=== hggdh is now known as hggdh|away [06:40] asac: there's some movement on the gstreamer backend for FF40, so there's a chance we'll get H.264 support [07:14] micahg: have you filed the nVidia bug on Mozilla's bug tracker (and/or nVidia's)? [07:15] magcius: not yet, I've been busy, it's on my list to do soon [07:18] micahg: is the mozconfig you use posted anywhere? [07:19] magcius: I don't think we use one, we use the debian/rules file to declare what flags to pass [07:19] micahg: uh oh [07:20] magcius: why? [07:21] it could be it's written out as part of the build, I haven't looked into that [07:21] micahg: hold on. I have no clue where to get a .tar.gz of the package [07:21] yeah [07:21] magcius: for what? [07:21] micahg: I want to see the build config you're using :P [07:22] magcius: easiest way is to go to the daily PPA and grab the orig.tar.gz and the diff.gz [07:22] micahg: hm ok. [07:22] magcius: the firefox one is all-in-one (FF+xul), firefox-4.0 is still split [07:22] micahg: unfortunately, I can't browse Launchpad right now due to the impending nVidia issue [07:23] micahg: ok... are you going to fix the spidermonkey issue that's been around for a while? [07:23] magcius: in that case, [07:23] bzr branch lp:firefox/3.6 firefox [07:23] ok [07:23] cd firefox; ./debian/rules get-orig-source [07:23] yep yep [07:23] magcius: what issue, the lack of mozjs? [07:24] micahg: lack of the /usr/lib/xulrunner dir being in ld.so.conf [07:24] that was never there AFAIK [07:24] we don't want it used as a system lib, that's why tehre's no conf file [07:24] micahg: well yes, but some things that build with it use it [07:25] micahg: there are plenty of things that use the SpiderMonkey API alone. [07:25] magcius: yes, and we have ways to make them work (rpath + rebuild or runtime wrapper) [07:25] micahg: rpath? [07:25] build time path linking [07:26] magcius: yes, there's an open upstream bug to make a spidermonkey that we can actual use as a system lib [07:26] ah, ok [07:27] hm, does building Firefox really require autoconf2.13? I don't remember this [07:27] * micahg doesn't remember [07:28] oh hell no [07:28] why was I denied checking out something out of lp [07:28] you shouldn't be [07:29] yep [07:29] $ bzr branch lp:firefox/3.6 ubuntu-ff [07:29] Permission denied (publickey). [07:29] bzr: ERROR: Connection closed: Unexpected end of message. Please check connectivity and permissions, and report a bug if problems persist. [07:29] do you have an ssh key set up w/bzre [07:30] micahg: I did. I don't see why I would need one to just make a brnach [07:30] magcius: oh, well checkout is anonymous unless you have a key associated with bzr in which case, you need that key on LP as well [07:30] oh wtf [07:31] why did my id_rsa disappear [07:31] that's not good [07:31] micahg: eh, I think it should be anonymous all the time [07:31] magcius: I agree :) I think it's a known bug [07:32] micahg: now to ssh-copy-id all over the place [07:32] I've probably forgotten half my passwords [07:33] magcius: create a new user and check it out as that user [07:34] micahg: it would be cool if Launchpad supported ssh-copy-id [07:36] magcius: Launchpad is open source :) [07:37] micahg: yeah, I know [07:37] micahg: doesn't mean if I code up an extra ssh server they'll deploy it [07:55] micahg: uh, unfortunately, I'm not on Ubuntu or Debian [07:57] micahg: I'm curious how you can use the official Firefox branding when you're making patches to the source distribution [07:57] magcius: the patches are approved by upstream before inclusion [07:58] micahg: ok. [07:58] micahg: so they are slated to go into the next minor release? [07:58] magcius: one of my goals is try to get them accepted upstream in a future release [08:01] micahg: may I ask what this "abrowser" stuff is? [08:01] magcius: a non-Firefox branded firefox [08:01] our version of Iceweasel [08:02] micahg: ok, where is this used? [08:02] magcius: by people who don't want firefox branding, but want the browser [08:02] micahg: ok [09:40] hi all. what PPA would you recommend for getting TB3.1.1? [09:40] (for lucid/32) === yofel_ is now known as yofel [15:47] micahg - i'll have the bot for pushing the crash symbols working later today hopefully, but i'm not going to switch it on straight away. i want to do some testing on it during next week to make sure it behaves properly. i'm not sure if you're interested in helping out with that? [15:47] i was just going to set it up at home to push the symbols to another machine of mine [15:48] and have it watch the daily builds [16:07] czr: on Monday, the daily PPA should have a 3.1.2pre, next Friday, Maverick will get it 3.1.1, then if, there are no bugs by Monday/Tuesday after, I'll push to thunderbird-stable [16:07] chrisccoulson: cool [16:09] micahg - i'm just uploading our first set of symbols now [16:09] i can't wait to get this in to the data center. 200MB upload from my connection takes a long time ;) [16:10] yeah === fta_ is now known as fta === fta_ is now known as fta [17:06] yay! our symbols finally uploaded [17:06] hopefully they are useful ;) [17:44] cool, the upload script seems to work ok [17:45] micahg - if you want to test it, it's hosted here: https://code.edge.launchpad.net/~chrisccoulson/+junk/mozilla-symbols-uploader [17:45] and the official configuration data will be here too: https://code.edge.launchpad.net/~chrisccoulson/+junk/mozilla-symbols-uploader-config [17:56] chrisccoulson: k, maybe over the weekend, thanks [18:28] asac - do you think we should switch the crash reporter on in lucid? [20:10] chrisccoulson_: \o/ they solved the gcc compiler issue, now we just need to wait for the fix to land in maverick === fta_ is now known as fta [20:37] micahg - cool. i'll look at the other build failures over the weekend hopefully [20:37] the crash reporter still isn't working in maverick btw :/ [20:38] it built with it enabled at the sprint, but the in-archive version is still building with it disabled [20:38] it's missing a "Enabled=1" from the application.ini still [20:38] chrisccoulson_: ugh, I'll keep an eye on uploads, if they get the new gcc up, I'll retry the failed armel builds [20:38] micahg - we can't retry the builds unfortunately (LP bug) [20:39] chrisccoulson_: huh? It works for me [20:39] oh, maybe they fixed it [20:39] it used to be the case where i could only retry stuff in universe, and not the things in main that are in packagesets i can upload to [20:39] chrisccoulson_: yeah, I think they fixed that [20:40] but i've not needed to retry a build for a while [20:40] i'll probably just switch off the crash reporter for the other failed architectures, i don't think mozilla support them anyway [20:40] ok, that makes sense [20:41] so, we'll get apport reports for them then? [20:41] micahg - are we going to keep building FF4.0 with system libraries? [20:41] chrisccoulson_: no, that's my project for next weekend, unless you want to do it [20:42] micahg - i don''t mind. it will make it easier to get crash reports to mozilla if we change it [20:42] i'm not sure how we'd handle those with external libraries [20:42] chrisccoulson_: right, also give us feedback on and regressions in mozilla libs vs system libs in enough time to fix before 4.0 is released [20:43] *any [20:43] and make a beta PPA easier :) [20:43] i think your idea about renaming the branches to -current and -next makes sense too [20:43] i had a think about it, and it will make it easier to transition between packages too (we can have a single set of package relationships that works all the time) [20:43] chrisccoulson_: actually was asac's idea from almost a year ago when we went to minimal system libs :) [20:44] ah, ok. but, still it seems like something we should just do [20:44] chrisccoulson_: right, and people with firefox-trunk can always have trunk [20:47] chrisccoulson_: k, also by the look of the schedule, we should probably hold off on Thunderbird 3.1.2 in maverick until after beta [20:48] 3.1.1 can go in after alpha3 though [20:48] yeah, we should aim to get that in after A3. when is 3.1.2 scheduled for? [20:48] chrisccoulson_: sept 7 release [20:48] well not true actually [20:49] that's the FF35/36/TB3 tentative release date [20:49] yeah, it says that for 3.1.3 [20:49] sept 7 is good, it gives us a little bit of breathing space ;) [20:50] I just don't want a potentially bad build on teh beta CD [20:50] I'm glad I have a month of hopefully calm to fix the gjs/gnome-shell issue [20:50] yeah, that point in the cycle is when we start thinking about putting it in maverick-security [20:50] else we risk releasing with an unreleased build [20:50] chrisccoulson_: IIRC, -security doesn't open until after release [20:51] yeah, that's true, but we can still put it in the PPA [20:51] we did that for lucid [20:51] to stage? yeah, we could do that [20:51] although, 3.6.4 ended up being released quite a long time after lucid was released [20:52] but the original schedule was around the same release date as lucid, so it got staged in the PPA during beta [20:53] yeah, that sounds good for the next round of releases [20:54] I just need to try to get everything else updated before feature freeze, last cycle I cut it WAY too close [20:54] i think feature freeze is pretty close isn't it? [20:54] as in an upload the day before final final final release freeze :) [20:54] chrisccoulson_: aug 12 [20:55] I still want to try to get some dpkg-vendor foo pushed up to debian for the rdepends [20:56] but that I can do after FF === fta_ is now known as fta