in book form, with all the proper entities. [21:36] OO.o's inserted some weeeiiird characters here and there, I have to hose them [21:37] Nubae: download http://logicalnetworking.net/other/urandomize.tar.bz2 if you want to test it - you need to compile urandomize.so and then load it so calls to /dev/random get pushed to /dev/urandom instead (see README) [21:37] plus, all of the screencap stuff like ps listings etc. has to be converted back into entities, so that's left to do [21:37] ok, if there is anything I can do, sbalneav, let me know [21:37] although I fear I've already done too much ;-p [21:38] sh*t happens, we deal :) [21:39] content's more important, so if content's been added, I'm happy to do whatever reformatting to get it into shape. [21:40] does read ok? [21:40] So far, the content's fine with me. I'm assuming you spell checked it? [21:41] yeah, the one good thing about open office :D [21:43] Hmm [21:43] Quanta for KDE looks like a promising docbook editor. [21:45] wont mess up formatting? guess I can compare against vi [21:48] I'm installing it now. [21:49] Ah, by the way, for spell checking an XML document, you can use aspell -H check [21:49] and aspell's smart enough to ignore all the XML tags [21:49] ah good to know