[00:00] bdrung: and I'll make a note in the changelog [00:13] who wants to see a crasher? [00:16] nobody wants to see a crasher? [00:18] * micahg wishes that this TB update was simple [00:24] BUGabundo, ping [00:26] pong [00:26] hai <3 [00:27] BUGabundo, mind checking something for me? [00:27] is it quick? [00:27] yep [00:27] replying to email [00:27] open a firefox that you don't care about [00:27] and trying to kill my feeds number [00:27] http://scaletech.cwillu.com/truck/ [00:28] too much work, can't read at work :( [00:28] tell me when you've got 10kg showing up :) [00:28] itsss loww [00:28] and sucking cpu [00:28] it's open already? [00:28] missed the connect [00:28] cwillu The connection has timed out [00:28] did it just work? [00:28] retying [00:29] chrisccoulson: bug 563441 [00:29] Launchpad bug 563441 in thunderbird-locales "Update thunderbird-locales to Thunderbird 3.0.4 translations" [Wishlist,Confirmed] https://launchpad.net/bugs/563441 [00:29] 95.172.225.58 [00:29] opens [00:29] okay, one sec [00:29] now what? [00:29] reload the page [00:29] ctrl-r [00:29] one [00:29] done [00:30] no crash? [00:30] none [00:30] okay [00:30] but I have noscript on that page [00:30] ... [00:30] :) [00:31] I don't trust the web [00:31] do you ? [00:31] I do <3 [00:31] the web would never hurt me [00:31] do you want me to allow it? [00:31] me and the web, we go way back [00:31] yes please :) [00:31] done [00:31] now ? [00:31] and ctrl-r [00:32] no crash? [00:32] nope [00:33] refreshing [00:33] slowwww [00:33] slow connection, and you're connected to a server on a beagleboard :p [00:33] as in lost CSS [00:33] reload again ? [00:33] nah [00:34] wonder if it got fixed in a point release [00:34] you're on 3.6.4? [00:34] I can crash on 3.5 and 3.6.3 [00:37] daily ppa [00:38] * cwillu_at_work sees another firefox build in his future [00:38] fta: are the dailies going back into production tonight? [00:52] micahg: lucid only [00:52] (dailies) [00:52] asac: k [00:52] asac: that'll be good so we can test OOPP [00:52] right [00:53] asac: you have time to push/check thunderbird-locales for me [00:53] thats why i ensured we can enable at least that for now [00:53] micahg: where are they? [00:53] micahg: are they fully tested? [00:53] its 7 minutes till freeze [00:53] and then i get painful whips if i broke something without testing ;) [00:53] asac: bug 563441 [00:54] Launchpad bug 563441 in thunderbird-locales "Update thunderbird-locales to Thunderbird 3.0.4 translations" [Wishlist,Confirmed] https://launchpad.net/bugs/563441 [00:54] * micahg test built, will test install/verfiy now [00:55] micahg: did i sponsor the first 3.0 upload ? [00:56] hmm. verison is bad [00:56] its native upload [00:56] no - [00:56] asac: I just did what was there before [00:56] micahg: was the 2.0 package native too? [00:57] asac: no [00:58] asac: so, do you want me to make it native versioning? [00:58] asac: or create a .orig.tar.gz [00:58] + Use 'install-xpi -i' to install to same directory as before [00:58] micahg: whats that? [00:58] why i sthat suddenly needed? [00:59] asac: mozilla-devscripts now installs in /usr/share/xul-ext and we didn't want to do that [00:59] micahg: we didnt use install-xpi before at all it seems [00:59] asac: no, we did [01:00] http://paste.ubuntu.com/414614/ [01:00] thats not clean diff [01:01] asac: I moved the code for modifying the maxversion out of the way [01:01] * micahg guesses he could just comment out in place [01:01] you should have removed it [01:01] no comments [01:01] asac: we'll need it for 3.1 release [01:01] that looks not clean enough ;) [01:02] asac: for Maverick I will make a separate target for it [01:02] asac: I guess I can drop it then and add it back next time [01:03] asac: should I do that? [01:03] micahg: why is our install-xpi now equivalent? [01:03] was it broken without the -i ? [01:03] asac: top line [01:03] in your diff [01:03] yes, that doesnt have -i [01:03] i dont get why we suddenly need it [01:04] did mozilla-devscript regress? [01:04] asac: right, new behavior is to install in /usr/share/xul-ext vs /usr/share/ [01:04] micahg: did you test the binaries? [01:04] asac: I just tested 1 and it worked fine [01:05] 1 usually isnt good enough [01:05] we need to test all primary locales [01:05] for an upload like this [01:06] fr, de, es, pt, en, en_GB [01:09] asac: all look good [01:09] asac: what should I do with that comment in rules? [01:09] asac: and the version? [01:09] thats ok [01:09] i am finishing testbuild and then upload (i wont test he outcome) [01:09] lets hope archive is still open then [01:09] asac: k, I got the FFe for Seamokey [01:10] no new packages? [01:10] micahg: we dont have seaemonkey ready afaik [01:10] asac: I'll have it ready thsi weekend [01:10] asac: should be similar to TB [01:11] asac: if it takes me too long, we'll have to drop from archive [01:13] ooops the upload just finished :) [01:13] asac: ? [01:13] guess too late if my testbuildfails [01:13] locales [01:13] ok [01:13] asac: I tested in pbuilder [01:14] asac: since I added the stub packages for 3.0, we have no new binary apckages [01:15] asac: also, I just modified the upstream .xpi since we're doing native for Tamil and Sinhala [01:15] got rid of the whitespace [01:15] tha upstream .xpi modifications ... are those reproduci ble` [01:15] ? [01:15] are upstream bugs filed? [01:16] asac: yes and yes [01:16] I just deleted the whitespace [01:16] I should have made a changelog note [01:16] asac: and I poked upstream on those 2 as there was no action, you're copied as well [01:17] asac: I was hoping the update would take 5 minutes and ended up taking an hour... [01:21] asac: thanks for the upload :) [01:22] asac: does blueprint = session at UDS or just an action item for next release? [01:23] depends [01:24] not all blueprints get sessions [01:24] asac: I wanted to make a blueprint to clean up the mozilla packages and enhance documentation/standards/consistency [01:24] if thats what you ask [01:24] don't need a session really [01:24] if someone wants a session he needs to propose it [01:24] yeah [01:24] but would like to get approval for it [01:24] you can write a blueprint and submit it for review to get approval [01:25] asac: k, who should review? [01:26] good question. [01:26] propose it for lucid ... [01:27] and if nothing happens give me the blueprint name [01:27] i will sort out [01:27] ;) [01:27] asac: lucid or maverick? [01:27] i guess you know the answer :-P [01:27] asac: k [01:29] asac: 10 insecure left for hardy [01:29] + testing for all [01:30] 10 out of how many ;)? [01:30] asac: 15 [01:30] heh ok [01:30] hurry ;) [01:30] what is with gluezilla in lucid? [01:30] asac: once I have hardy + karmic/lucid changes, jaunty should be easy [01:30] asac: I'm arguing that out with directhex [01:30] I think I fixed the GRE patch [01:30] micahg: is there any way to fix it? [01:31] I think I got it working [01:31] asac: but he dropped the xul patch in debian [01:31] and then sync'd to Ubuntu [01:32] and since then its broken? [01:32] i thought he wanted to sync a new version [01:32] asac: yes, new Debian version w/out the GRE patch [01:32] with the right thing i place [01:32] micahg: with what kind of plan? [01:32] just ignorant? [01:32] asac: "it seems to work" plan [01:33] and now? [01:33] when was that sync? [01:33] asac: with xul193 it crashes [01:33] earlier today [01:33] micahg: what does gluezilla use now? [01:33] links directly with -rpath or what? [01:33] asac: was using 191 [01:33] no [01:33] i mean ... what ws the GRE patch? [01:33] wasnt that the standalone glue booting? [01:33] asac: I think so [01:34] debian/patches/xulrunner_1.9.1_by_default.dpatch [01:34] that one is still in there [01:34] and there is still the booting code [01:34] isnt that what you meant? [01:34] asac: commented out in series [01:35] micahg: so did you agree on a way forward? [01:35] or should i poke him=? [01:35] asac: well, I need to prove that I'm right that xul193 crashes which I think I'll do in the morning [01:35] err [01:35] he was going to grab xul193 from mozila daily repo [01:35] why do you need to proof that? [01:35] prove ;) [01:36] asac: doesn't crash on xul191 or xul192 [01:36] micahg: if its built against xul 1.9.2? [01:36] it should crash on 191 [01:36] or at least its bad [01:36] let me check with him [01:36] actually, I think it'll try to pull 192 if 191 and 192 are avaialble [01:36] the problem is when it tries to pull193 [01:37] i think thats in theory really ok [01:37] the problem is that it might pull 191 even though its linked against 192 [01:37] thats wrong [01:37] and to be safe we should constraint maxVersion too ... yes. [01:38] asac: the default in gluezilla is max system GRE [01:42] micahg: use 1.9.2 and 1.9.* [01:42] or 1.9.3 [01:42] whatever you want [01:43] asac: this seems to work: http://pastebin.com/LjcszwPB [01:45] asac: I was going to try to talk to upstream about all the rdepends after lucid is released and try to get everything updated to xul192, most are already moving in that direction though [01:45] *upstreams [01:47] bbia 2hrs [01:50] ok [01:50] i will see if i can check the gluezilla thing [03:11] fta: you enabled all chromium ? [03:11] intrepid to karmic? [03:11] thougth we wanted lucid first to give them some hope ;) [03:12] ok thats beta. guess that wasnt disabled [03:18] asac - do you not sleep? ;) [03:50] I'm still having this issue: https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/513887 [03:50] Launchpad bug 513887 in firefox "Flash, Java, etc. does not work with out of process plugins and causes Firefox to become unresponsive" [High,Triaged] [04:21] great, and there are literally NO 3.6.3plugin1 Linux64 builds I can find, ANYWHERE, that don't have the brokenness described there. [04:22] DanaG: we'll have new builds tonight that will hopefully work [04:22] ah, cool. [04:27] weird... trying to add security exception for here ( https://users.csc.calpoly.edu/~bellardo/courses/2104/464/index.shtml ) causes a crash in hash4/ [04:28] . It happens even before asking for username and password, and I believe it even happens in safe-mode. [04:29] DanaG: I don't know what hash4 is [04:29] Nor do I. [04:29] But it's the name of the function that's crashing. [04:30] DanaG: can you use firefox -g to get a backtrace? [04:33] www.csc.calpoly.edu/~dgoyette/firefox-crash.log [04:33] that's with "thread apply all bt" [04:33] er, looks like I need to install more debug symbols. === Mook_ is now known as Mook [04:41] #0 0x00007fffdafb3d66 in hash4 () from /usr/lib/firefox-3.6.3/libnssdbm3.so [04:41] hmm, still missing debug symbols? [04:42] www.csc.calpoly.edu/~dgoyette/firefox-crash.log -- replaced file; still missing some debug symbols despite having installed everything. [04:56] DanaG: which firefox do you have installed? [04:57] ii firefox 3.6.3+nobinonly-0ubuntu2 [04:57] ii firefox-dbgsym 3.6.3+nobinonly-0ubuntu2 [04:57] DanaG: it's been fixed in the latest upload [04:57] and same for gnome-support and gnome-support-dbgsym [04:58] The hash4, or the symbols? [04:58] DanaG: the crash [04:58] Cool. For reference, do you have the bug number on hand? [04:58] And I'm wondering why there still seem to be some missing debug symbols. [04:58] DanaG: no, it was an NSS hash issue [04:58] ah. Anyway, once the 3.7 plugin-isolation is fixed, I'll be using that again. [04:59] oh yeah, and I went to try the plugin-isolation on OS X... and boy was I surprised when that feature didn't work there. [05:00] The release notes don't easily say "plugin isolation (excluding OS X)". [05:00] DanaG: hasn't been released yet [05:00] And only in the FAQ do they ask WHY it doesn't work... before they've even annouced the fact that it doesn't work. [05:01] About all I'd ask for is a change to the wording on that page... say right at the top "(except OS X -- see FAQ)" [05:10] http://www.mozilla.com/en-US/firefox/3.6.3plugin1/releasenotes/ -- this page, to be specific. Read downwards, you'll see what I mean. === DanaG1 is now known as DanaG [08:38] bRoas [10:19] chrisccoulson_: seems like it's too late for packagekit [10:28] chrisccoulson: seems like it's too late for packagekit [10:29] micahg - it could probably still be rebuilt, but i can't do that (I can't upload it) [10:29] chrisccoulson: do we need to rebuild? it has no binary depends AFAICT [10:30] micahg - i was wondering the same actually. we already rebuild a few with no binary depends, but AFAICT it doesn't actually have any runtime xulrunner depends [10:30] (it's just the browser plugin) [10:32] chrisccoulson: BTW, when do you think we should drop xul191? The last rdepends AFAICT is gjs and that's on an older version than in archive [10:32] i was going to try and fix gjs today [10:32] binary rdepends [10:32] chrisccoulson: k, if you can go for it [10:33] micahg - and sugar [10:33] chrisccoulson: I still think it has to do with that return type no longer being the expected null [10:33] chrisccoulson: yeah, almost forgot about that... [10:34] chrisccoulson: is sugar going back in? [10:34] micahg - so, i talked with asac about that yesterday, and we sort-of agreed that the only way forward now is the unpopular one (we have to drop it from the archive) [10:35] chrisccoulson: k, I think we can probably make something work for maverick [10:35] micahg - yeah, that sounds good. but, unforunately we're going to have to drop it in lucid [10:35] chrisccoulson: once we get the packages for maverick, we can help them create a PPA for Lucid [12:06] fantastic, none of the gjs tests run when xulrunner is built with --enable-debug === nikolam_ is now known as nikolam [12:13] chrisccoulson: thats normal [12:14] if you build xul with --enable-debug you also have to rebuild rdepends [12:14] as some headers are changed by DEBUG [12:14] :/ [12:14] asac - that's what i'm doing though (i'm investigating a gjs test-suite failure) [12:14] so, gjs is built against the debug version of xulrunner in my build environment [12:15] but all tests now are bailing out with: [12:15] (cx)->requestDepth || (cx)->thread == (cx)->runtime->gcThread, at jsapi.cpp:2876 [12:15] oops [12:15] i missed the "Assertion failure" message off the start of that ;) [12:16] chrisccoulson: right. but the headers have #ifdef .... [12:16] so you need to have the same defs set [12:16] if thats the problem at all here [13:09] just a guess here but is the daily bot down? === gnomefreak76 is now known as gnomefreak [13:15] not sure if my question was posted it seems i lost connection but here it is again just in case. is the daily bot down? [13:33] also Chromiun-browser does not print correctly, it prints too far to right and cuts off a bit of each line [14:15] fta - there? [14:30] * gnomefreak missing Firefox, Chromium for me is not an everyday browser [14:30] * gnomefreak be back [15:05] bug 90652 [15:05] Launchpad bug 90652 in thunderbird "MASTER mozilla-thunderbird crashed [@NS_InitXPCOM3_P] [@nsVoidArray] [@??]" [Unknown,Confirmed] https://launchpad.net/bugs/90652 [15:59] anyone remember the master bug on firefox not opening from a link in tb? [15:59] it seems i found 2 duplicates for that bug [16:01] chrisccoulson, hi [16:02] hi fta, how are you? [16:02] chrisccoulson, fine, you? saw your merge request, thanks! I will process that shortly. does it work as expected? [16:03] fta - yeah, good thanks. thanks for looking at the merge, it seems to be working as expected [16:04] one thing i noticed though is that it only applies to fresh profiles, but i guess that's just how chromium works right now [16:05] fta: chrisccoulson how do you bookmark a page using Chromium? I dont see a way to do it [16:05] gnomefreak: click the star [16:05] jcastro: thanks [16:13] chrisccoulson, once it's in ~/.config/chromium/Default/Preferences, it won't be changed unless the user returns to the pref UI and change it [16:14] fta - ok, no worries. i'm not particularly concerned about that anyway [16:24] there used to be a setting in thunderbird that allowed you to change the placement of your reply. looking for setting to allow me to answer under the senders comments [16:24] i cant find it [16:27] gnomefreak: account preferences [16:32] micahg: thanks [16:33] micahg: you did mean account settings right? [16:33] for bug 561667, is this wording ok? (as in not offensive): http://paste.ubuntu.com/415023/ [16:33] Launchpad bug 561667 in chromium-browser "Package description does not contain the word "chrome"" [Undecided,Confirmed] https://launchpad.net/bugs/561667 [16:34] micahg: ok i found it but it is saying that i am replying under the comments but it is not working that way when i reply [16:35] jcastro, ^^ what do you think? [16:40] anyone? [16:40] gnomefreak: it's separate for each account you mail from [16:40] fta: i like it [16:41] fta: I think it should be the opposite and say what chromium is, not what Chrome is [16:41] micahg: i know they are all set to below but when i am replying it is on the same line as the first comment of email [16:41] fta: assuming we can even mention Chrome as a brand [16:41] micahg, yes, but chromium is not a stripped chrome, it's the opposite [16:42] fta: micahg i see it in software center [16:42] gnomefreak, only if you type "chrom", not "chrome" [16:42] fta: well, Chrome is built on Chrome [16:42] oops [16:42] * micahg needs caffeine [16:42] i searched with chromium [16:43] and it is the 2nd package in list [16:43] gnomefreak, it's not what the bug is about [16:43] see also bug 561667 [16:43] hmmmmmm [16:43] Launchpad bug 561667 in chromium-browser "Package description does not contain the word "chrome"" [Undecided,Confirmed] https://launchpad.net/bugs/561667 [16:44] grrr. bug 559987 [16:44] Launchpad bug 559987 in chromium-browser "searching google chrome should find chromium (dup-of: 561667)" [Undecided,New] https://launchpad.net/bugs/559987 [16:44] ah ok [16:45] fta: I see I have a lot of daily breakage to fix [16:45] yep, me too [16:46] the one thing I know is the dash/bash patch landed on all stable branches [16:48] gnomefreak: if I get to Seamonkey this weekend, it'll be in archive otherwise not [16:50] micahg: 2.0? if so it needs alot of work from what the other person did. he removed all binaries other than the *-browser and removed files (not sure what ones) so it pretty much needs an overhaul unless you use the 1.* branch [16:50] * gnomefreak doesnt recall his name. buts name is not important [16:50] gnomefreak: I'm sticking with the dev branch that asac worked on and am using what I did for TB3 [16:50] micahg: good thanks [16:50] gnomefreak: should only be a few hours [16:51] micahg: ok. what are we doing with gnomefreak: well, we'll have to migrate to 2 at some point [16:51] gnomefreak: also TB3 [16:52] i'm thinking since 1.1.19 is last supported re;lease it doesnt make sense to keep it [16:52] unless they decide to do another release for 1.8.1.x [16:52] IIRC 1.1.19 is last but i havent gotten to looking into it [16:52] nothing other than SM1 is on 1.8.1 [16:53] gnomefreak: TB2 :) [16:53] oh i thought that was 1.9 [16:53] gnomefreak: same problem [16:53] yep [16:53] oh btw bot is dead i think [16:53] gnomefreak: which bot? [16:54] micahg: daily. i didnt get any daily updates for chromium ff tb nothing [16:54] out of our packages [16:54] gnomefreak: bot was disabled because of an archive rebuild [16:54] gnomefreak: was enabled last night, but a lot of breakage [16:54] micahg: ah [16:54] I'll work on the mozilla stuff and fta is working on the chromium stuff [16:55] gnomefreak: you have time to finish lightning for Lucid? [16:55] micahg: no not for this dev cycle. im backed up on work [16:56] micahg, http://paste.ubuntu.com/415033/ better? [16:56] micahg: we are packagin it alone this time arnt we? [16:56] asac: chrisccoulson: anyone doing lightning [16:56] fta: works much betterr for me, but I have no authority :) [16:56] micahg - i've not done anything wit it [16:57] micahg: debian has iceowl done [16:57] micahg: but really. we need to focus on the porting project [16:57] asac: I don't want to do it :) but I don't want it forgotten [16:57] micahg, i'm not native, that's why i ask. it's more about the form/style that i'm worried about [16:58] fta: it works for me because the focus is Chromium [16:58] iceowl is built off our lightning-sunbird package (at least its the same) but if we are dropping sunbird it needs alot of work, or have we decided to include sunbird in Lucid? [16:59] http://www.intac.net/a-comparison-of-dedicated-servers-by-company_2010-04-13/ [17:00] asac: by Tuesday I hope to have the rest of hardy done [17:01] ok i have to get ready for a meeting i should be back tomorrow at least i hope to be. [17:12] hi guys [17:13] hi dholbach [17:13] for some reason my tb does not start any more [17:13] can anybody help me with that? :) [17:13] dholbach: ugh [17:13] dholbach: yes [17:13] 3.0.4+nobinonly-0ubuntu2 [17:14] dholbach: what it the terminal output? [17:14] nothing [17:14] let me get you an strace [17:14] http://paste.ubuntu.com/415039/ [17:14] micahg: ^ [17:15] uh oh... [17:16] dholbach: ls -ld ~/.mozilla-thunderbird ~/.thunderbird [17:17] daniel@miyazaki:~$ ls -ld ~/.mozilla-thunderbird ~/.thunderbird [17:17] lrwxrwxrwx 1 daniel daniel 25 2010-04-15 15:18 /home/daniel/.mozilla-thunderbird -> /home/daniel/.thunderbird [17:17] lrwxrwxrwx 1 daniel daniel 25 2010-04-15 15:11 /home/daniel/.thunderbird -> /home/daniel/.thunderbird [17:17] daniel@miyazaki:~$ [17:17] how did that happen... [17:18] WTF [17:18] daniel@miyazaki:~$ LC_ALL=C ls -la .thunderbird/ [17:18] ls: cannot access .thunderbird/: Too many levels of symbolic links [17:18] daniel@miyazaki:~$ [17:18] yeah, I'm wondering how that happened [17:18] but that means that I can delete it [17:18] and set up all my stuff again [17:19] dholbach: you didn't modify the .sh script for TB, did you? [17:19] no [17:20] there's also ~/.thunderbird.upstream [17:20] dholbach: that's probably where your data is [17:21] oh, that's not good [17:21] -d shows the symlink is a dir :( [17:21] * micahg needs to make a new release for this... [17:22] this is weird [17:22] dholbach: I'd venture to say it's a bug in the shell, but I may be wrong in my limited knowledge of shell scripting [17:22] chrisccoulson: problem :( [17:22] hey micahg - whats up? [17:23] chrisccoulson: -d tells me my symlink is a dir so the migration script keeps running and creating a symlink loop in TB [17:23] chrisccoulson: I'm preparing an update with importance set to high [17:23] ok, got it working again [17:24] micahg - oh, i didn't notice an issue when i reviewed the initial change [17:24] chrisccoulson: I didn't either because I have a .upstream folder already [17:24] so the migration script ignored it for me [17:24] ah [17:24] chrisccoulson: but I don't get why -d shows the symlink as a dir [17:25] ok, i will get that uploaded right away and have it poked through the queue [17:26] chrisccoulson: I should set high, right? [17:26] micahg - yeah, should be fine [17:26] chrisccoulson: should I add a listchanges file or add something in the posinst script to fix [17:27] postinst script to fix the profile? [17:27] chrisccoulson: yes [17:27] that will be pretty difficult, as it doesn't run in the users context [17:28] chrisccoulson: is the listchanges file where I can make an announcement about how to fix? [17:29] i'm not familiar with that actually, but for an announcement of how to fix it you can normally mail u-d-a [17:29] chrisccoulson: I'm saying in the package on install [17:29] oh, i'm not too sure about that [17:29] mvo might know though [17:43] chrisccoulson: do you know about a news file? [17:48] micahg - one second. do you have a number for the thunderbird bug? [17:54] chrisccoulson: bug 563893 [17:54] Launchpad bug 563893 in thunderbird "Thunderbird will not launch do to a recursive symlink" [High,Triaged] https://launchpad.net/bugs/563893 [17:57] micahg - thanks. do you know how to fix this quite quickly? [17:57] chrisccoulson: yeah, I have the fix already [17:57] micahg - ok, thanks [17:57] chrisccoulson: I just need to know how to let people know what to do [17:57] chrisccoulson: will 2 hrs make a difference? [18:03] fta: I think that's a good idea. [18:04] jcastro, already committed, i hope it won't offense anyone [18:05] I don't see why it would [18:06] donno [18:07] micahg: chrisccoulson: so tbird not starting is covered? [18:07] chrisccoulson: is this ok: http://pastebin.ubuntu.com/415059/ [18:07] asac: yep [18:08] asac - yeah, we're just discussing blocking them in the archive [18:08] chrisccoulson: blocking what? [18:08] chrisccoulson: do I note adding a .NEWS file in changelog? [18:08] asac - the thunderbird binaries. the update is moving the profile folder for some users :( [18:09] oh ... bad bad ... fix it if you have a fix!! rather than bother about blocking imo ;) [18:09] asac - there is a fix too, i just need to test and upload it and wait for it to be built [18:09] but i think pitti has volunteered to get it published asap [18:09] chrisccoulson: you can test the fix by reaplacing the script localley [18:10] ok [18:11] just ensure the builds get full build score [18:11] ;) [18:11] asac: I just posted the workaround in the bug [18:11] asac: can you look at my .NEWS post? http://pastebin.ubuntu.com/415060/ [18:12] a line break is missing [18:12] but otherwise its ok. when was this uploaded? [18:12] asac: it's weird, in my file it's there [18:12] why wasnt this properly tested? [18:12] asac: not yet [18:13] no ... when was the problem causing this uploaded? [18:13] asac: I did tested it, but I didn't expect -d to return true on a symlink [18:13] it does [18:13] now I know :( [18:13] so you didnt test it ;) [18:13] yeah. these things are hairy [18:13] i tested it too, and i'm not seeing the issue at the moment (i'm repeatedly opening and closing it) [18:13] every corner case needs to be tested ... especially after we had the first problem with reappearing .mozilla-thunderbird [18:13] asac: the problem is the migration script migrated the profile out and then the symlink to where the profile was [18:13] what condition causes it to happen? [18:14] * micahg thinks you shouldn't be able to mv symlinks [18:14] why not? [18:14] anyway. get this fixed. at best even fix the .thunderbird self symlink [18:14] micahg - where did you put the fix? [18:14] but if thats too hard NEWS is ok [18:14] chrisccoulson: I'm pushing up in a minute [18:14] asac: do I note .NEWS file in changelog? [18:14] when was this regression introduced? [18:15] when was the upload for this? [18:15] 3.0.4-0ubuntu2 [18:15] WHEN? [18:15] yesterday [18:15] ah [18:15] i think it may have only just published though [18:15] ok. get it up asap ... but properly tested [18:15] you dont need to wait for full build to test it [18:15] just test the script [18:15] so put this stuff in NEWS ... upload the fix ... get it up quickly [18:16] not want to hurry you ... so take your timme [18:16] asac: yes, do I need to note NEWS in changelog? [18:16] * micahg is rushing anyways [18:16] well. if your script doesnt fix that situation then yes [18:16] asac - ok, no more than 6 hours ago [18:16] i never used NEWS [18:16] its imo useless [18:16] noone reads that [18:16] at least none of our target audience [18:17] but well. in this case i would push stuff up ... then if folks complain tell them what to do [18:17] if we hav a window less than 24h the amount of folks running into this is low [18:17] and all are on pre-release software so they should be happy that their profile isnt gone [18:17] anyway ... you will handle it [18:22] chrisccoulson, merged, i will build trunk once with it, then merge in the everywhere else, incl lucid [18:22] -the [18:22] chrisccoulson: ok. maybe help verify the new fix ;) ... one thing i wanted to check though is if all the search locales are now done or if there is still something needed? [18:23] asac - the localised search bits are done, just waiting for langpacks [18:23] thanks fta [18:23] chrisccoulson: should I tag for release? [18:24] micahg - did you start writing a mail to ubuntu-devel-announce? [18:24] chrisccoulson: thx for confirming. [18:24] chrisccoulson: no [18:24] chrisccoulson: maybe write that for him. so it doesnt hold back the fix [18:24] chrisccoulson: I have to leave soon, I can write that on my way in to work [18:24] yeah, i can do that [18:24] personally i wouldnt send an email there because of this ... but well ;) ... [18:24] chrisccoulson: so should I tag for release or leave open [18:25] if -desktop folks think thats needed, then so e it [18:25] fix and NEWS are ready [18:26] chrisccoulson: I have to go, tag for release or leave open [18:26] micahg - i can tag that once i've tested it [18:27] chrisccoulson: k, pushed [18:30] "Add symlink from old profile directory to new profile directory" [18:30] that was never discussed was it? [18:30] didnt we say we check if .mozilla-thunderbird and .upstream exists and if so dont do anything? [18:31] micahg: can you repeat for what condition that was the fix for? [18:32] i think we discussed something with links ... just dont remember what that was [18:34] asac: right, that's why I didn't have the issue originally [18:34] asac: if there was no .upstream folder, after the symlink was created, it moved the profile to .upstream and moved the symlink thinking it was an old profile === yofel_ is now known as yofel [18:48] ok, update is working good [18:49] chrisccoulson: test more conditions after uploading ;) [18:50] and do a thorough review of the script code ... not sure how much cruft that has accumulated by now ;) [18:50] (maybe) [18:50] asac - will do. i'm sure i tested all conditions last night, but i think the failure to start only mainfests on a second attempt to start [18:50] and then i ended up with a .thunderbird.upstream folder and didn't see the issue at all :( [18:51] chrisccoulson: yes. starting 4 times is usually a good idea [18:51] to test such profile operations [18:51] yeah, i'll remember that in future [18:51] :) [18:51] i am sure you will ;) [18:52] imo its not a big deal [18:52] much too much stuff was triggered by calling for an emergency [18:52] most folks dont upgrade twice a day ... those that do, know where they can get support [18:53] and if the upload happens without like 12h at best the damage is often minimal ;) [18:53] i noticed such things because the complains usually happen with quite a good delay [18:53] if one listens carefully you can get to know about those really early [18:54] usually there was a bug opened somewhere [18:54] so after uploads having constant refresh for 4 hours gives you the ability to fix >99% of the users before it happens [18:54] acutally let me ask something ;) [18:56] s/refresh/refresh of bugpage/ [19:18] https://bugs.launchpad.net/ubuntu/+source/thunderbird/+bug/563893 => you can just replace the .thunderbird symlink to a symlink to .thunderbird-upstream ? [19:18] Launchpad bug 563893 in thunderbird "Thunderbird will not launch do to a recursive symlink" [Critical,Fix released] [19:37] dupondje, you should remove the recursive symlink (or rename it if you are unsure) [19:37] then move ~/.thunderbird.upstream to ~/.thunderbird [19:40] got it working :) [19:45] dupondje, that's good [19:45] sorry about that [19:46] if you don't change anything you can't break anything :) [19:46] ah well :d [20:10] asac: now it's clear why 1 line fixes that look harmless can be hurtful :) [20:11] hehe [20:11] right [20:11] and more importantly: the world still exist ;) [20:11] chrisccoulson: everything going ok? [20:12] micahg: he is having dinner ;) [20:12] afaik its all good [20:12] asac: I think I check the docs twice on the behavior of the shell test flags, but will reread and propose changes if there's no warning about detecting through symlinks :) [20:12] thanks a lot for fixing this ;) [20:13] asac: np [20:17] asac: about the dailies, should I bump the version in .head for xul192 and firefox36 to 3.6.4+nobin since we only have one version ATM and could add a +build2 later or should I just use +build1 now? [20:18] micahg: +buil1 [20:18] build1 [20:18] asac: k [20:19] i386 18 11325 jobs (47 hours) [20:20] yeah [20:20] seems completely impossible [20:20] its the rebuild ;) [20:20] asac: I assume I should wait on TB before bumping .head, or can we just branch to .lucid if we need to? [20:20] going backwards ... lol [20:20] yeah, but i meant 11325 in 47h [20:20] micahg: bump .head to what? [20:20] fta: yes, that measure is off [20:20] asac: 3.0.5~hg [20:20] fta: maybe if nothing new came in [20:20] asac: daily breakage [20:21] micahg: yeah bump [20:21] asac: k [20:21] we can create .lucid from previously upload tag/commit [20:21] that's one every 15 sec [20:21] fta: 18 builders ;) [20:22] strong buildds ;) [20:22] so its 270sec avg [20:23] problem is that we have bunch of idle ppa builders because that many things in the queue make the scheduler crawl. [20:23] hm, maybe, we'll see in 2d [20:24] so its probably at least double [20:24] fta: btw, i dont think disabling dailies meant keeping /dev going ;) [20:24] i dont care ... just looking at the builders ;) [20:25] also wasnt intrepid disabled ;) [20:25] ? [20:25] i've reenabled everything, and sent lots of fixes [20:25] k [20:25] isn't what you said i should do yesterday? [20:26] i said lucid ;) [20:26] intrepid is gone from umd/ucd but not from ucd-dev/ucd-beta [20:26] well firest i said that, but then they appealed to just do lucid for a few more dails [20:26] fta: yes. but intrepid is dead for everything ;) [20:26] oh, i missed that part then, sorry [20:26] no problem [20:27] iirc, micahg told me it's end of april [20:27] as i was unsupportive aobut the request tod disable dailies [20:27] because of the buildds [20:27] but i am supportive on not doing anything with intrepid anymore that causes buildd slowdown ;) [20:43] I'm seriously affected by bug 556829, and the patch in bugzilla.mozilla.org fixes it for me. What should I do to ensure someone considers adding it to the Ubuntu package for lucid? [20:43] Launchpad bug 556829 in thunderbird "thunderbird-bin crashed with SIGSEGV in nsQueryInterface::operator()" [Medium,Confirmed] https://launchpad.net/bugs/556829 [20:45] maxb: landing on 3.0.5 was already requested which is due out May 11th or so [20:46] But won't that miss lucid-release? [20:46] maxb: thunderbird is upgraded with upstream minor versions [20:48] It still doesn't seem ideal to release in a known crashy state, but I suppose it matters less if a prompt SRU is forthcoming. [21:03] chrisccoulson: you missed the workaround for people with the old package in teh -devel-announce post [21:03] chrisccoulson: BTW, thank you for all your work on this [21:04] micahg - no worries, thank you too :) [21:33] chrisccoulson: is it worth sending a second note to the -devel list with the workaround for the current package [21:33] micahg - what was the other part of the workaround that i missed? [21:34] chrisccoulson: creating an empty .thunderbird.upstream folder [21:34] micahg - that probably doesn't matter too much now. the updated packages are already built (and probably published now) [21:35] if they're not published, then they will be at the next publisher run [21:36] chrisccoulson: they're published [21:37] yeah, i just noticed that too (since 20:14 UTC) apparently [21:37] so, i wouldn't worry about it now [21:37] that was a pretty quick turnaround :) [21:38] we hopefully minimized the number of affected users by deleting the old packages from the archive, and i've not noticed a flood of bug reports from people with the issue [21:42] mirrors probably still have this with some bad luck for 24h or so [21:43] but yes, i think it was quick enough to avoid causing pain on many sides [21:44] asac: I'll keep an eye on the bugs reported and dupe as needed [21:50] chrisccoulson: so i checked gluezilla yesterday and its really in a badish state .... i think the whole xpcom bindings need to be regenerated using this xpidl2cs.pl script that is gone from the mono svn :( [21:50] chrisccoulson: at best we would check if we can rip that mozilla part out of it [21:51] and drop the mono.WebBrowser parts [21:51] if we can figure who needs that [21:51] asac: have you tried my package from ffox36 PPA? [21:54] micahg: the gluezilla? does it patch something else than the xulrunner patch? [21:54] with that you get a grey screen [21:54] asac: no, but it works for me [21:54] which can be worked around by not passing the locationprovider [21:54] micahg: what testcase are you using? [21:54] can you give that to me and instructionshow to run it? [21:54] the one that directhex suggested [21:55] maybe its just the example code directhex gave me [21:55] micahg: yes, with just the xulrunner version constrained its busted here [21:55] removing location provider helps getting stuff rendered reliably [21:55] but there are binding problems on certain evengts [21:55] the whol stuff needs to be regenerated as the xulapi probably changed [21:56] asac: k, is there an easy way to do that or do I need to rewrite stuff? [21:56] its development work [21:56] asac: project for m? [21:56] as i said it would involve first finding the xpidl2cs.pl script that was previously in the mono tree [21:56] and then regenerating everything and potentially fixing loads of things after that :/ [23:15] asac: armel failed on TB3 [23:18] asac: which is weird, because it built before with teh same code abse [23:21] micahg: link to the build entry on launchpad (not the log) [23:21] asac: https://edge.launchpad.net/ubuntu/+source/thunderbird/3.0.4+nobinonly-0ubuntu3/+build/1695070 [23:22] thx [23:22] checking [23:26] looks bad [23:27] asac: I wonder what changed though, ubuntu2 and ubuntu3 were the same code, ubuntu2 built [23:28] asac: https://edge.launchpad.net/ubuntu/+source/thunderbird/3.0.4+nobinonly-0ubuntu2/+build/1693030 [23:36] yeah. thats bad [23:36] trying to figure out if toolchain changed in last minute atm [23:39] asac. users are reporting that ambiance is causing the addressbar drop down and/or menus to be orange and grey. [23:39] occurred with latest updates [23:40] should i contact ken? [23:47] LLStarks: yes [23:47] kwwii [23:49] http://ubuntuforums.org/showthread.php?t=1455112 [23:49] in case you are interested [23:51] nope ;) [23:51] chrisccoulson might be ;) [23:51] (a good safety net to poke kwwii) [23:52] hmmm, when did that happen? [23:52] i haven't noticed any change yet [23:53] oh, the dropdown in the addressbar [23:53] yes. thats the painful place fr dark themes [23:54] yeah, it is quite difficult. the previous colour worked well in firefox but made it difficult to read URLs in other GTK apps [23:55] so evolution is considered more important than firefox ;)? [23:55] heh ;) [23:55] well. so in the end the problem is that firefox should allso a separate class [23:55] gtk resource thing ... not sure ... so you can have a different color than "link color" [23:55] at least thats my thought [23:56] anyway, this half dark approach i dont like at all [23:56] i am using the other ambience like theme [23:56] yeah, a different class would make more sense [23:56] * micahg misses all the *fun* of Gnome by using Xubuntu :) [23:56] i am really close moving to kubuntu or kubuntu-netbook [23:57] asac - i would advise that if we ever ended up with gnome-shell ;) [23:57] espscially if it's still using mozjs... [23:57] chrisccoulson: speaking of that, any luck with gjs? [23:58] isnt gjs just a wrapper? [23:58] micahg - not yet. i got distracted with the thunderbird issue [23:58] not saying thats a bad idea if it hides the bakend completely [23:58] tbh, i'm getting to the stage where i'm thinking about just turning off the test-suit enow [23:58] chrisccoulson: doesn't help [23:59] micah - it would get it to build wouldn't it? (but doesn't fix the real issue) [23:59] i know there's also an armel crash i need to look at [23:59] chrisccoulson: no, I think RAOF tried that after I originally suggested that to get us across the finish line and it didn't help [23:59] oh, how come?