| === rickspencer31 is now known as rickspencer3-afk | ||
| === asac_ is now known as asac | ||
| hyperair | asac: ping (regarding some ia32-libs stuff) | 10:14 |
|---|---|---|
| asac | hyperair: yeah. go ahead | 10:18 |
| hyperair | asac: pulse is broken. | 10:18 |
| hyperair | the pulse bits in ia32-libs anyway | 10:18 |
| hyperair | skype throws out a whole lot of stuff about not being able to connect to pulseaudio | 10:18 |
| hyperair | and flash refuses to use it either | 10:19 |
| asac | so flash doesnt use pulse anyway | 10:19 |
| asac | it uses alsa | 10:19 |
| asac | skype i am not sure | 10:19 |
| hyperair | er | 10:19 |
| hyperair | i meant the libasound2-plugins bit | 10:19 |
| hyperair | the pulse plugin in alsa | 10:19 |
| hyperair | both of them use it | 10:19 |
| asac | yeah | 10:19 |
| asac | are you on karmic? | 10:19 |
| hyperair | yes i am | 10:19 |
| asac | for me flash sound work | 10:20 |
| hyperair | it does? | 10:20 |
| asac | with 32bit plugin on amd64 | 10:20 |
| hyperair | it does for me when i have nothing running | 10:20 |
| hyperair | =\ | 10:20 |
| asac | not sure if that means that it uses puslse | 10:20 |
| hyperair | check using pavucontrol | 10:20 |
| hyperair | pulseaudio auto suspends-on-idle | 10:20 |
| hyperair | so if nothing uses pulse, it'll release your sound card | 10:20 |
| hyperair | and flash will "work" even if not through pulse | 10:20 |
| asac | hyperair: i can play youtube and rhythmbox | 10:21 |
| asac | but youtube doesnt show up in pavucontrol | 10:22 |
| asac | but afaik it never did that anyway | 10:22 |
| asac | do you have a 32-bit system to check? | 10:22 |
| hyperair | er.. it's on intrepid | 10:23 |
| hyperair | i'll have to upgrade to karmic first. | 10:23 |
| asac | well. intrepid has severe issues with pa afaik | 10:23 |
| asac | so better upgradfe | 10:24 |
| hyperair | intrepid was the rock solid one | 10:28 |
| hyperair | for x86 anyway | 10:28 |
| hyperair | i used it on my notebook | 10:28 |
| hyperair | anyway is it possible to drop libasound2-plugins stuff from ia32-libs? | 10:29 |
| hyperair | there seems to be a lib32asound2-plugins | 10:29 |
| hyperair | perhaps it'd be better to depend on that | 10:29 |
| asac | hyperair: hmm. not sure. most likely we cannot do such things in intrepid anyway. so please checkout jaunty/karmic | 10:33 |
| hyperair | i think you're mistaking something | 10:33 |
| hyperair | my 32-bit machine is running intrepid. | 10:34 |
| hyperair | my 64-bit machine which is having problems is karmic. | 10:34 |
| hyperair | my 32-bit machine is a desktop, the 64-bit is a laptop. | 10:34 |
| hyperair | i personally think that the issue comes from using an outdated version of libasound2-plugins in ia32-libs | 10:36 |
| hyperair | by the way, the 64-bit flash works with pulseaudio. | 10:37 |
| fta | drop it manually, i have an update of ia32-libs planed for really soon (just need the gtk patches in) | 10:40 |
| hyperair | alright | 10:40 |
| fta | it shouldn't be that outdated, as i totally refreshed ia32-libs not so long ago | 10:40 |
| hyperair | i think it's using jaunty debs. | 10:41 |
| hyperair | at least, sources.list.deb shows jaunty deb lines | 10:41 |
| fta | right, but it was fine until recently. btw, it's far less outdated than it was before | 10:41 |
| hyperair | heh i see | 10:42 |
| hyperair | i only started using x64 in jaunty | 10:42 |
| hyperair | and now in karmic | 10:42 |
| hyperair | so i didn't notice =\ | 10:42 |
| hyperair | hmm just replaced the alsa-libs one, and it didn't work. guess it's an outdated libpulse then | 10:47 |
| fta | could be anything, who knows with pulse | 10:48 |
| hyperair | heh yeah | 10:48 |
| asac | fta: did you check whether the gtk i built once fixes your issues? | 11:03 |
| asac | if so i willl just push it upstream now | 11:03 |
| fta | hm, nope, will try later today | 11:04 |
| asac | let me check if the version is more or less recent still | 11:04 |
| asac | @time | 11:07 |
| asac | !info xulrunner | 11:07 |
| ubottu | xulrunner (source: xulrunner): XUL + XPCOM application runner. In component universe, is optional. Version 1.8.1.16+nobinonly-0ubuntu1 (jaunty), package size 279 kB, installed size 1020 kB | 11:07 |
| hyperair | fta: aha! it's libpulse0! | 11:19 |
| hyperair | skype goes through a series of underruns now but flash works nicely. | 11:19 |
| asac | makes sense | 11:20 |
| asac | so just outdated | 11:20 |
| hyperair | yep | 11:21 |
| asac | fta: new gtk uploaded to sandbox ppa https://edge.launchpad.net/~asac/+archive/sandbox | 12:11 |
| asac | i will check if that works too after build has finished | 12:11 |
| asac | fta: so please rename /usr/lib/gtk-2.0/2.10.0/i686-pc-linux-gnu to i486 and also add another link: /usr/lib32/gtk-2.0/ /usr/lib/gtk-2.0/i486-pc-linux-gnu | 13:07 |
| asac | fta: then add libgail.so ... and _remove_ libatk-bridge.so | 13:08 |
| asac | which will tear down everything because of dbus issues | 13:08 |
| asac | fta: and try the gtk package in my sandbox (which is likely the upstream fix) | 13:08 |
| asac | questions? ;) | 13:10 |
| fta2 | mconnor, is smil supposed to be working in trunk now? | 13:11 |
| fta2 | asac, no question ;) | 13:11 |
| asac | fta2: actually i think you might even not need the gtk patch at all ;) | 13:14 |
| asac | so just try to setup the links like above | 13:14 |
| asac | yeah. pretty sure gtk patch is superfluous. its just that the GTK_HOST link needs to be directly in pkglibdir ;) | 13:15 |
| asac | glib patch is still needed though, but i will double check that after lunch | 13:16 |
| asac | now going for lunch | 13:16 |
| fta2 | asac, wtf? http://paste.ubuntu.com/191612/ | 13:30 |
| fta2 | ok, damned Recommends | 13:33 |
| fta2 | Apt::Install-Recommends "false" to the rescue | 13:33 |
| fta2 | asac, now i remember why i stopped working on the ia32-libs update last month, it's because of this: https://edge.launchpad.net/ubuntu/karmic/+source/db4.6/4.6.21-13 | 14:08 |
| asac | hmm | 14:10 |
| asac | right | 14:10 |
| fta2 | dtchen, what happened to libpulsecore9 in karmic? has it been merged in something else? | 14:44 |
| fta2 | asac, damn, i need this one too: https://edge.launchpad.net/ubuntu/karmic/+source/xcb-util/0.3.5-1 | 14:51 |
| asac | fta2: so i think that db-4... "just" needs a libtool update | 14:56 |
| asac | "just", because I tried it and failed miserably ;) | 14:56 |
| asac | let me check xcb | 14:57 |
| === rickspencer3-afk is now known as rickspencer3 | ||
| fta2 | asac, http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=487788#10 | 15:08 |
| ubottu | Debian bug 487788 in gtk+2.0 "gtk+2.0: Searches for 32 bits stuff on 64 bits path" [Important,Open] | 15:08 |
| asac | fta2: seems they didnt attach the actual patch they are talking about | 15:14 |
| fta2 | asac, it's not in 2.12.9-4? | 15:18 |
| asac | oh right. most likely yes. | 15:19 |
| asac | but it only really fixes the meta files from what i understand | 15:19 |
| asac | the rest is a missing link to libdir like i said above ;) | 15:19 |
| asac | bug taching the patch directly (not folded= | 15:27 |
| asac | oops | 15:27 |
| asac | bug 357651 | 15:27 |
| ubottu | Launchpad bug 357651 in firefox-3.5 "firefox-3.5 not adding Shiretoko entry to preferred applications" [Undecided,New] https://launchpad.net/bugs/357651 | 15:27 |
| fta2 | asac, i fixed that for 3.6 and tb3, but not for 3.5 | 15:31 |
| fta2 | asac, the reason is that we will migrate soon | 15:31 |
| fta2 | asac, https://bugs.edge.launchpad.net/ubuntu/+source/ia32-libs/+bug/190227/comments/10 | 15:50 |
| ubottu | Launchpad bug 190227 in ia32-libs "ia32 apps look for libs on the wrong place" [Low,Confirmed] | 15:50 |
| asac | fta2: migrate soon? | 16:13 |
| asac | fta2: i think we should treat this similar to how we treat .desktop files atm | 16:14 |
| asac | fta2: makes sens? | 16:14 |
| asac | should match quite well imo | 16:14 |
| asac | maybe we can even create this .xml file from the .desktop file itself | 16:14 |
| asac | e.g. no need to maintain a copy | 16:14 |
| fta2 | i mean, firefox is listed in gnome-control-center, so if it's also in ff-3.5, no idea what will happen, a duplicate? | 16:16 |
| asac | fta2: i would think the idea is to remove all hard-coded stuff from control-center and make our packages ship stuff instead | 16:18 |
| asac | so we can decide what shows up there and in which way | 16:18 |
| fta2 | oh, ok then, but it was hard to justify just because of PPAs | 16:18 |
| fta2 | that's why i touched everything but ff3.5 | 16:19 |
| asac | fta2: so does gnome-control-center currently ship a firefox.xml? | 16:19 |
| fta2 | no, everything in one giant xml file | 16:20 |
| asac | fta2: maybe it does smart stuff to not duplicate entries if they refer to the same path? | 16:21 |
| asac | anyway. think i will check that out | 16:21 |
| fta2 | it's probably doign smart things. i has ff-3.6 as custom browser, i upgraded to a 3.6 shipping the xml file, it moved automatically to ff-3.6 as preferred | 16:23 |
| asac | hmm | 16:23 |
| asac | fta2: do you have ffox-3.0 removed? | 16:23 |
| fta2 | s/has/had | 16:23 |
| fta2 | no | 16:23 |
| asac | fta2: http://paste.ubuntu.com/191734/ thats gnome-control-center content | 16:25 |
| asac | fta2: ah fount the xml | 16:25 |
| asac | so tbird 3.0 already ships stuff? nice | 16:26 |
| asac | so one idea would be to do the same we do with .desktop wrt to branding. but also dont install if the current package ships the /usr/bin/firefox link | 16:27 |
| fta2 | -rw-r--r-- 1 fta fta 577632132 2009-04-17 23:18 ia32-libs_2.7ubuntu5.tar.gz | 16:28 |
| fta2 | d'oh! | 16:28 |
| fta2 | oh, it's an old one | 16:29 |
| fta2 | -rw-r--r-- 1 fta fta 617070041 2009-06-09 17:20 ia32-libs_2.7ubuntu7.tar.gz | 16:29 |
| fta2 | <asac> fta: so please rename /usr/lib/gtk-2.0/2.10.0/i686-pc-linux-gnu to i486 and also add another link: /usr/lib32/gtk-2.0/ /usr/lib/gtk-2.0/i486-pc-linux-gnu | 16:36 |
| fta2 | asac, it's still called i686 | 16:37 |
| fta2 | and what about the 2nd part? you want a link to /usr/lib/gtk-2.0/2.10.0/ix86-pc-linux-gnu *and* to /usr/lib/gtk-2.0/ix86-pc-linux-gnu ?? | 16:37 |
| asac | fta2: i686 is wrong now | 16:38 |
| asac | change that to i486 again | 16:38 |
| asac | it was just because the last jaunty build of gtk was busted | 16:38 |
| asac | fta2: yes. ship the current link (with i486) and ship a link /usr/lib/gtk-2.0/ix86-pc-linux-gnu -> /usr/lib32/gtk-2.0 | 16:38 |
| asac | that should work | 16:39 |
| fta2 | asac, i still have i686 on karmic.. | 16:42 |
| asac | fta2: how do you know? | 16:42 |
| fta2 | root@cube:~ # ls -l /usr/lib/gtk-2.0/2.10.0/ | 16:43 |
| fta2 | total 24 | 16:43 |
| fta2 | drwxr-xr-x 2 root root 4096 2009-06-08 21:23 engines | 16:43 |
| fta2 | lrwxrwxrwx 1 root root 29 2009-05-18 11:23 i686-pc-linux-gnu -> ../../../lib32/gtk-2.0/2.10.0 | 16:43 |
| fta2 | root@cube:~ # lsb_release -d | 16:43 |
| fta2 | Description: Ubuntu karmic (development branch) | 16:43 |
| asac | fta2: well. that link is created by ia32libs package. ... which is wrong | 16:43 |
| fta2 | oh | 16:43 |
| fta2 | lol | 16:43 |
| asac | fta2: run strings on libgtk-x11.so | grep i.86 | 16:43 |
| asac | what i am saying is that the current link is wrong now and should be i486 again ;) | 16:44 |
| asac | fta2: you can also check that by strace -f -estat chromium-browser | grep 486 | 16:44 |
| asac | hmm. | 16:44 |
| asac | of course not if you are using ia32libs (which is still jaunty state) | 16:44 |
| asac | trust me ;) | 16:44 |
| asac | after updating to gtk from karmic its i486 again | 16:45 |
| BUGabundo | hey asac fta | 16:48 |
| fta2 | hi | 16:48 |
| asac | hi BUGabundo | 16:49 |
| fta2 | asac, "ship a link /usr/lib/gtk-2.0/ix86-pc-linux-gnu -> /usr/lib32/gtk-2.0" why is that needed? for modules? | 16:49 |
| asac | fta2: yes. for everything that isnt in the versioned dir | 16:50 |
| asac | so we basically have /usr/lib/gtk-2.0/VERSION/ix86-pc-linux-gnu and /usr/lib/gtk-2.0/ix86-pc-linux-gnu | 16:50 |
| asac | with ix86 being i486 | 16:50 |
| asac | for everything but jaunty | 16:50 |
| fta2 | asac, what's wrong with libatk-bridge.so? | 16:52 |
| asac | fta2: give it a try ;) | 16:53 |
| asac | fta2: i think that the corba impl it uses doesnt speak proper network encoding over the wire | 16:54 |
| asac | so running 32-bit calls on 64-bit causes failure | 16:54 |
| asac | leading to apps not starting at all | 16:54 |
| asac | canberra works well though | 16:54 |
| asac | so i suspect its really the corba that isnt up to multiarch | 16:54 |
| fta2 | hm, i wanted to document the reason | 16:55 |
| asac | fta2: say it doesnt work in multi-arch cases; symptom: 32-bit apps dont start anymore | 16:56 |
| fta2 | # Drop libatk-bridge.so for now, it's causing troubles | 16:56 |
| fta2 | rm -f $(DEST)$(ROOT)/usr/lib$(SUFFIX)/gtk-2.0/modules/libatk-bridge.so | 16:56 |
| fta2 | ;) | 16:56 |
| asac | fta2: yeah make that: | 16:56 |
| asac | # Drop libatk-bridge.so which seems to be not ready for multiarch | 16:57 |
| asac | but your comment is also ok | 16:57 |
| fta2 | i thought that guy meant it was ok for him: https://bugs.edge.launchpad.net/ubuntu/+source/ia32-libs/+bug/190227/comments/10 | 16:58 |
| ubottu | Launchpad bug 190227 in ia32-libs "ia32 apps look for libs on the wrong place" [Low,Confirmed] | 16:58 |
| asac | fta2: check it on your own ;) | 16:59 |
| asac | for me it killed everything (chromium, flash, firefox upstream) | 16:59 |
| asac | i can check firefox upstream again - i am not 100% sure about that atm | 17:00 |
| asac | fta2: have you verified if chrome really does this ugly daily cronjob? | 17:01 |
| fta2 | no, based on the paste from jorge and report from several users, it seems it's doing it | 17:08 |
| asac | good. thats definitly adware/malware like behaviour | 17:08 |
| asac | adding the sources line on install is ok ... but readding ...pfft | 17:09 |
| asac | so is there a huge rant somewhere about this behaviour already? | 17:10 |
| asac | did you spot something? | 17:10 |
| BUGabundo | I didn't | 17:10 |
| fta2 | nope | 17:11 |
| BUGabundo | not at the blogsfer at least | 17:11 |
| BUGabundo | I could check the foruns dough | 17:11 |
| asac | no need to ;) | 17:11 |
| asac | just was curious if something spilled over already | 17:11 |
| fta2 | a lot of confusion between chrome and chromium, official and non official, packaging vs fork | 17:11 |
| BUGabundo | eheh | 17:11 |
| BUGabundo | yeah | 17:11 |
| BUGabundo | and google builds, cx, and chromium too | 17:11 |
| asac | fta2: so do we know how their release process work? they branch once a week and then roll out for a week to reach 100% diffusion? | 17:13 |
| asac | and then start the process again? | 17:13 |
| asac | (i am talking about how their stable process will look like) | 17:13 |
| fta2 | http://ubuntuforums.org/showthread.php?t=1176370 | 17:13 |
| fta2 | asac, i'm not sure. they have the Dev Channel which is about weekly | 17:14 |
| asac | fta2: i thought they do weeklies on stable channel too | 17:15 |
| asac | at least on win | 17:15 |
| asac | ok break before meeting starts | 17:15 |
| BUGabundo | hey micahg | 17:17 |
| micahg | hi BUGabundo | 17:17 |
| fta2 | asac, http://paste.ubuntu.com/191766/ | 17:19 |
| asac | fta2: yeah. wanna push that to a ppa so i can check? or directly in the archvie (both works for me) | 17:20 |
| asac | fta2: i would say: | 17:21 |
| asac | "add a symlink for the gtk+2.0 modules (LP: #369498) | 17:21 |
| asac | " | 17:21 |
| asac | -> | 17:21 |
| fta2 | the src package should be ~ 630 MB | 17:21 |
| fta2 | inflation | 17:21 |
| asac | "add a GTK_HOST symlink to gtk+2.0 pkglibdir to support modules" | 17:21 |
| asac | s/GTK_// | 17:22 |
| BUGabundo | fta 19 pages about chrome vs chromium autch | 17:22 |
| fta2 | BUGabundo, it starts ~ comment #134 | 17:23 |
| BUGabundo | and who is this ubulette ? | 17:23 |
| asac | good question ;) | 17:23 |
| asac | i think i know him | 17:23 |
| BUGabundo | "I try to keep the packaging in sync with all the good stuff the Chromium authors are bringing in. Sometimes, the daily PPA is making them nervous as it constantly exposes unfinished and untuned features, leading to unwanted bug reports." | 17:23 |
| BUGabundo | duh | 17:24 |
| BUGabundo | its FTA | 17:24 |
| BUGabundo | lol | 17:24 |
| BUGabundo | http://ubuntuforums.org/member.php?u=249378 | 17:24 |
| BUGabundo | "Launchpad Account fta" | 17:24 |
| BUGabundo | eheh | 17:24 |
| fta2 | what's so funny about this? | 17:26 |
| asac | i think BUGabundo just makes fun on of his own this time | 17:27 |
| BUGabundo | asac: :) | 17:27 |
| fta2 | http://ubuntuforums.org/showpost.php?p=7420120&postcount=37 | 17:27 |
| BUGabundo | fta I did not know it was you ! | 17:27 |
| hyperair | speaking of daily ppas.. | 17:29 |
| hyperair | fta2: i'm curious to know if you use cron to automate that. | 17:29 |
| * BUGabundo hides | 17:29 | |
| fta2 | yes i do | 17:30 |
| hyperair | BUGabundo: unwanted bug reports are why the banshee devs didn't want a banshee-daily ppa. they agreed to it due to the banshee-as-default-for-karmic-or-something spec | 17:30 |
| hyperair | fta2: and your gpg key isn't protected by a passphrase? | 17:30 |
| fta2 | correct, the one i use to sign the packages | 17:30 |
| hyperair | aah i see. | 17:30 |
| fta2 | but noone has access to that box | 17:31 |
| hyperair | aah so in that sense it's secure. | 17:31 |
| hyperair | if only i could leave my desktop running like that. =\ | 17:31 |
| hyperair | (mom won't let me) | 17:32 |
| fta2 | :) | 17:32 |
| hyperair | heheh =p | 17:32 |
| asac | now meeting in #ubuntu-desktop | 17:32 |
| hyperair | i was considerign using cron on my notebook for that, but i couldn't come up with a secure enough way to debsign my packages automatically. | 17:32 |
| fta2 | asac, when you create a 2nd ppa, do you share the quota too? | 17:32 |
| asac | fta2: i think the quota is not shared | 17:34 |
| fta2 | hyperair, even if you write a wrapper to sign correctly, you have to store the password in the box somehow | 17:34 |
| BUGabundo | ohh great | 17:34 |
| hyperair | fta2: yeah, that's right. i'm considering it hooking up to my gpg agent | 17:35 |
| hyperair | seahorse-agent. | 17:35 |
| BUGabundo | that's not safe :\ | 17:35 |
| hyperair | yup, it isn't | 17:35 |
| hyperair | because i'll get accustomed to giving my gpg agent my key | 17:35 |
| hyperair | without knowing what it wants it for | 17:35 |
| fta2 | asac, chromium now has zygote by default in linux. it means you can continue to use the browser even if your upgrade or remove it | 17:37 |
| fta2 | -your+you | 17:37 |
| BUGabundo | hyperair: and it stays open a few extra minutes too | 17:37 |
| hyperair | BUGabundo: that's fine by me. i set it to 10 mins. | 17:37 |
| fta2 | daily ppa is task for a server, not for a laptop | 17:38 |
| fta2 | +a | 17:38 |
| hyperair | heh yeah =( | 17:38 |
| hyperair | don't have a safe server anywhere, though | 17:38 |
| BUGabundo | fta just saw a bug on chromium | 17:38 |
| BUGabundo | ctrl+enter doesn't place www. & .com | 17:39 |
| BUGabundo | it works on Windows | 17:39 |
| hyperair | well, i guess i'm content with running the script manually =\ | 17:39 |
| asac | fta2: dont sure i understand what that means | 17:40 |
| fta2 | asac, the problem with have with xul upgrades while ff is running | 17:41 |
| fta2 | -with+we | 17:41 |
| fta2 | d'oh! | 17:41 |
| hyperair | fta2: having issues with the keyboard today? =p | 17:41 |
| fta2 | with my brain | 17:42 |
| hyperair | hehe | 17:42 |
| asac | fta2: so what does zygote do? | 17:47 |
| asac | load libs/files in background? | 17:47 |
| fta2 | it's now the 1st/main process, it opens fd for the binary and the .pak files and keep them forever | 17:49 |
| fta2 | so all children receive those fds from zygote instead of trying to open their files by themselves | 17:50 |
| fta2 | ok, pushing ia32-libs to my new ppa | 17:53 |
| fta2 | done | 17:53 |
| hyperair | woohoo | 17:57 |
| hyperair | does it have a new libpulse0? =P | 17:58 |
| fta2 | yes | 17:58 |
| fta2 | https://edge.launchpad.net/~fta/+archive/staging | 17:58 |
| fta2 | asac, ^^ | 17:59 |
| asac | great. watiting for the bits to appear | 17:59 |
| fta2 | pf, it's already almost full | 18:00 |
| fta2 | ok, i have to run. cu in ~2h. | 18:01 |
| BUGabundo | fta: how can I search for xubuntu-desktop seed on popcon? | 18:44 |
| BUGabundo | I must be blind, can't find a search anywhere | 18:45 |
| asac | BUGabundo: use firefox ctrl+f | 18:46 |
| asac | fta: there is a problem with your dailies | 18:47 |
| asac | at least the mail doesnt look right ;) | 18:47 |
| asac | "svn: Unrecognized URL scheme for 'http://src.chromium.org/svn/trunk/src'" | 18:47 |
| BUGabundo | asac: but how can i filter karmic ? | 18:47 |
| asac | not sure where you are looking at ;) | 18:50 |
| BUGabundo | asac: http://popcon.ubuntu.com/by_inst | 18:52 |
| BUGabundo | trying to get ~ number of xubuntu karmic users for charlie-tca | 18:52 |
| BUGabundo | bye | 19:03 |
| fta | back | 19:20 |
| === rickspencer3 is now known as rickspencer3-afk | ||
| fta | asac, hmm.. indeed | 19:30 |
| fta | $ svn log --limit 1 http://src.chromium.org/svn/trunk/src | 19:30 |
| fta | svn: Unrecognized URL scheme for 'http://src.chromium.org/svn/trunk/src' | 19:30 |
| fta | new subversion is broken | 19:31 |
| asac | fta: how about having your bot machine run a stable release ;)? | 19:46 |
| fta | lol | 19:46 |
| asac | jk | 19:46 |
| fta | bug 385318 | 19:47 |
| ubottu | Launchpad bug 385318 in subversion "svn: Unrecognized URL scheme" [Undecided,New] https://launchpad.net/bugs/385318 | 19:47 |
| fta | asac, so... ia32-libs? | 19:50 |
| asac | checking | 19:52 |
| asac | had to redownload ... no clue where i put it here ;) | 19:52 |
| asac | hmm | 19:52 |
| asac | chromium-browser | 19:52 |
| asac | /usr/lib/chromium-browser/chromium-browser: error while loading shared libraries: libxcb-render-util.so.0: cannot open shared object file: No such file or directory | 19:52 |
| asac | fta: ? | 19:53 |
| fta | lol, ok, i dropped it remember? | 19:53 |
| asac | yeah. but it seems to not work without | 19:54 |
| asac | firefox also fails to start now | 19:54 |
| asac | so maybe just update the really required packages (pulse etc) and leave the rest untouched ... just adding the i486 link | 19:54 |
| * asac has to run to get some food before store closes ... bbl | 19:55 | |
| NCommander | Are we planning to move to thunderbird-3.0 for the karmic cycle? | 20:11 |
| asac | NCommander: 50/50 | 20:20 |
| asac | no spec on this. no upstream commitment on release and so on | 20:20 |
| NCommander | asac, ok, I'm currently trying to run down why thunderbird 2.x series goes kaboom on ARM | 20:24 |
| NCommander | asac, I just broke in the armel retracer on it, but the thunderbird codebase is MASSIVE so I'm not sure where to begin on this | 20:25 |
| NCommander | Especially because I think we have a stack of patches on their stack which may (or may not) be related :-/ | 20:25 |
| asac | NCommander: most likely they are unrelated. | 20:26 |
| asac | NCommander: does tbird 3 work well? | 20:26 |
| NCommander | https://bugs.edge.launchpad.net/ubuntu/+source/thunderbird/+bug/385325 | 20:26 |
| ubottu | Launchpad bug 385325 in thunderbird "[armel] thunderbird-bin crashed with SIGSEGVI" [Medium,New] | 20:26 |
| NCommander | I dunno, I can't find armel binaries which means to test would require a compile | 20:26 |
| asac | give it a try | 20:27 |
| NCommander | asac, is it int he archive? | 20:27 |
| NCommander | and if not, where can I find the deb, and the rdepends | 20:27 |
| asac | https://edge.launchpad.net/~ubuntu-mozilla-daily/+archive/ppa | 20:27 |
| asac | grab the latest daily and spin it on armel | 20:28 |
| asac | does the mobile team have a ppa with armel builders? | 20:28 |
| NCommander | asac, not one that would be acceptable for this, but my personal PPA is completely devirtualized | 20:28 |
| NCommander | (i.e., all architectures are built) | 20:28 |
| asac | so then you use that | 20:28 |
| NCommander | which packages do I want/ | 20:28 |
| NCommander | all of them? | 20:28 |
| asac | just thunderbird-3.0 | 20:29 |
| asac | the rest is in main archive | 20:29 |
| asac | well. the rest that is required for building/running tbird 3 | 20:29 |
| NCommander | meh, there is a pretty good backlog on the armel builder, I'll build on my board | 20:29 |
| asac | NCommander: just copy sources using launchpad feature ;) | 20:29 |
| asac | whatever suites you well ;) | 20:30 |
| NCommander | asac, no, I need to wait for the armel builder to free up, that will take awhile | 20:30 |
| asac | NCommander: you can just copy it ... it will get build when a slot is avail ;) | 20:30 |
| NCommander | :-p | 20:30 |
| NCommander | asac, well, that was quick (https://edge.launchpad.net/builders/nageia) | 20:32 |
| NCommander | asac, BTW, I'll be seeing you at DebConf 9 it seems ;-) | 20:32 |
| asac | NCommander: heh. I will not go ;) | 20:38 |
| NCommander | asac, your not going? | 20:39 |
| asac | i think not. | 20:39 |
| NCommander | asac, oh :-/. I thought you were on the list | 20:39 |
| asac | i am i think. but i still think i am not going ;) | 20:39 |
| asac | have to confirm that though. | 20:40 |
| NCommander | asac, ah ;-). So if thunderbird 3.0 works, then what? (I'm not sure what upstreams plans with the 2.x series are, as this is my first major fray with mozilaa hacking) | 20:40 |
| asac | so not 100% sure;) | 20:40 |
| fta | asac, libxcb needs to buid-dep on xcb-proto (>= 1.5), then xcb-util needs a respin (probably a bump of libxcb too) | 20:40 |
| asac | fta: do we have all the required versions already? | 20:44 |
| asac | or is it because of outstanding merges? | 20:44 |
| asac | NCommander: not sure yet. first check if it works. then we can think what we can do | 20:44 |
| fta | i think so, the deps were not strict enough so pkgs have been built in the wrong order | 20:45 |
| asac | there is an outstanding merge for libxcb | 20:45 |
| asac | https://merges.ubuntu.com/libx/libxcb/REPORT | 20:45 |
| asac | seems someone wants to have it synched | 20:46 |
| NCommander | asac, if you sponsor the result, I'll handle it unless we want the diffs | 20:46 |
| * NCommander hasn't looked yet to see what we are carrying | 20:46 | |
| asac | fta: bug 383954 | 20:46 |
| ubottu | Launchpad bug 383954 in libxcb "Sync libxcb 1.3-2 (main) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/383954 | 20:46 |
| asac | NCommander: not sure what you mean exactly ;) | 20:47 |
| NCommander | asac, I said if a merge was necessary, I'd be glad to do it in return for your thunderbird assistance | 20:47 |
| asac | heh. no its fine | 20:47 |
| asac | thanks! | 20:47 |
| asac | i can assist you, but i cannot debug anything as i dont have a arm board | 20:48 |
| NCommander | asac, want a root shell? | 20:48 |
| asac | point is that tbird 2 is really in permafrost mode | 20:48 |
| NCommander | Yeah, that's the impression I got upstream :-/ | 20:48 |
| asac | so we wont get any assitance from upstream. i expect there to be various issues on that branch, which all got fixed in 1.9 branch and later | 20:48 |
| NCommander | I'm not sure if this is broken as Debian as well | 20:48 |
| asac | NCommander: feel free to compare the patchsets | 20:49 |
| asac | maybe i dropped some arm patches compared to icedove | 20:49 |
| asac | but i don think i did that | 20:49 |
| asac | icedove-2.0.0.x$ ls debian/patches/*arm* | 20:49 |
| asac | debian/patches/18_arm_xpcom_unused_attribute.dpatch debian/patches/38_arm_xpcom_optim.dpatch | 20:49 |
| asac | thunderbird.dev$ ls debian/patches/*arm* | 20:49 |
| asac | debian/patches/18_arm_xpcom_unused_attribute.dpatch debian/patches/38_arm_xpcom_optim.dpatch | 20:49 |
| NCommander | There are enough differences between Debian armel and Ubuntu armel w.r.t. to glibc and toolchain that it could easily flush out a very subtle alignment bug | 20:50 |
| asac | could be. are modern armels (which we support) really that hard about alignments still? | 20:51 |
| === micahg1 is now known as micahg | ||
| NCommander | asac, yeah, but the kernel is supposed to trap it | 21:17 |
| NCommander | but even with the traps on, nothing (which isn't THAT suprising, I've seen the alignment catcher fail miserably before) | 21:18 |
| asac | NCommander: from what i see it crashes in *xc = 0 ... but according to backtrace xc = 0x310000 which is aligned to 4 bytes | 21:25 |
| asac | though that number looks suspicious in general | 21:25 |
| asac | http://mxr.mozilla.org/mozilla1.8/source/js/src/jsdtoa.c#659 | 21:25 |
| asac | hmm. | 21:28 |
| asac | does that for make sense at all? | 21:28 |
| asac | oh its incremented at the end so probably yes | 21:28 |
| NCommander | asac, WTF?!?!?!? | 21:32 |
| NCommander | asac, that code BURNS | 21:32 |
| NCommander | asac, I'm not even sure what its supposed to do | 21:34 |
| asac | ;) | 21:36 |
| asac | NCommander: it does bigint math ;) | 21:36 |
| asac | the variables could be more verbose i agree ;) | 21:36 |
| NCommander | oh god | 21:36 |
| asac | so the diff of that file from 1.8 branch (tbird 2) to trunk CVS (ffox 3) is http://paste.ubuntu.com/191900/ | 21:36 |
| NCommander | Bigint is a struct | 21:36 |
| NCommander | >.<; | 21:36 |
| asac | of course it is | 21:37 |
| asac | its there to hold larger numbers than you can hold by normal types ;) | 21:37 |
| asac | NCommander: the diff doesnt look like it could fix this | 21:37 |
| NCommander | asac, I get it, but *twich* :-/ | 21:37 |
| NCommander | asac, i think its more this is where its segfaulting, but thats because a bad pointer is going in I think | 21:38 |
| NCommander | I hate to ask this, but does thunderbird-3.0 make a -dbg package? | 21:38 |
| NCommander | (please say yes) | 21:38 |
| asac | NCommander: thats what i mean . the address looks suspicious even | 21:38 |
| asac | which would mean that Balloc fails | 21:38 |
| asac | NCommander: yes we have -dbg packages i think | 21:38 |
| asac | fta: ? | 21:38 |
| asac | too bad ... seems we didnt add them yet | 21:39 |
| asac | NCommander: so build locally with pkg-create-dbgsym installed | 21:39 |
| asac | sucks i know | 21:40 |
| asac | thought we added the -dbg packges everywhere again | 21:40 |
| asac | but most likely it doesnt crash with tbird 3.0 ;) | 21:40 |
| NCommander | well, if it works, then I don't care | 21:40 |
| NCommander | if it goes boom, then we got a problem :-/ | 21:40 |
| asac | NCommander: dont you build locally? | 21:41 |
| fta | nope, no -dbg in tb3 | 21:41 |
| asac | if so, just build with pkg-create-dbgsysm | 21:41 |
| asac | installed | 21:41 |
| asac | that will produce dbgsym things | 21:41 |
| NCommander | Looking at the stack trace, #5 in the backtrace caught my eye | 21:41 |
| NCommander | js_ReportUncaughtException | 21:41 |
| NCommander | asac, looking at Debian, icedove is not built on armel (just arm) | 21:45 |
| asac | sure. isnt the name armel an invention of ubuntu? | 21:46 |
| NCommander | Nope | 21:46 |
| NCommander | armel is New EABI ARM | 21:46 |
| NCommander | arm is Old EABI ARM | 21:46 |
| NCommander | (in Debian and ubuntu) | 21:46 |
| NCommander | asac, and actually, its FTBFS on armel, | 21:46 |
| asac | ubuntu didnt have old arm support, did we? | 21:46 |
| asac | tbird 3? | 21:47 |
| NCommander | nope | 21:47 |
| NCommander | I'm just looking at the icedove package in Debian | 21:47 |
| NCommander | which FTBFSed in the configure step, due to the Debian compiler not being happy | 21:47 |
| asac | dont bother with debian compiler i would think ;) | 21:48 |
| NCommander | I was just looking on why it didn't build on armel in Debian | 21:48 |
| asac | ok ... so config.guess/sub update most likely | 21:49 |
| asac | oh. really, i didnt know that we had armel in debian too now ;) | 21:50 |
| * asac feels out-of-date ;) | 21:50 | |
| asac | https://buildd.debian.org/pkg.cgi?pkg=icedove | 21:50 |
| asac | ok i think ill add -dbg packages to tbird 3.0 ... will be in daily tomorrow i guess | 21:52 |
| fta | asac, while you're at it, http://launchpadlibrarian.net/27708075/buildlog_ubuntu-jaunty-amd64.thunderbird-3.0_3.0~b3~hg20090609r2812%2Bnobinonly-0ubuntu1~umd1~jaunty_FAILEDTOBUILD.txt.gz | 22:00 |
| === rickspencer3-afk is now known as rickspencer3 | ||
| asac | sure | 22:04 |
| === Moot2 is now known as MootBot | ||
| asac | hmm. wonder if this will break it | 22:11 |
| fta | ? | 22:11 |
| asac | they dropped the install:: for defaults/prefs.js completely from that file | 22:12 |
| asac | maybe its the result of making it more "make install" friendly | 22:13 |
| asac | but most likely they only added libs:: again | 22:13 |
| asac | :( | 22:13 |
| asac | http://paste.ubuntu.com/191934/ | 22:14 |
| fta | hmm | 22:15 |
| asac | http://hg.mozilla.org/comm-central/log/d1a9b79cf5c2/mail/app/Makefile.in | 22:15 |
| asac | yeah so they removed the prefs.js deliberately | 22:15 |
| asac | ignorant | 22:15 |
| asac | -libs:: $(srcdir)/profile/prefs.js 60 -$(INSTALL) $(IFLAGS1) $^ $(DIST)/bin/defaults/profile 61 - 62 -install:: $(srcdir)/profile/prefs.js 63 -$(SYSINSTALL) $(IFLAGS1) $^ $(DESTDIR)$(mozappdir)/defaults/profile | 22:16 |
| asac | ouch | 22:16 |
| asac | well. they remove libs:: and install:: and only add libs:: again :( | 22:16 |
| asac | http://hg.mozilla.org/comm-central/rev/22248bfae2b7 | 22:16 |
| asac | *why are you guys so ignorant* | 22:16 |
| asac | the bug doesnt talk about install:: ;) | 22:17 |
| asac | https://bugzilla.mozilla.org/show_bug.cgi?id=496857 | 22:17 |
| ubottu | Mozilla bug 496857 in Build Config "license.html and profile/prefs.js not packaged in clobber Mac builds" [Normal,Resolved: fixed] | 22:17 |
| asac | so its an oversight droppage as it seems | 22:17 |
| asac | bugzilla: review+ ? | 22:17 |
| asac | ok mark banner ;) | 22:18 |
| asac | me lurks on #maildev | 22:18 |
| fta | :) | 22:19 |
| asac | lets see ... i complained to the reviewer | 22:20 |
| asac | they say they dropped it semi-intentionally | 22:26 |
| asac | because of https://bug496857.bugzilla.mozilla.org/attachment.cgi?id=382095 | 22:26 |
| asac | err | 22:26 |
| asac | https://bugzilla.mozilla.org/show_bug.cgi?id=394046#c23 | 22:26 |
| ubottu | Mozilla bug 394046 in Release Engineering "Tracking bug: do major update from FF2.0.0.x to FF3.0 to make sure it all works" [Normal,Resolved: fixed] | 22:26 |
| dtchen | fta: yes, libpulsecore9 is now NBS, so it should be removed from karmic. it is superceded by libpulse0, which now contains libpulsecommon-0.9.15.so due to upstream restructuring of libs | 22:39 |
| fta | dtchen, good, dropped from ia32-libs too then | 22:39 |
| asac | fta: wanna kick off a tbird umd2 round? | 23:01 |
| fta | k | 23:02 |
| fta | done | 23:02 |
| asac | seems it still work | 23:02 |
| asac | great | 23:02 |
| asac | grr. apport retracers marks dupes just by top most function name now as it seems :( | 23:13 |
| fta | i wish my UNR bug was moving forward | 23:13 |
| fta | or just moving at all | 23:14 |
| asac | bug 258629 | 23:14 |
| ubottu | Launchpad bug 258629 in nspluginwrapper "npviewer.bin crashed with SIGSEGV in pthread_mutex_lock() (dup-of: 192888)" [Undecided,New] https://launchpad.net/bugs/258629 | 23:14 |
| ubottu | Launchpad bug 192888 in ia32-libs "firefox crashes on flash contents when using libflashsupport" [High,Fix released] https://launchpad.net/bugs/192888 | 23:14 |
| asac | this mass duping cant be good ;) | 23:15 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!