[00:00] if i drop the sed, http://paste.ubuntu.com/55440/ [00:00] the script just need APPNAME to run $LIBDIR/$APPNAME [00:06] fta: /usr/bin/x-www-browser => APPNAME=firefox-3.1 [00:06] does x-www-browser link to firefox-3.1? [00:06] does it work if it links to firefox? [00:07] fta: can you run the same check with the patch as of now? [00:07] mine, yes [00:07] would like to see the difference ;) [00:07] and file a bug about that :) ... i am sure i will not be able to remember which corner cases were fixed by this ;) [00:08] do you need $0 later on inside firefox ? [00:08] yes thats the whole point of all these changes :) [00:09] i need a good system $0 [00:09] which isnt versioned if user runs unversioned link (or link of link etc.) [00:09] well not system $0, but good basename $0 ;) [00:10] i already gave up on the hope that $0 might point to the right ;) [00:10] what should I get with 3.1 ? firefox or firefox-3.1 (ie the Name from application.ini) [00:11] i mean, what do you expect inside ff [00:15] well, i'll have a closer look tomorrow [00:16] i need to reboot, new kernel, new nvidia, new everything.. [00:16] i have a bad feeling about this [00:25] as usual, X broke [00:25] :( [00:25] fta: when last rebooted? [00:26] each time there's an update of either the kernel or the nvidia driver, the next reboot / X restart is a nightmare [00:27] yeah. that reminds me, does fglrx exist in intrepid now? [00:27] gasp, the pc speaker is back [00:27] xorg-driver-fglrx | 2:8.532-0ubuntu5 | http://archive.ubuntu.com intrepid/multiverse Packages [00:27] fglrx-installer | 2:8.532-0ubuntu5 | http://archive.ubuntu.com intrepid/multiverse Sources [00:28] fta: i mean ... does it work? [00:29] no idea, my ati box is still running hardy [00:29] i didnt try to use that for a while because people said that we might not have a working solution for release [00:29] but that was 4 weeks back [00:29] maybe there was a miracle ;) [00:32] apps/metacity/general/audible_bell has been (re)set [00:34] fta: anything else we need to fix for next upload? [00:35] https://bugs.edge.launchpad.net/ubuntu/intrepid/+source/firefox-3.0 [00:35] looks like the script is the only targetted fix left [00:35] https://bugs.edge.launchpad.net/ubuntu/intrepid/+source/xulrunner-1.9 [00:36] hmmm [00:36] why is that thing still open? [00:44] fta: there are still a bunch of abrowser transitions things open :( [00:44] https://bugs.edge.launchpad.net/ubuntu/intrepid/+bugs?field.searchtext=&orderby=-importance&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field.omit_dupes=on&field.has_patch=&field.has_no_package= [00:44] i guess thats me? [00:45] i was done with all mine [00:46] Bug 197786 [00:46] Launchpad bug 197786 in prism "Prism apps think I'm offline and no way to change it" [Undecided,Invalid] https://launchpad.net/bugs/197786 [00:46] yeah you can inject the pref ... but that would make NM completely ignored [00:47] most likely we will have a real fix for that in time (i hope) [00:47] the master is bug 191889 [00:47] Launchpad bug 191889 in pidgin "[MASTER] [WORKAROUND] "Offline Mode" feature fails to detect proper online state for networks that are managed outside of network manager." [Undecided,Confirmed] https://launchpad.net/bugs/191889 [00:49] i wish i could update prism too [00:49] and push ff3.1 [00:49] <[reed]> I wish 802.1x worked on 8.10 [00:49] <[reed]> but nooo [00:49] <[reed]> we all have to make sacrifices [00:50] lol [00:50] [reed]: have you tried the module paramter? [00:50] [reed]: does your wired work again? [00:50] <[reed]> wired works again [00:50] <[reed]> thanks to the new kernel [00:50] <[reed]> but still no 802.1x [00:50] ok. thats good news at least ;) [00:50] [reed]: have you tried the module parameter? or didnt i ask you to test that? [00:50] <[reed]> you asked me, and it didn't work [00:50] <[reed]> but I can try again later [00:50] <[reed]> that was two kernels ago [00:50] <[reed]> :p [00:51] [reed]: the disable_hw_scan=1 thing? [00:51] ok [00:53] <[reed]> asac/fta: seen http://lists.gnu.org/archive/html/savannah-hackers-public/2008-10/msg00006.html ? :) [00:55] [reed]: did you write that by yourself? :-P [00:55] [reed]: but anyway. good info there [00:56] [reed]: you should send that to debian too ;) [00:57] "Free software != always secure" :) [01:02] http://ubuntuforums.org/showthread.php?t=942023 [02:22] <[reed]> asac: yes, I wrote it myself! [09:30] https://wiki.ubuntu.com/UbuntuBugDay/20081009 it's moving [09:54] yeah [09:55] strange why they look at 111 New bugs instead of 631 though ;) [10:16] \o/ -> moving everything to inbox-seen :/ [10:16] lets hope that allows me to process mails again :( [10:18] asac: flashupdate \o/ [10:19] yay i found it ;) [10:26] sddoes moz-devscripts need mercurial as a depend? [10:26] asac: s/sddoes/asac does [10:27] # Error: missing dependencies. Please install mercurial at /usr/share/mozilla-devscripts/mozclient/mozclient.pl line 206 [10:27] sebner: yeah ... that reminds me that I should try it ;) [10:27] gnomefreak: apt-get install hg [10:27] gnomefreak: it should be a recommend [10:27] ok thanks [10:27] as not all features depend on hg [10:28] but in general users should have it installed [10:28] fta: ^^ [10:28] hg == not a package [10:28] asac: ^^ like me but no time for now. soon I have my 2nd "perfektionsfahrt" if you know what's that ;) [10:29] damn i changed my password and now i forgot it :( [10:29] cannot install anything right now :( [10:29] asac: wth O_o [10:29] thats bad [10:29] so lets hope that this gnome sessions doesnt crash :/ [10:30] wait there are a few password "cracker" packages in repos [10:30] hmm ... maybe i added my ssh key so i can log in as root [10:30] asac: write a posteit :P [10:30] asac: hg-buildpackage hg-load-dirs? [10:30] gnomefreak: ? [10:31] not sure what you mean ;) [10:31] we dont use hg-buildpackage [10:31] we only use it to produce the orig from hg.mozilla.org [10:31] asac: what hg package? there is no "hg" package in repos [10:31] gnomefreak: yeah its simply called "mercurial" [10:31] thought you already looked for that [10:31] ;) [10:32] ah ok already installed did that when i got there error [10:32] sebner: when i made my fuehrerschein i didnt have a "perfektionsfahrt" ... only "ueberlandsfahrten" [10:32] ;) [10:32] :) error is gone [10:34] cool i can use my virtual box VMs again [10:34] something fixed the vboxdrv for me [10:37] asac: grrr [10:37] too bad ... but my debian VM doesnst start X anymore :/ [10:37] so most likely i have to trash that anyway [10:37] asac: btw, do you know if there exist a openjdk6-plugin for firefox? Couldn't find one and had to install sun java plugin -.- [10:38] jdk == devel right? [10:39] if so i dont remember a plugin for FF to use *jdk [10:41] gnomefreak: well there is a sun-java6-plugin that's the one for firefox und I couldn't find a openjdk6-plugin (openjdk6 is the opensource java) [10:42] sebner: im looking in archives there were ~6 that i saw none said "6" and none were a plugin but still running through archives. im assuming you are still using intrepid? [10:42] gnomefreak: sure ^^ [10:44] gnomefreak: though it's for my parents pc [10:44] hmmmmmm seamonkey-browser i guess that means seamonkey has been separated and now "seamonkey" is a meta [10:44] sebner: it is Ubuntu? or Winblows? [10:45] gnomefreak: ubuntu of course ;) [10:45] maybe it is [10:45] sebner: it is Ubuntu? or Winblows? [10:45] argh [10:45] sry [10:45] icedtea-gcjwebplugin [10:45] that is not jdk ;) [10:46] I searching for a firefox web plugin [10:46] not developing java things ;) [10:46] sun-java6-jdk is it [10:46] nope [10:46] java development kit is for developing java applications [10:47] the right thing is *-plugin [10:47] but I want the opensource java ;) [10:47] sebner: than thats should be good but it should only be pretty much same as sun-java6-plugin [10:48] yes but as I said there is no openjdk6-plugin ;) but I think it's the icedtea-gcjwebplugin [10:48] yeah i noticed [10:48] kk ^^ [10:49] asac_: got your pw? [10:49] doesnt openjdk-6-jre bring in plugin for web? [10:49] nope [10:50] it is the plugin as im seeing [10:51] or atleast a bit different than the plugin [10:51] jre is normally for normal java application and it wasn't working here so I installed the sun java plugin which worked [10:51] fta: you think this is "unsafe": [10:51] for i in /usr/lib/firefox-3.*/.autoreg; do [10:51] touch $i [10:51] done [10:51] ? [10:51] in case there is no firefox install that would break, woulnt it? [10:51] Provides: java-runtime, java2-runtime, java5-runtime, java6-runtime [10:52] that should work *runtime [10:53] gnomefreak: its probably in a separate package (e.g. the plugin) [10:53] asac_: its not unless he uses gcj [10:53] openjdk-6-jre provides the above [10:54] not really sure why jave2 is in that or java5 as there are not "free" [10:55] be assured, i am at least as confused about what is what for java in our repo as you are ;) [10:56] thats why we have doko and/or the java team [10:56] to cause confusion ;)? [10:56] OpenJDK Team [10:56] imo they should clean this up :-D [10:56] to fix the confusion ;) [10:56] yeah. doko is working hard ... and what he gets from sun isnt always that great ;) [10:57] that is true [10:57] but thats just my guess :) ... nothing official [10:57] java has sucked for years IMHO open java doesnt help [10:57] well. open java would have helped [10:57] but it taking so many years, that it doesnt help in the end anymore ;) [10:57] its still not "open" its restricted as i recall [10:58] yeah at least parts of it [10:58] makes no sence unless "parts" is just logo, name [10:58] sense [10:59] well ... i think one part is the crypto part [10:59] that one can be replaced by the bouncycastle free stuff afaik [11:00] i think you are right however i havent read the license since it was first included === asac_ is now known as asac [11:00] me neither ... and i dont want to think about that ;) [11:01] have enough bad dreams without that already ;) [11:01] ill be back updates are killing my proc for some reason [11:15] ill be back she wants breakfast and coffee before she leaves [11:24] so foks [11:24] folks [11:24] wish me luck [11:24] bye bye [11:25] sebner: toi toi toi [11:48] Hello. Someone pointed at the Fennec announcement to me today. I seem to remember seeing that string in some mozilla packaging stuff when I was looking at the common scripts with fta. [11:48] Is the packaging preparation already underway, or is assistance required? (Jaunty goal) [12:37] any good guide on quilt? [12:38] gnomefreak, What are you trying to do? [12:40] gnomefreak, I usually just follow the recipe at https://wiki.ubuntu.com/PackagingGuide/PatchSystems for basic stuff. If you need to do things like insert a patch in the middle of the stack or something, it's a bit more complicated. [12:40] persia, i have a branch for fennec since 0.3 (it's at 0.8+ now) but it needs my build-system patch in xulrunner-1.9, that's why i didn't push it to intrepid [12:41] fta2, OK. I won't try to recruit someone to chase it and come here for integration then. Thanks for the confirmation. [12:42] persia, no assistance required required for the packaging, and i guess for jaunty, it's just a matter of pushing it [12:42] persia, is fennec needed for ubuntu ? or just ubuntu mobile ? [12:44] fta2, Dunno yet. Currently Ubuntu Mobile uses midbrowser, but some people would like to change. A couple of the mobile developers were all excited about today's press release, so I figured I'd stop by and ask. [12:44] (my package needs /usr/lib/xulrunner-devel-1.9.0.*/sdk/build-system.tar.gz from xulrunner-1.9-dev) [12:44] I wouldn't think that Ubuntu Desktop would want it, just because their target devices can usually run full firefox. [12:46] persia, ok, i will start to publish previews in ppa then, so people interested could give it a try. [12:46] I suspect Ubuntu MID will probably want Minimo or MicroB. [12:46] fta2, Don't rush on my account, although if you've previews available, I'm sure you'll have some users. [12:47] I heard a couple complaints about usability (like URL bar and scrolling), so it may make sense to get something to testers early, to help the feedback cycle and get something even more solid for Jaunty. [12:50] i was just forgetting a command but i found it thanks [12:58] persia: we have a tutorial on how to package fennec [12:58] with xulrunner-1.9.1 [12:58] persia: https://wiki.ubuntu.com/MozillaTeam/XulApps/Packaging [12:59] asac, Thanks. I'll pass that on. [13:00] persia: fta is the author of that page iirc [13:01] persia: if mobile team wants to package this, you should really line up with mozillateam [13:01] doing this on your own isnt really a great idea IMO :-D [13:01] Absolutely. That's why I'm here even before there's any decision or discussion on whether to package it :) [13:01] we would certainly be happy to have mobile team members contributing to our packages and would help to improve know-how diffusion i guess [13:01] cool [13:02] I'd much rather coordinate to find out status, and if you guys don't have enough time, try to get you some more resources to help, rather than package it separately. [13:02] yeah. that makes more sense [13:03] most work is probably initial bootstrapping of the packages. but if they are done using our best practices, maintaining them could probably be mostly done by us when bumping ffox and so on [13:03] Right, or if someone is *really* interested, I suppose they'd do best to help you guys, rather than trying to manage it themselves. [13:04] ack [13:12] asac, it's not with xul 1.9.1, it's with 1.9 [13:12] god i hope i fixed this [13:12] ill be back while this builds, [13:13] fta2: oh opk [13:21] sunbird 0 gnomefreak 1 ;) [13:22] atleast fixed patch [13:22] 0.9 package ready? [13:22] building [13:23] but got past the failure [13:23] good [13:23] if all goes well maybe an hour or so till its done and shortly after it will hit my PPA [13:24] just wish i could fix my sat. dish so i can watch tv again [13:39] is there a problem with FF 3.1? its not showing up in preferred apps [13:39] it just lists firefox [13:41] asac: sunbird wont be ready for a while its a failure now [13:42] ok fixed the prefered apps just have to use custom command [13:46] ok [13:46] asac: fta2 what the hell am i missing? http://pastebin.mozilla.org/551199 [13:46] am i missing a flag in rules? [13:47] gnomefreak: yeah it lacks a ptch [13:47] gnomefreak: look in the tbird package [13:47] the patch should be there [13:47] look at seamonkey-1.1.dev, there's a patch called debian/patches/11_fix_ftbfs_with_fontconfig.patch [13:48] or there ;) [13:49] looking [14:36] please remind me why we use quilt [14:37] oh yeah because i finally got used to dpatch than we switch to quilt [14:38] * gnomefreak crosses fingers and goes outside [14:42] ok damnit i did everything right and its failing after applying the new patch [14:48] imho, quilt is much better, cleaner, faster [14:48] asac: fta2 http://pastebin.mozilla.org/551239 would trailing whitespace in configure-autoconf2-13-reconfigure.patch cause this failure? [14:48] fta2: dont forget a pain in the ass ;) [14:49] well, i won't call quilt new/add/push/pull/refresh a pain in the ass [14:49] i like it [14:50] # [14:50] cd build-tree/mozilla && QUILT_PATCHES=/home/gnomefreak/lightning-sunbird-builds/work/build-area/lightning-sunbird-0.9~cvs+nobinonly/debian/patches quilt --quiltrc /dev/null push -a || test $? = 2 [14:50] # [14:50] Patch bzXXX_ftbfs_fontconfig.patch does not exist [14:50] yeah [14:50] you probably forgot the bzr add [14:50] gnomefreak: bzr add debian/patches/bzXXX_ftbfs_fontconfig.patch [14:50] forgot add [14:50] ok brb trying it [14:59] ok that fixed it thanks i knew i forgot something :) [15:06] * gnomefreak looking over error and i dont see anything that points to forgetting bzr add [15:07] maybe not all patches being there [15:24] asac, nspr/nss.head reverted [16:01] fta2: yeah ;) [16:01] ifupdown is really painful [16:01] i rarely say that i hate something. but touching that is really pain imo [16:02] i really cant believe that we have such an esoteric package as the foundation of ubuntu/debian [16:02] networking [16:12] next try [16:12] asac : To answer your question for bug 259214, it is reproducible on a fresh intrepid installation. It is due to the fact that ethernet connexions use "system settings" so if you change the settings to static IP, the computer will use DHCP on the next reboot [16:12] Launchpad bug 259214 in network-manager "wired connection settings are lost after reboot" [Medium,Confirmed] https://launchpad.net/bugs/259214 [16:13] saivann: its fixed in upload i did a few minutes ago [16:13] asac : Really, nice to know, thanks! [16:13] saivann: please verify and let me know if its not fixed ffor you :) [16:13] but i hope it is [16:13] well i verified that it is here [16:13] asac : Which upload version is that? [16:13] there might still be rough edges [16:13] saivann: svn20081008... [16:13] asac : Thank you, I'll give feedback [16:13] saivann: https://edge.launchpad.net/ubuntu/intrepid/+source/network-manager/0.7~~svn20081008t224042-0ubuntu1 [16:15] asac : I'll have to wait for AMD64 build, thanks [16:16] asac: i'm quite confused on how to patch mozilla [16:16] nxvl: just out of curiosity, is this for a security update? [16:16] nxvl: we use quilt and branches [16:16] asac: and i'm not sure if it is because i don't know bzr or that everything is tarballed :P [16:16] jdstrand: yup [16:16] nxvl: please state first what you are working on :) [16:16] jdstrand: well, stuff like [16:16] nxvl: those kind of patches need to go through asac first [16:17] asac: due to our relationship with mozilla [16:17] nxvl: ^ [16:17] ok [16:18] nxvl: whats this about? [16:34] * asac reboots and hopes for the best [17:05] ;) [17:05] uploading sunbird [17:08] asac: does jazzva maintain foxyproxy? [17:09] bdmurray: i think he built it last but let me check [17:09] bdmurray: he is the "extensions" lead ... but its team maintained in the end [17:10] upload+browser makes browser not work [17:11] Hmm, I took the liberty of subscribing him to bug 279466 [17:11] Launchpad bug 279466 in foxyproxy "Error writing settings file: file:///usr/lib/firefox-3.0.3/foxyproxy.xml" [High,Confirmed] https://launchpad.net/bugs/279466 [17:12] Is he on irc much? [17:13] bdmurray: he usually is here every night. but then he is on a holiday trip till next week atm :/ [17:14] Ah, good for him! ;) [17:14] when did bzr-builddeb start making a dir backup.bzr in the top level dir? [17:17] asac : I tried network-manager svn20081008 in a i386 VirtualBox installation (not completely up-to-date) and the problem persist, if I try to change the system setting from DHCP to static, it fall back to DHCP after a reboot [17:19] saivann have you restarted everything after the update? [17:19] gnomefreak: thats not builddeb, but bzr upgrade [17:19] ah [17:21] asac : Yes [17:22] asac : I'll still try the AMD64 build on a completely up-to-date computer when it will get in the repositories [17:24] it works but i get javascript errors "window not defined [17:24] " [17:24] saivann: soudns wierd that it doesnt work for you. not sure what this is up to. it will certainly fail if you have policykit issues [17:24] saivann: do you get at least asked for a passphrase? [17:24] * asac reboot again [17:25] asac : Yes I guess so, that's why I'll try with a up-to-date intrepid [17:25] asac : No, not asked for the passphrase [17:28] asac: pushing 0.9 to branch if you need anything with it [17:28] ow iceowl [17:30] asac: do you have an up-to-date iceowl branch? mines a bit old [17:32] i still get bzr upgrade warnings but i have latest [17:32] bzr: ERROR: The branch format Bazaar-NG meta directory, format 1 is already at the most recent format [17:33] ok its building in my PPA atm i will be back later today i still have to fix my sat. [17:34] oops forgot to test smart gui for sticking at ldconf [17:37] gnomefreak: isnt the iceowl branch in mozillateam up to date? [17:38] gnomefreak: you have to upgrade the remote branch too [17:38] not only your local one [17:38] you can do that with bzr upgrade URL [17:38] but take care to not abort that [17:38] it takes quite a while [17:48] fta_: there? [17:48] can you paste your /var/run/network/ifstate please :) [17:51] asac: I've tested bug 277063 and it doesn't seem fixed to me [17:51] Launchpad bug 277063 in network-manager "Wired and wireless connection to the same router cause confusion" [High,Fix released] https://launchpad.net/bugs/277063 [17:52] bdmurray: does the order of the interfaces still change? [17:53] asac: what do you mean? [17:54] bdmurray: look at your routing table [17:54] before you have a certain order (e.g. eth0 eth1) ... after its (eth1 eth0) in the routing table [17:54] i am quite sure this has something to do with it [17:55] asac: I'll have to check again when first booted. [18:20] is there a version-specific xulrunner-addons i should use for 1.9-only or 1.8-only plugins [18:30] flash is soooo f*cking b0rken. more than normally xD [18:31] sebner, i can never tell [18:31] directhex: ^^ [18:31] sebner, right now i'm trying to work out the best way to let some foreigners have proxy access inside the uk, to help debug a plugin [18:31] directhex: O_o [18:33] sebner, you're aware of ITV catchup? [18:33] directhex: when you'd tell me what ITV is xD [18:33] sebner, one of our main TV networks [18:33] ahhh [18:34] ITV catchup is their streaming browser-based service, which allows you to watch any of the shows they air, as long as you come from a british ip [18:34] except... the implementation is sick. twisted and sick [18:35] * sebner needs a proxy xD [18:35] silverlight? [18:35] sebner, if it was *just* silverlight, it wouldn't be sick [18:35] sebner, it's silverflashlight. [18:36] sebner, watch http://retro.apebox.org/moon/moon-itvcatchup.ogg [18:36] * sebner is scared now [18:36] * sebner hide [18:36] s [18:36] xd [18:37] directhex: omg xD [18:37] sebner, sick, isn't it :) [18:38] sebner, it's a very high profile site, i want it working in ubuntu. so i need to find a way to let the devs be british, for debugging [18:41] directhex: you're a nice guy ^^ [19:00] directhex: there should be no 1.8 only plugin [19:01] asac, so there's no mechanism for it? [19:01] hello, [19:01] i'm trying to get started with triaging. I have some time at the moment, so i thought, why not start with the firefox hug day. (since firefox is part of my most frequently used software ;) ) [19:01] So i was looking at bug #280101 to which i commented. What steps should i do to complete this triage. [19:01] (https://bugs.launchpad.net/ubuntu/+source/firefox-3.0/+bug/280101) [19:01] Launchpad bug 280101 in firefox-3.0 "Firefox search bar points to wrong language wikipedia" [Undecided,New] https://launchpad.net/bugs/280101 [19:01] Launchpad bug 280101 in firefox-3.0 "Firefox search bar points to wrong language wikipedia" [Undecided,New] [19:11] asac: yes, the route ordering does change [19:20] bdmurray: ok ... and you have two net routes right? [19:20] (e.g. not one host route + one net route) [19:20] directhex: well. if your plugin only uses npapi it should work everywhere [19:21] directhex: if it uses XPCOM you need to link against the dependent glue of 1.8 ... then you can use the binary on 1.8 and 1.9 [19:22] asac: I have two routes to the same subnet if that is what you mean. [19:22] yeah thats dumb [19:23] bdmurray: i will test that now :) [19:24] asac, there's a core plugin, which is NPAPI, and some per-engine glue, needed for advanced bits & pieces (e.g. allowing the plugin access to the DOM). currently there are two bits of glue, ff2bridge (1.8) and ff3bridge (1.9) [19:26] bdmurray: i have almost the same setup. doesnt happen here [19:26] bdmurray: what i do: [19:26] 1. up two devices that have the same subnet [19:26] 2. modprobe -v -r forcedeth (e.g. my wired driver) [19:26] 3. check that the routing was removed for the wired driver [19:26] 4. modprobe -v forcedeth [19:27] asac, it's not a huge problem for me as long as i'm only targetting debian (since only 1.9 exists in debian), but... [19:27] 5. check that the routing was happily added again [19:27] bdmurray: ^^ for me ping never looses a package when doing that [19:28] bdmurray: can you reproduce that by removing your module too instead of unplugging? [19:30] bdmurray: unplugged and plugged again [19:30] worked well too [19:30] bdmurray: maybe your dns server changes and that causes issues? [19:30] bdmurray: do you get the same dns through wired and wireless? [19:32] bdmurray: hmm do you have different metrics for both devices? [19:32] bdmurray: http://paste.ubuntu.com/55743/ [19:32] bdmurray: so wlan and eth have metric 2 and 1 respectively [19:33] asac: no all the metrics are 0 except for 169.254.0.0 [19:34] bdmurray: yeah. cant tell then thats the latest code :/ it shouldnt be 0 [19:34] bdmurray: did you setup this stuff in connection editor? [19:35] bdmurray: please paste the output gconftool-2 -R /system/networking [19:39] asac: http://paste.ubuntu.com/55746/ [19:40] bdmurray: so those are not system connections? [19:42] asac: what do you mean by system? [19:43] bdmurray: in connection editor "system" [19:43] e.g. not bound to your user account [19:43] bdmurray: "system setting" ... thats a check box [19:43] which means that those connections can be started even before you logged in [19:44] asac: the one's in pastebin are not however eth0 is a system connection [19:56] bdmurray: ok so you have one system and the wireless is user [19:56] can you try to make both system or both user? [19:57] asac: okay, fwiw I've never used the connection editor before so I'm uncertain how it got this way [19:58] bdmurray: are you using managed=true in /etc/NetworkManager/nm-system-settings.conf? [19:58] if so then your eth0 is most likely configured in ENI [19:58] which would make that a read-only system config [19:58] err connection ;) [19:59] asac: okay managed is false and eth0 is in ENI. However, wasn't ENI the right way to go a release ago? [20:00] bdmurray: well. it means that eth0 isnt managed by NM [20:00] which might explain why NM doesnt know about the metric [20:00] bdmurray: please set it to =true [20:00] reboot [20:00] and see if you still reproduce it [20:00] or if your metric is fine [20:01] my guess is that it works with "managed=true" [20:01] geez, reboot really? [20:01] bdmurray: well ... stop NM [20:01] then sudo killall nm-system-settings [20:01] then start NM [20:01] is the other way [20:01] okay, but nm-system-settings.conf is new for Intrepid right? [20:01] bdmurray: yes [20:01] we use the same by default that we used before [20:01] e.g. managed=false -> meaning that devices configured there are not managed by NM [20:02] -> which causes issues as you can see now ;) [20:02] Okay so this should be release noted then right? [20:02] hopefully until jaunty we can flip the default to managed=true and say good bye to ifupdown :) [20:02] bdmurray: no i dont think so [20:02] bdmurray: the behaviour didnt change [20:02] if you dont do anything your /etc/network/interfaces devices will not be managed by NM [20:03] thats the behaviour we had before [20:03] i will blogg about it ;) [20:03] i also sent an announcement to ubuntu-devel [20:03] -> RFC: [20:04] bdmurray: you can also read some details in README.Debian [20:04] asac: I'm looking at that mail now [20:04] cool [20:04] bdmurray: the README.Debian might be more comprehensible [20:04] :) [20:04] asac: Why did it used to work in Hardy though? [20:04] bdmurray: by pure accident then [20:05] bdmurray: you would need to observe the behaviour in hardy to tell for sure [20:05] Having eth0 unmanged and wireless managed seemed to be fine [20:05] bdmurray: first try to verify that this is the bug please :) [20:06] if it is, we can probably fix it [20:10] asac: I've set managed=true and rebooted and see no change in behavior [20:14] bdmurray: no metric set on wifi connection? [20:14] bdmurray: does grep NULL /var/log/syslog give you anything? [20:15] just CPU stuff [20:15] bdmurray: really. the metric in route -n should be 1 for ETHERNET and 2 for WIFI [20:15] if that isnt the case something is seriously broken on your system ;) [20:16] my system! hmmm [20:16] i looked at the code. and there is really no other way ;) [20:17] okay, I'll try and find what I broke then [20:18] bdmurray: rebooting helps ;) [20:18] bdmurray: do you see the eth0 connection now in NM? [20:18] I'm rebooting again - give me a second [20:26] asac: I see "Auth eth0" in nm now - it still seems to be a system setting [20:28] bdmurray: thats ok [20:28] Auth? [20:28] still no metrics for the interfaces [20:28] bdmurray: http://paste.ubuntu.com/55758/ [20:28] Auto - my bad [20:29] bdmurray: is there a ifupdown (eth0) connection now? [20:29] bdmurray: you need to apply that patch. maybe your find_route doesnt work. thats the only reason i see that could hinder NM from setting a proper priority on your device [20:38] bdmurray: its really confusing that you always switch systems when testing this [20:38] sometimes you have wlan0 and eth0 ... then eth1 and eth0 ;) [20:39] sometimes you have nothing in /etc/network/interfaces (e.g. like in the bug log) ... sometimes you have configs there [20:46] asac: I've always used the same system - the description, which has wlan0, is from the original reporter - not me. [20:48] bdmurray: anyway. i could reproduce the bug i think with the hybrid approach (e.g. one unmanaged and a managed connection) [20:48] bdmurray: however, the bug is particularly caused by a metric on the NM managed device [20:48] if you have 0 everywhere on the metric then i cannot say [20:48] its definitly another bug [20:49] bdmurray: the unmanaged default would always have a default route with a metric of "100" [20:49] thats the way ifupdown handles that [20:49] also if i have one unmanaged and one mangaed device i will always have two default routes [20:49] but you all the time have only one [20:51] bdmurray: http://paste.ubuntu.com/55770/ [20:51] thats the routing table that i get ... and which causes packet loss [20:52] the default route of wlan has a higher priority than the default route of eht0 (which comes from eni) [20:52] but the net route of eth0 has ahigher prio than the wlan0 route [20:52] which explains why it doesnt work [20:52] packets want to go the wlan default route ... but try to do that through the eth net [20:53] and i get packet loss [20:53] anyway. whatever i do, the metric will be 2 for wireless devices and 1 for wired (given they are not unamanged) [20:53] and i know that this was a bug fixed in this last upload :) [20:54] which is the 8th oct build ;) [20:55] i will think about the hybrid bug ... most likely i have to patch ifupdown again to use a default metric of 10 or so on nets that dont have a metric :( === fta_ is now known as fta [21:26] the hug day seems staled :P [21:26] * sebner hugs fta :P [21:26] lol [21:36] no hugs for me? :( [21:39] * sebner also hugs the poor directhex [21:39] well you are my mono hero so this is a true hug ^^ [21:40] sebner, meebey's the hero. i'm more like robin to his batman, if anything [21:40] directhex: sure but he's not here ;) btw, good progress with mono2? [21:41] sebner, i'm not getting very involved with that right now, mono is meebey's baby. mainly been working on moonlight and xsp [21:41] directhex: well, I refered to him ::P [21:41] sebner, but i understand he's making progress when he finds time. there were licensing problems to solve first and foremost [21:41] grrrrrr [21:41] licensing [21:41] yuk [21:42] licensing is important. it's patents that can take a flying leap! [21:42] fta: https://lists.ubuntu.com/archives/ubuntu-devel/2008-October/026676.html [21:44] however folks, I wish everybody a good night (if they are in nearly the same timezone as I am ^^) [21:46] "A debdiff is just a kind of patch and any Ubuntu developer who can't figure out how to apply some slightly differently-formatted patch in a matter of seconds shouldn't be an Ubuntu developer." [21:46] eheh [21:47] worse than debdiffs are debdiffs which ignore a package's patchsys [21:47] i can spend ages sometimes merging ubuntu changes that were poorly applied outside the scope of the existing patchsys [21:47] fta: yeah ... i was mostly refering to the statements about MOTUs and debdiffs and sending people back and forth for nothing ;) [21:50] sebner, anyway, you know i only do backports for LTS, right? some other mug will need to offer mono2 packages for intrepid [22:00] asac, i agree with most of it. yet, i don't always like to send patches upstream.. for some projects, i don't want to have to create an account there and deal with a complex/weird/unknown/alien workflow, i expect someone else to take care of that, but i understand that probably noone will :( [22:12] asac: what more needs to happen with bug 279083? [22:12] Launchpad bug 279083 in firefox-3.0 "firefox 3.0.3 on intrepid reports 3.0.1 as user agent" [Undecided,Confirmed] https://launchpad.net/bugs/279083 [22:16] bdmurray, i think i already fixed that [22:17] fta: is there somewhere I could check? I noticed some of firefox is in bzr [22:21] bdmurray, hm, i fixed it for sure in abrowser, let me check [22:22] it's in a branch, not released yet [22:24] fta: okay, should I do something to the bug or can it get closed by a changelog? [22:25] bdmurray, please assign it to me, i'll take care of it [22:26] okay, thanks!