[00:01] lol: http://www.rumblingedge.com/2009/03/18/thunderbird-20021-released/ [00:01] "Thunderbird 2.0.0.21 (not yet) Released" - what kind of news is that? [00:02] it's just an activity report, like you do to your boss [00:03] fta: maybe. [00:03] most likely they thought it was announced and changed the title after they noticed it was wrong ;) [00:03] at least the link is still reading: thunderbird-20021-released [00:04] "Mozilla Thunderbird 2.0.0.21 has (not yet) been released. Release notes are available. This post lists the improvements in Thunderbird 2.0.0.21 over 2.0.0.19. This list encompasses almost every single known fix that went into this release. Do check out the known issues as well." [00:05] I read a lot of "xxx (not yet) Released" blog entries in the past [00:06] jtv: We were unable to import your translations because you did not update [00:06] the time stamp in its header to state when you added your translations. [00:07] jtv: i think thats a problem. we should either take care that this happens automgically during build [00:07] or make rosetta take them regardless [00:09] fta: i really think we should spot why isdnutils is in ia32 and take care that its not in there anymore [00:09] Debian bug 479662 in ia32-libs "ia32-libs: libcrypto is broken, causes Wine FTBFS" [Serious,Closed] http://bugs.debian.org/479662 [00:09] Debian bug 479662 in ia32-libs "ia32-libs: libcrypto is broken, causes Wine FTBFS" [Serious,Closed] http://bugs.debian.org/479662 [00:09] it's mentioned in there [00:10] wtf? http://paste.ubuntu.com/133323/ [00:11] fta: thtas interesting [00:11] fta: i would think that the tarballs are just out of sync a bit [00:12] maybe bad luck [00:12] either get a new xul or firefox [00:12] i would think you should get a new one for the currently older one ;) [00:12] well, i wanted to test the transition, i'll just respin the bit [00:12] bot [00:12] obviously, my xul was a bit older than ff [00:13] not really obvious from the paste. unless you know the signature of run out of your head [00:13] could as well have been firefox older [00:14] fta: is 3.1-> 3.5 done? [00:14] both [00:14] committed [00:14] fta: did you rename branches yet or do i need just update? [00:15] ok i seem to have gotten 3.5 things [00:15] not yet, everything else should be ok [00:16] whats the story about libstartup-notification0-dev ? [00:16] on lpia, it's not dragged in automatically by something else [00:20] fta: is that linked into mozgnome? [00:20] started a few days ago, no idea why, the commit fixed it [00:23] we should ask mobile team about that [00:23] maybe it was something transitional even [00:29] http://www.mozillamessaging.com/en-US/thunderbird/ ... so is it out or not ;)? [00:31] 21 ? yes [00:31] THUNDERBIRD_2_0_0_21_RELEASE => 03/03 [00:33] asac, lol, there *is* a ia32-libs-dev in the rules file, but it's only on ia64, not amd64 [00:37] Note that ia32-libs-dev was added specifically to allow building of an amd64 [00:37] biarch compiler suite, it is not intended that anyone use this package to do [00:37] "real" development work for ia32 systems! [00:37] hm [00:39] i always found this odd ;) [00:39] fta: maybe check what debian ships [00:39] and how far we have diverged ;) [07:28] asac: we're planning to change that this cycle, so that at least an _unchanged_ date will be accepted. (An actually _older_ date should still be rejected of course). [09:13] fta: a tag doesnt mean its out ;) [09:46] asac: ping [09:46] around? [09:46] only a bit [09:46] * asac gets coffee [09:46] that --print-reply [09:46] but yes. i am here because its business hours ;) [09:47] didn't work until reboot [09:47] after that at least suspend would work [09:47] haven't tried hibernate while plugged in [09:48] BUGabundo: i uploaded a fixed pm-utils [09:48] try that [09:48] so when you get the update [09:54] ok [09:54] running NM now [09:54] ahhh [09:54] UM [11:22] ok seems that tbird 2.0.0.21 is finally finally released ;) [11:22] http://www.rumblingedge.com/2009/03/18/thunderbird-20021-released/ [12:22] asac, [01:31] THUNDERBIRD_2_0_0_21_RELEASE => 03/03 [12:39] fta: what does that mean? [12:45] that 2.0.0.21 has been tagged 2 weeks ago [12:56] asac, so, i want to update ia32-lib, but the isdn thing is a blocker, now what? [12:58] fix isdn thing [12:58] is the simple answer [12:58] finding someone to fix it might be more difficult ;) [12:59] it's in main [13:00] fta: its on its way down [13:00] fta: you can just do it and either we demote it right away or upload it somehow [13:00] or i can cheat and use the previous build [13:01] providing the debs are still somewhere [13:15] fta: if that works ... of course next time it will reappear. but if we want something before beta, why not [13:15] (note i am not ia32libs guy ... you could ping \sh who did that in the past too) [14:12] http://crad-vador.net/blog/dotclear/index.php?post/2009/03/18/Pr%C3%A9-alpha-de-chromium-pour-Ubuntu [14:12] :) [14:13] http://www.stefanoforenza.com/chromium-on-ubuntu-how-to/ [14:39] asac, what do you think about bug 190227? i'm quite reluctant to patch binaries with sed :P [14:39] Launchpad bug 190227 in ia32-libs "ia32 apps look for libs on the wrong place" [Low,Confirmed] https://launchpad.net/bugs/190227 [14:40] for some reason i thought we dropped ai32-libs but i may be thinking of something else or we used a diff source [14:42] fta: no english link for chrmium-pour... [14:42] gnomefreak, read the 2nd [14:42] k [14:44] i pasted the one in french because we already had one in german and a few in chinese [14:46] ah. [14:59] fta: doesnt mozilla make chromium [14:59] lol, no [14:59] google does [14:59] fta: and your jaunty build failed [15:00] i know, that's why i'm working on fixing ia32-libs [15:00] fta: [15:14] fta: let me know when you have a 64bits to test [15:19] https://edge.launchpad.net/+builds :) [15:19] trying to complete that !#~?@ versioning transition [15:20] fta: ok latest fontconfig-config will fix it ... you have to downgrade though as it only cleans up if you have lt ..ubuntu12~ [15:20] what a mess [15:20] its really interesting how much cruft can accumulate over time [15:20] i mean. debian has zero patches ;) [15:20] because debian guy is Keith Packard [15:21] who we should assume is always right ;) [15:22] fta: any time estimated? [15:22] hate keep getting those warnings [15:23] * BUGabundo reminds: make FF profile backups [15:23] if it works, ~1h or 2 [15:25] http://www.sofaraway.org/ubuntu/tmp/PPA.png [15:34] http://www.vivanno.com/aggregator/?p=17732 [15:34] lol, the new buzz === Ampelbein is now known as ampelbein [15:35] we'll see if popcon reacts to that [15:35] 9629 cxchromium 3041 361 2462 159 59 (Unknown) [15:35] 22949 chromium-browser 314 16 0 298 0 (Unknown) [15:35] 54196 ia32-libs-chromium-browser 5 0 0 0 5 (Unknown) [15:35] hmm [15:35] catching up on the wine package.. [15:37] 23590 firefox-3.2 287 22 84 181 0 (Unknown) [15:51] fta: asac: flash 64 bits on FF 3.1 using youtube FullScreen shows up on laptop display and not external LCD where FF is opened! [15:52] bug is on FF, flash, nvidia driver, or you tube player? [15:52] i have no time right now. still in beta rush (something new came up) [15:52] okay [15:54] anyone care to fix firefox-gnome-support-3.1? [15:54] touch: cannot touch `/usr/lib/firefox-3.1*/.autoreg': No such file or directory [15:54] dpkg: error processing firefox-3.1-gnome-support (--configure): subprocess post-installation script returned error exit status 1 [16:00] gnomefreak, it's part of the transition, currently building<;;; [16:00] fta_: thanks [16:33] asac, when i resume on my netbook, it's not able to reconnect to wifi. after a while i get a popup asking me to enter a password for wpa something, if i unhide the password, it's a hash, not my actual password [16:36] fta: right. thats a known usability issue [16:36] it reminds me that i should poke upstream abou tit [16:36] fta: but its ok in general [16:36] it doesn't work for me [16:37] fta: well. the password probably works. you just cannot connect after resume -> driver issue [16:37] what should i enter there? my passphrase or my password? [16:37] fta: try to unload the driver together with the max80211 and cfg80211 modules [16:37] fta: it doesnt matter. you can either use the hex key thats alrewady prefilled [16:37] fta: or enter your passphrase/password [16:37] both are complex, and type that on a netbook is a pain [16:37] fta: you dont need to retype them [16:37] fta: they are remembered [16:38] fta: unless you know you mistypes you shouldnt delete it when it pops up [16:38] just dismiss or press ok [16:38] the ath5k_pci process is sucking all the cpu :( [16:38] fta: yeah. driver bug [16:38] fta: try to remove all modules [16:40] i reentered my passphrase, nada [16:40] fta: oh todays -gnome-support failed because it tried to touch /usr/lib/firefox-3.1*/.autoreg [16:40] maybe you have forgotten to change that [16:41] no, it's a 2~3 days old ff3.1 [16:41] fta: if you are in a state where it fails to connect it will not help to reenter password [16:41] it's fixed in the branch [16:41] its just borked [16:41] fta: hmm. ok [16:41] fta: just because i got that today ;) [16:41] ppa is building [16:41] amd64 is ready [16:42] k [16:42] i will try a bit later today then [16:43] fta: 64bits ready of what? [16:47] ff 3.1->3.5 transition [16:47] well, i hope so [16:49] ok [16:49] updating now [16:49] oh wait.. im on daily [16:49] can you spin that too fta? [16:50] that's the one [16:50] upgrading now [16:50] should I close FF first? [16:51] as you want [16:53] from my experience FF upgrades with it running don't go so well [16:53] it is easilly repoducble with ibex [16:53] fresh install, open FF, upgrade system [16:54] the upgrade itself should run fine, the running ff otoh will be confused [16:54] yeah [16:55] updates sources and aint seeing it [16:55] but I have xulrunner-1.9.1-dbg updates [16:55] should not matter [16:55] will I need to change my default browser EVERYWHERE? [16:55] its on gnome and KDE settings [16:56] 1.9.2 is broken, i fixed it but i will let the bot catch up, it starts at 7pm [16:56] eh? [16:56] oh, if you have firefox-3.x hardcoded, yes [16:56] touch: cannot touch `/usr/lib/firefox-3.1*/.autoreg': No such file or directory [16:57] getting hit by gnomefreak bug too [16:57] hm, which version? [16:57] BUGabundo: i had that today too ;) [16:57] fta claimed its fixed since 2 days ;) [16:57] eheh [16:57] slow repos [16:57] fixed but not built [16:57] its due to transition as stated above :) [16:57] I read it [16:57] The following packages have unmet dependencies: [16:57] libfontconfig1: Depends: fontconfig-config (= 2.6.0-1ubuntu9) but 2.6.0-1ubuntu12 is installed [16:57] downgrading that [16:58] * BUGabundo gets is gun out! grrrrrrrrr [16:59] audio is broken [16:59] * c0p3rn1c ducks [16:59] again [16:59] after a single day 100% perfect [16:59] BUGabundo: what kind of audio system do you have ? [17:00] my pulseaudo recently also stopped working, I'm using also atm [17:00] its pulseaudio (0.9.14-0ubuntu14) [17:00] ALSA* [17:01] I was soundless for 2 months [17:01] I used to have pulseaudio :( [17:01] something is wrong here :( good news not mozilla related [17:01] this morning updates fixed it [17:01] and now its broken again [17:01] damn it [17:02] gnomefreak: I have some more good news, I have discoverd(through a forum) a very easy way to solve most of 64 bit flashplugin-nonfree issues [17:03] it is all described at http://ubuntuforums.org/showthread.php?t=772490&page=64 [17:03] asac, http://paste.ubuntu.com/133764/ (from etckeeper) [17:03] c0p3rn1c: i was thinking of packaging that for PPA but i cant test it so not sure if i will or not [17:03] fta: looks good, right? [17:03] gnomefreak: well it solves alot of bugs so I would surely recommend it [17:03] fta: e.g. first it goes missing. then the cleanup transition code kicks in and puts an end to it [17:04] gnomefreak: flash runs olmost perfect on my system atm [17:04] fta: also we want -slight and not medium ... which also seems to be right [17:05] great. [17:05] c0p3rn1c: i will most likely do it this weekend if im going to. today is just way too busy for me [17:06] gnomefreak: ping me when you do [17:06] archive or PPA? [17:06] gnomefreak: ok great! Many users will affected by this, surely I don't need to remind you that flash has become a vital internet plugin [17:06] BUGabundo: ok [17:06] c0p3rn1c: flash is nothing more than a piece of shit app [17:07] gnomefreak: lol, maybe, but it's a widely used piece of shit app ;-) [17:08] c0p3rn1c: if adobe ever starts supporting it maybe than it will be worth it [17:09] ok brb trying to break system [17:09] gnomefreak: yeah, I know, linux isn't receiving the love it should from alot of corporations out there and adobe is surely one of them [17:12] asac, nada, wifi's dead on my netbook, rmmod of the drivers, even reboot [17:12] asac, now i have the ath5k phy0 logs i pasted a few days ago [17:16] BUGabundo, the latest valid build of chromium amd64 + the ia32-libs from my PPA works fine, just tested it on my laptop [17:16] ok [17:16] but no FF3.5/6 yet here [17:16] one thing at a time please [17:17] thouht you said it was built [17:18] yep, but i haven't tested it myself, i'm currently upgrading my desktop [17:18] and fighting with my netbook for regain wifi [17:19] fta: post modinfo DRIVER [17:19] can't post anything, no network [17:19] asac: FYI that --reply-print seems to be working 100% now [17:19] sometimes poking a few parameters liks "disable-all-the-magic-a-driver-should-never-do-anywa" help [17:19] fta: what module is that? [17:20] just ath5k or something else on top? [17:20] ath5k. [17:20] donno [17:20] fta: anything that depends on ath5k? [17:20] fta: lsmod [17:20] if that has a driver on the right side [17:20] it depends on it [17:20] or the other way ;) [17:20] asac: many user on #+1 asking about those atheros [17:21] I at least counted 5 [17:21] making fta 6 [17:21] ok, it's working after a cold restart [17:22] fta: chrome: [17672:17672:24111581544:ERROR:browser/tab_contents/web_contents_view_gtk.cc(66)] Not implemented reached in virtual void WebContentsViewGtk::CreateView() [17:23] copy paste from it won't work either [17:23] mouse triple click doesn't select the all line [17:23] xeee 3 sec, 3 bug [17:23] better close it now [17:23] well, you should see tons of those in the term, forget about them [17:24] ok [17:24] but it is faster!!! [17:24] way faster then my FF3.1 with 53 addongs [17:24] but doing far less [17:26] fta: i would think trying the daily vanilla kernels our kernel team provides might be a thing to do [17:27] at least if its fixed there we can check what to cherry pick [17:27] why dont we have port scanners in repos? i need to test a windows firewall and i dont want to do it by hand [17:28] asac, i need this thing working for this week-end [17:28] gnomefreak: nmap? [17:28] asac: nmap will scan ports? [17:28] fta: well. if vanilla works ;) .... [17:28] the upstream RCs are usually not that bad [17:28] so you dont need to try daily [17:28] just RCs [17:28] ah it does :) [17:28] gnomefreak: yes it does [17:28] gnomefreak: thats the main tool ;) [17:29] gnomefreak: it can even determine uptime of systems ;) [17:29] welll ... at least for those that dont do anything against it (like some bsds) [17:29] cool thanks ill look into it [17:30] gnomefreak: http://www.jwsdot.com/tuptune/ ;) [17:30] asac: thanks [17:30] that was real fun ;) [17:30] "With TUPTUNE (Tcp UPtime TUNE) you can easily spoof your uptime detected by network scanners like nmap or netcraft. " [17:32] asac: http://paste.ubuntu.com/133781/ [17:33] yeah its ath5k [17:33] fta_: are there -backport-modules in jaunty? [17:33] thanks, i can read :) [17:33] try that if thats the case [17:33] otherwise really try the vanicall 2.6.29 RCs at least [17:35] gnomefreak: we do! nmap -A SERVER or IPrange [17:36] fta: you can also try to blacklist ath5k and use the madwifi dogshit ;) [17:36] thats ath_pci [17:36] i think [17:36] but i guess it doesnt deal with your card [17:37] fta 3.1->3.5 is here [17:41] did it work? [17:43] hm, last update gave me a new default wallpaper [17:45] fta: do you have the dmesg/syslog kernel messages you get ? [17:47] fta: see -devel. so LBM which keeps up with upstream pace [18:00] asac, http://paste.ubuntu.com/133801/ bunch of stuff: power on, all fine, suspend/resume, no more wifi, reboot, still no wifi, halt, wait, power on, wifi's back [18:02] gym time [18:03] fta: have you seen -devel discussion? [18:04] fta: so LBM is really the package you should try [18:04] ;) [18:04] it seems to be more or less what goes into .29 [18:05] http://linux.derkeiler.com/Mailing-Lists/Kernel/2008-10/msg12772.html [18:41] asac, ff3.1 left some files behind in /etc. I should add a postinst too :P [18:43] fta: so are the files removed from your dpkg database? [18:43] fta: dpkg-query ? [18:43] yes [18:43] fta: so it worked. [18:43] fta: not sure. i am a bit scared doing the same for a good package like firefox ;) [18:44] http://paste.ubuntu.com/133838/ [18:45] fta: why do we ship all that stuff there? [18:45] it was already there in ff2, no? [18:45] yeah makes sense [18:46] fta: in the end i dont care for inter 3.1 transitions [18:46] fta: if we want to create infrastructure code we can reuse for all kind of conffile things [18:46] we should do it right [18:46] otherwise you can do rm_conffile in postinst [18:47] a bunch of people installed it from universe [18:47] if you do it in preinst, ensure that you do the proper abort case (e.g. copy -back stuff back) [18:47] 12210 firefox-3.1 1712 260 767 685 0 (Unknown) [18:48] 20620 chromium-browser 448 19 0 429 0 (Unknown) [18:48] d'oh [18:48] fta: we didnt send out a call for testing yet [18:48] 3.1 is mostly a covert thing still [18:48] folks dont know [18:48] [16:35] 22949 chromium-browser 314 16 0 298 0 (Unknown) [18:48] unless the find it by accident [18:49] yeah. growth is healthy ;) [18:49] fta: now that we have 3.5 in ppa only for a week we might see how many run dailies [18:51] popcon is only a subset, on new installs, it's not even proposed anymore [18:51] its not? [18:51] thats a problem [18:52] i think everybody installing an alpha or daily shuld get it enabled [18:52] those volunteer to help [18:52] so they volunteer to call home ;) [18:52] at least a checkbox during install thy should get if its daily or alpha [18:52] on my netbook, it's off [18:52] fta: did you install plain jaunty or remix? [18:53] remix [18:53] not sure i will keep it though [18:53] * asac looks up his company VoIP data [18:55] Decompressing /mnt/senica_home/alex/Maildir/ARCHIVE/inbox_archive.gz... [18:55] that will take a bit ;) [19:15] asac, do you get that too? http://paste.ubuntu.com/133874/ [19:16] i should not run my bot on a box running jaunty, i keep hitting regressions :( [19:28] fta: http://identi.ca/notice/2870530 [19:28] so yes. its a mess [19:29] gwibber keeps freezing here :P [19:29] here too [19:30] and it's a pig [19:30] fta 3603 1.4 2.8 102688 58372 ? Sl 18:20 1:49 /usr/bin/python /usr/bin/gwibber [19:30] USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND [19:30] for me it felt like they do synchronouzs networking and that blocke or something ;) [19:34] there's a socket leak too [19:43] asac, is there a fix for bzr-bd? [20:54] asac: ping [20:59] !info libnss3-1d jaunty [21:00] libnss3-1d (source: nss): Network Security Service libraries. In component main, is optional. Version 3.12.2~rc1-0ubuntu2 (jaunty), package size 1049 kB, installed size 2796 kB [21:00] !info libnss3-1d intrepid [21:00] libnss3-1d (source: nss): Network Security Service libraries. In component main, is optional. Version 3.12.0.3-0ubuntu5.8.10.1 (intrepid), package size 1030 kB, installed size 2724 kB [21:00] !info libnss3-1d intrepid-security [21:00] intrepid-security is not a valid distribution ['dapper', 'gutsy', 'gutsy-backports', 'hardy', 'hardy-backports', 'intrepid', 'intrepid-backports', 'jaunty', 'jaunty-backports', 'kde4-ppa', 'kubuntu-members-kde4', 'medibuntu', 'partner'] [21:00] !info libnss3-1d intrepid security [21:00] libnss3-1d (source: nss): Network Security Service libraries. In component main, is optional. Version 3.12.0.3-0ubuntu5.8.10.1 (intrepid), package size 1030 kB, installed size 2724 kB [21:11] Jazzva: saw a dent about nspluginwrapper fix? [21:11] dent? [21:11] asac: I noticed a mail [21:11] Jazzva: identica [21:11] ah... [21:11] or twitter [21:11] not sure [21:11] both ;) [21:12] me neither... [21:12] fta: i dont think so [21:12] ? [21:12] fta: james_w doesnt answer to my pings anymor ;) [21:12] asac: I pushed a fix that will fix just the part "don't install to default dirs if NSPLUGIN_DIR is specified" [21:12] maybe i complained too much ;) [21:13] Jazzva: good [21:13] Jazzva: how about fixing -l to look at NSPLUGIN_DIR too? [21:13] (ashamed that it took me so long to send a one-liner, but the school obligations came in :(... and I wanted to send it together with "list plugins in NSPLUGINDIRS") [21:13] Jazzva: archive is frozen for another week ;) [21:13] yeah, that's next on the menu. [21:13] good [21:14] asac: Isn't it that just main and restricted are frozen? [21:14] Jazzva: did you request a merge or sent a mail? [21:14] requested a merge... [21:14] or did you just commit [21:14] ah [21:14] Jazzva: probably true [21:14] (frozen of main/..) [21:14] bug 321814 [21:14] Launchpad bug 321814 in nspluginwrapper "nspluginwrapper creates "default" links even though NSPLUGINDIR is set" [High,Triaged] https://launchpad.net/bugs/321814 [21:15] that's the report, and it contains a linked branch [21:15] yeah. thats bad bug [21:15] asac ^ [21:15] for sure [21:15] bug 345606 [21:15] Launchpad bug 345606 in nspluginwrapper "list_plugin() doesn't take NSPLUGINDIRS in consideration" [Undecided,Confirmed] https://launchpad.net/bugs/345606 [21:15] asac: that's the report for -l [21:16] how can we do the cleanup for cruft created by 321814 [21:16] its a bit tricky because the locations are kind of valid [21:16] if you run it from command line [21:16] can we figure out where such a wrapper points to? [21:16] e.g. to which dir [21:16] well, it installed only in /usr/lib/{mozilla,firefox,iceweasel/plugins/ [21:17] does that directory show up when running "strings" on the so? [21:17] and I think all of nsplugin wrappers begin with npwrapper. or so? [21:17] Jazzva: yeah. most likely we should just remove it [21:17] yep. [21:17] Jazzva: but there could be users that have their own nspluginwrapper and crated their own stuff with that [21:17] but lets focus on the important part ;) [21:17] asac: I'm not sure if we can check that much :) [21:18] asac: we can see if the link is broken... but that only happens when original wrapper is removed [21:19] Jazzva: i thought that the iceweasel files are actual wrappers and not links if you just use nspluginwrapper [21:19] i mean if you just run the command line thing [21:19] does it create links? [21:19] i would think it duplicates the wrapper everywhere [21:19] asac: check initial report in bug 321814 [21:19] Launchpad bug 321814 in nspluginwrapper "nspluginwrapper creates "default" links even though NSPLUGINDIR is set" [High,Triaged] https://launchpad.net/bugs/321814 [21:20] asac: bugabundo's listing shows them as links [21:20] asac: and I noticed the same too... [21:21] I'm away for 15 minutes, need to go and buy cigarettes [21:23] oh [21:23] Jazzva: so if the links point to a known place then we can clean that up [21:28] i'm not sure i should fix bug 190227 [21:28] Launchpad bug 190227 in ia32-libs "ia32 apps look for libs on the wrong place" [Low,Confirmed] https://launchpad.net/bugs/190227 [21:29] it would be nice to have, but the workaround is ugly, and it's marked invalid for gtk itself [21:29] asac, what do you think? [21:32] i think that i am almost falling asleep ;) [21:32] dinner was too much for me [21:41] asac: good [21:45] fta: ok trying to read that bug now ;) [21:46] asac, http://paste.ubuntu.com/133943/ [21:47] so many clones [21:47] yeah [21:47] fta: well. right [21:47] i think its a problem of gtk ;) [21:47] or well [21:47] actually its a problem of ia32libs [21:48] you need to patch that during build [21:48] e.g. binary patch [21:48] maybe canberra should provide something in /usr/lib32/gtk-2.0/2.10.0/loader-files.d/ [21:49] there's a sed in /usr/lib32/gtk-2.0/2.10.0/*.d/* but it contains only text files [21:49] i dont know what loader files do [21:50] is it really for loading gtk modules? [21:50] for me its rather the compiled in PKGLIBDIR [21:50] it seems to have fixed the issue for some [21:50] could be [21:50] let me check [21:51] fta: well. the loader files seem to be kind of related to loading files [21:52] maybe thats a submechanism just for that part of gtk framework? [21:52] Gtk-Message: Failed to load module "canberra-gtk-module": /usr/lib/gtk-2.0/modules/libcanberra-gtk-module.so: wrong ELF class: ELFCLASS64 [21:52] this is not provided by gtk obviously [21:53] but maybe canberra should provide those loader files too [21:54] fta: that sed statement is really ment to be binariy [21:54] (hm, my nano is not stable) [21:54] fta: try to run that on gtk [21:54] x11.so [21:54] asac, yes, of course, but doko seemed to dislike it [21:55] sudo sed -i.bak 's/\/usr\/lib\//\/usr\/l32\//g' /usr/lib32/libgtk-x11-2.0.so.0 [21:55] fta: so doko says that loaders work? [21:56] unclear [21:56] fta: have you ztried the command? [21:57] so the backup thin wasnt really smart ;) [22:00] hold on pushing that 550M src package once again [22:00] done [22:01] ok i guessi broken my lib32 ;) [22:01] i will wait till a new ia32libs arrives ;) [22:02] http://paste.ubuntu.com/133953/ that's what i have so far [22:04] basically with that, my ia32-libs-chromium-browser is empty on jaunty [22:04] i just have that canberra thing left (and other libs from the same bug) [22:22] fta: ok i think gtk needs to be fixed ;) [22:22] not sure [22:22] fta: i look at the code [22:22] it just tries multiple dirs [22:22] takes the first lib that it find that matches the name [22:22] and if the load of that fails it just gives up [22:22] why is it working for the other modules then? [22:23] and doesnt try more [22:23] fta: its not working for at-spi [22:23] it seems to me i just have to add libcanberra to ia32libs [22:23] fta: the other modules are loader [22:23] fta: well. thats obviously a prerequisite [22:23] but still [22:23] the code tells us that it really just looks at /usr/lib/ [22:24] Makefile.am: -DGTK_LIBDIR=\"$(libdir)\" \ [22:25] default_dir = g_build_filename (GTK_LIBDIR, "gtk-2.0", NULL); [22:25] so it has the capability to look at more than one dir [22:25] but unfortunately it doesnt probe libs ;) [22:25] so that feature needs to be extended [22:25] you can set GTK_PATH=/usr/lib32/gtk*/ [22:25] or something [22:25] but that will break your normal 64bit lib a bit [22:25] yes but remember we jsut steal the i386 files and move them [22:26] fta: yes. thats why it doesnt look at /usr/lib32/ [22:27] could be that we use /usr/lib32 during build time as libdir everywhere [22:27] but i dont think we do [22:27] build time of what? [22:27] of gtk on 32bit [22:27] thats the time when the static part gets plumbered into the binary [22:27] i don't think we do [22:28] e.g. we build gtk on 32 bit ... that compiles /usr/lib as libdir in the binary [22:28] then we copy that to /usr/lib32/ ... and everything works except the module path [22:28] that looks at $libdir from compile time [22:28] anyway. i see the issue ;) [22:29] i think i will push my ia32-libs without that [22:30] lpia chroot is roten; http://launchpadlibrarian.net/24118454/buildlog_ubuntu-jaunty-lpia.firefox-3.5_3.5~b4~hg20090319r23829%2Bnobinonly-0ubuntu1~umd1_FAILEDTOBUILD.txt.gz [22:30] fta: http://paste.ubuntu.com/133967/ [22:30] first half of patch [22:30] GTKLIB32_DIR should be done properly in configure.ac [22:30] its hacky that way obviously [22:31] not sure seb will like that [22:35] http://paste.ubuntu.com/133969/ [22:35] unteltes [22:35] fta: heh [22:35] of course he will [22:35] its the right fix ;) [22:35] its just that configure the constant in makefile is done hacky ;) [22:35] also might want to use some ifdef BIARCH [22:35] love [22:36] the whole /usr/lib32/ is hacky anyway ;) [22:37] doesnt fedora have the same dir? [22:37] they have proper biarch support afaik [22:39] let me check if it builds [22:41] hmm. did i svn update after last tiome i built this gtk tree or why is it building everything [22:41] * asac takes a break [22:42] fta: anyway. have you actually tried to put those modules there? [22:42] i mean if everything links against /usr/lib32/ nothing is needed [22:46] that's what i meant earlier [22:46] but it was just a guess, i didn't try [22:47] fta: you shouldnt ask without trying. i mean the lib needs to be added anyway :) [22:47] just add everything that is complained about [22:48] it's already 550M :) [22:48] i don't really mind, i can upload than in 20 secs [22:56] sebner: ok so i get Failed to load module "atk-bridge": /usr/lib/gtk-2.0/modules/libatk-bridge.so: wrong ELF class: ELFCLASS64 [22:56] as this is shipped in lib32 its confirmed that it doesnt work [22:56] let me check if my patch does the trick [22:57] if it would build ;) [22:58] ok it continues ;) [22:58] wrong seb? wrong channel? [22:58] yeah [22:58] wrong seb [22:59] seb == fta [22:59] ;) [23:00] now that i found that notify-osd and gnome-terminal both did the same error about forcing font sizes, i dont feel that comfortable anymore ;) [23:01] seems everyone using cairo or pango directly usually doesnt know exactly what they do ;) [23:01] good that gtk gets that right at least [23:02] evolution is horrible now, huge fonts [23:02] in the message preview [23:02] fta: screen ;)? [23:03] * asac adds evo to todo list [23:06] asac, http://www.sofaraway.org/ubuntu/tmp/evo.png [23:15] Failed to load module "atk-bridge": /usr/lib/gtk-2.0/modules/libatk-bridge.so: wrong ELF class: ELFCLASS64 [23:15] sigh [23:15] pate [23:15] paste [23:15] http://paste.ubuntu.com/133985/ [23:15] ok have to apply that on i32 obviously ;) [23:21] seems the wifi on my netbook is not stable === asac_ is now known as asac [23:24] fta: are you on lbm finally? [23:35] eh, no, not clue what lbm is [23:35] -t [23:36] fta: err, linux-backport-modules [23:36] oh [23:36] thats what i am talking about since ... 24 hours ;)? [23:37] if that doesnt work with that its really doomy i would say [23:38] except trying the madwifi drivers [23:38] from restricted modules [23:39] * asac copyies so from laptop to amd64 [23:42] linux-backports-modules-jaunty then [23:44] fta: yeah [23:44] not sure if thats the pkg name [23:45] fta: can you put canberra into your ia32libs? [23:45] fta: the patch i have works but there are other issues with libspi on 32 bit as it seems [23:45] i think canberra would work. have to check that [23:45] but having it in ia32libs somewhere makes it easier [23:45] e.g. amybe ppa [23:47] yep, done that already [23:47] i need to repush [23:58] asac, http://dvlabs.tippingpoint.com/blog/2009/03/18/pwn2own-2009-day-1---safari-internet-explorer-and-firefox-taken-down-by-four-zero-day-exploits [23:58] asac, pushing a new ia32libs to my own ppa (~fta3) [23:58] with canberra