[00:04] asac, quick-n-dirty hack to fix gwibber: perl -i -pe 's,https://identi.ca,http://identi.ca,' /usr/share/pyshared/gwibber/microblog/identica.py [00:24] fta: ok, will take a look. i'm recovering from food poisoning ATM. [00:26] dtchen, oh, sorry to hear that. [00:26] dtchen, http://paste.ubuntu.com/116259/ here is what i see from pa [00:28] going to bed. i'm too tired. let me know if you need more info, i'll post that in the morning [09:20] fta: so why did i get the tbird build failures today? [09:20] fta: are you using the last commit author as changelog author on upoad now? [11:49] asac, strangely, i got all the ftbfs logs from the ubuntu-mozilla-security PPA [11:50] fta2: wow [11:50] fta2: cant be a coincident [11:51] fta2: did those go through mailing list maybe? [11:52] donno. no sig, and the To: is my own email [11:54] fta2: hmm. so maybe bigjools went a head and manually fixed our PPAs to send messages to whole team (even though he said this would require a bug) [11:55] hm, do you really believe that? [11:55] otoh, i didnt get any -security build failsures [11:55] just the -daily ones [11:55] https://launchpad.net/~ubuntu-mozilla-security/+archive/ppa/+build/862939/+files/buildlog_ubuntu-gutsy-sparc.firefox_2.0.0.21~20090209t122238+nobinonly-0ubuntu0.7.10.1_FAILEDTOBUILD.txt.gz [11:56] thats the crash right? [11:57] https://launchpad.net/~ubuntu-mozilla-security/+archive/ppa/+build/862954/+files/buildlog_ubuntu-hardy-hppa.xulrunner-1.9_1.9.0.6+nobinonly-0ubuntu0.8.04.1_FAILEDTOBUILD.txt.gz [11:57] yeah i see them https://edge.launchpad.net/~ubuntu-mozilla-security/+archive/ppa [11:57] ok [11:57] seems we never had successful builds on sparc/hppa in hardy [11:58] only firefox gutsy is a compiler crash [11:59] I can never help wondering how much you guys would get done if you weren't fixing sparc/hppa/ppc/omgwtf compiler issues :) [12:01] mconnor: armel ;) ... oh, wait you want that too :) [12:01] well... someone wants that [12:01] me? it's not my problem :) [12:02] and once compiler issues are fixed you get aligment crashes out of the nowhere [12:06] mconnor: did you really say that bundling doesnt lead to market share? http://tech.slashdot.org/article.pl?sid=09/02/09/1537203 ? [12:07] asac: I said that it doesn't prevent others from taking market share [12:08] ok good. i already guessed you ment it that way [12:09] I mean, it's pretty hard to assert that bundling prevents a better product from being adopted [12:09] right. but still it would give the better product a boost [12:09] sure [12:13] asac: it was a little too much shooting from the hip, but basically, I look at it as one of three situations: [12:13] 1) you're bundling a better product. users win. the competition loses. [12:14] 2) you're bundling a similar product. users don't win or lose. the competition loses. [12:15] 3) you're bundling crapware, users get annoyed and find something better. the competition starts taking share. [12:15] of course, smart people than me have cited the must-carry Java stuff as a huge hit to java [12:17] from a software market perspective its probably healthier to not bundle, i agree. OTOH, i also feel bad about training users to search for alternatives and install stuff from the net as they then start to install various things that often comes with spyware and stuff. not sure what is more important. i think i would prefer if the oem market would ship various bundles and not a monoculture [12:26] jcastro: can you please fix gwibber-team PPA package asap? ;) [12:27] asac: I have the daily build from fta and its also broken [12:27] not sure if it still is the SSL prob [12:27] or not [13:05] what is broken? [13:52] any plans on updated Sm 1.1.X or should i do it when im done with 2.0 [13:52] pinery kept crashing yesterday so i havent gotten to sing/upload yet [13:52] sign even [13:59] asac: what's wrong with the gwibber-team package? [14:00] anyone care to update m-d for seamonkey-1.1.x? sunbird 1.x but seamonkey is mail one [14:01] jcastro: it doesnt work anymore [14:02] with identi.ca [14:02] jcastro: well. in fact identi.ca seem to have dropped support for https [14:02] I thought they turned it back on [14:02] checking [14:02] didnt work a few hours ago [14:03] they shut it off for 24 hours or so [14:03] I saw evan telling upstream to not use https, I'll check with ryan when he wakes up [14:03] jcastro: indeed its back now [14:03] thanks [14:03] anyway, they should give a warning before turning off imo [14:05] yeah they were troubleshooting and turned it off, according to them gwibber was using an undocumented thing anyway [14:05] I don't even know why you'd care if your dents were over https anyway [14:11] jcastro: huh? afaik their turned of https [14:11] damn ... me cannot type [14:11] afaik they turned off https ;) [14:11] Yeah I think no one was supposed to be using https yet [14:12] i think polling for new dents is ok but login + posting over https makes sense [14:12] agreed [14:12] ok. but in the end i dont mind that much :) [14:12] as long as the package is working i am happy with whatever ;) [14:12] jheh [14:12] I hope he finishes the stupid keyring soon [14:13] * jcastro looks at the calendar [14:13] keyring? [14:13] you mean for gnome? [14:13] damn ... /me restarts X session to get back window decorations [14:13] yeah, right now he stores the login stuff in a plain text file [14:13] he wants that in the keyring before we put it in for jaunty [14:14] oh right [14:14] but now that i know that we will use http it doesnt really make a differnce ;) [14:15] heh [14:17] fta2: Does https://wiki.ubuntu.com/DailyUpstreamBuildsPOCSpec do what you need? [14:20] james_w, reading... [14:24] can someone try opening a text file by double clicking. let me know if it opens. this is jaunty [14:25] james_w: is lp:firefox supposed to be upstream branch? [14:27] yes [14:28] asac: yes [14:29] i understand that the pre-requisite is that the full upstream sources are cloned into a bzr branch, that excludes all projects using multiple vcs (thunderbird, seamonkey, songbird, chromium, ...) [14:29] james_w: ok, is there a good read-only hg integration yet? [14:30] asac: I don't think so [14:30] james_w: yeah. what fta says is also important: we have nested assembled trees from multiple vcs's [14:30] .. and all projects not currently importable by lp (hg, git, ...) [14:30] nesting is explicitly addressed in the spec [14:30] does that not cover your use of it? [14:31] nesting is not easy, most of the time, you have to run a command after you fetched the 1st tree, to get the rest [14:32] a command provided by the 1st tree? [14:32] tb3 wants python client.mk [14:32] chromium wants cd blabla ; ./gclient in most cases its a command from first tree [14:33] -mk+py [14:33] but first tree now comes from bzr ... second tree would come from hg again [14:33] or cvs [14:33] if thats possible we should also allow first tree to come from other VCS#s [14:33] that's the problem is have with my bot. my local branch feature doesn't work for those [14:34] so i do it without local branch (equiv to your lp:firefox) [14:59] james_w, i'm also wondering how it's supposed to work for packages using embedded tarballs [15:01] and how will you name the versions?? [15:11] oh, I forgot the section on version numbering, thanks [15:11] and do embedded tarballs make sense for daily builds? [15:12] do you extract tip and tar it up, then drop it in the package? [15:21] james_w, well, i decided to stick with the original packaging, there's why i invoke get-orig-source, so all the dh_* rules work, pre-build, patch, etc.. [15:22] so yes, i create a tarball, preferably from a local branch, but i only unpack it to test-build the package (which i don't do right now), the real build happens in the ppa. [15:23] james_w: couldnt we just use custom commands to plumber orig? even without getting some orig branch with bzr? [15:23] i mean if we get comm-central from bzr ... and then mozilla-central from hg we could just take everything from hg ;) [15:27] here are my confs so far, the bot expects nothing else: http://paste.ubuntu.com/116481/ [15:30] (i should really document my stuff) [15:48] ff-4.0 is really 3.2? [15:49] asac: in my opinion that is a small thing [15:49] gnomefreak: undecided [15:50] the revision # sugest it is [15:50] james_w: you mean a small thing to add custom command for orig? [15:50] asac: to allow grabbing branches from hg [15:51] does this look right for dput incoming = ~ubuntu-mozilla-daily/ubuntu/ [15:51] james_w: yes, but we need to run a custom command for nested branches anyway [15:51] gnomefreak, don't dput anything in ubuntu-mozilla-daily, it's meant for the bot only [15:52] ah ok [15:52] gnomefreak: and neither in -security please :) [15:52] i dont have security [15:52] good [15:52] ;) [15:53] new dput line should prefix /ubuntu with /ppa ... e.g. ~asac/ppa/ubuntu/ [15:53] asac: the problem is that if you add in custom commands there is little need for a common tool in the first place [15:53] i'm may even be impossible, i don't think -daily is owned by the team, just me at the moment [15:53] -i'm+it [15:53] I understand you are in a bit of a bind though [15:53] incoming = ~gnomefreak/ubuntu/ is mine [15:53] same with mt ppa [15:54] james_w: well. it provides cron job and auto upload at least [15:55] I understand you are in a bit of a bind though <= was that for me? [15:55] I meant the mozilla team [15:55] your packages are generally outliers in their requirements, and they are not easy to "fix" [15:55] james_w: atm we run this bot on a private machine. having something more officially deployed would be great in the long run. [15:56] there are no plans for that currently [15:56] james_w: also you need to provide hooks for custom commands anyway. e.g. +dfsg tarballs [15:56] need to get stripped or modified [15:56] why can't that be a branch? [15:57] yeah. my fault. i still think about not supported bzr imports ;) [15:57] so if i remove or replace files i need to bzr merge new upstream releases [15:58] err ... upstream dailies. that would work most likely [15:58] depends on git and hg support though [15:59] james_w: once we can sync from such branches, we could also try to provide a "hg", "git" wrapper that maps the sub-checkouts to the right bzr equivalent [15:59] not sure how easy it would be to guess bzr rev-ids from git/hg commits [16:00] but thats usually in such branches right? [16:00] so we would have: [16:00] lp:comm-central (aka tbird) [16:01] and when client.py runs something like: hg clone -r REVISION http://hg.mozilla.org/mozilla-central, we would do bzr branch -r MAPPEDREVISION lp:mozilla-central [16:02] sounds hard to do ;) [16:02] too bad [16:05] asac: where is the nobinonly script? sm1.1.x needs to be done by hand [16:06] gnomefreak: in debian/? [16:06] remove.nonbinonly.sh [16:06] gnomefreak, no, it has everything needed [16:06] in ubuntu [16:06] or something? [16:06] hmm [16:06] fta2: for new tarball? [16:06] then follow fta ;) [16:06] release == 1.1.15 IIRC we have 1.1.13 in repos [16:06] i remember it is able to repack the official tarballs [16:07] looking in m-d its not listed [16:07] get-orig-source & get-current-source are there [16:08] it's using uscan and performs a repack afterwards [16:08] so just do ./debian/rules get-orig-source [16:08] ok ill try it [16:09] will try it [17:08] * asac spins an unoptimized -central build [17:27] * gnomefreak reallying starting to hate quilt [17:27] be back maybe it will work this time [17:34] after autoconf2.13 quilt shows ~17 patches but in debian/patches(our dir) shows ~24 or so === asac_ is now known as asac [18:07] asac: what am i missing? I looked at the orignal file that was patched by the zip patch that failed and 1st thereis only 109 lines not 112+. here is output of erverything i ran so far let me know what im missing? http://pastebin.mozilla.org/620869 [18:07] * gnomefreak goes for smoke [18:14] scratcvh that seems the killall patch was applied upstream [18:16] fta: miro2 for jaunty? ;p [19:16] hello all [19:17] I am able to get the date of a selected message in thunderbird but its not coming in as a formated date it looks like a timestamp [19:17] how do i convert that to a formated date [19:23] ::echo:: :: echo:: [19:23] lol [19:40] anyone awake in here? [19:48] are you fucking kidding me [19:49] ive removed a shit load of patches nad still going :( [20:08] this better work this time, i have a life to get back to :( [20:08] 17 patches checked adn removed === fabrice_sp_ is now known as fabrice_sp [20:12] jcastro, what about miro? [20:14] fta: 2.0 got released just a bit ago, I saw an svn snapshot from a while back in your PPA [20:14] jcastro, i can sure bump mine but i'm no way the maintainer. i think we get it from debian. === thunderstruck is now known as gnomefreak [20:15] ah ok, just wondering [20:16] jcastro, i have the tendency to bump a lot of packages that i feel are not moving fast enough, neither in ubuntu, nor in debian, or even package stuff myself, but most are not ending in any repo, just my ppa :( [20:17] users seem to like my ppa because of that [20:17] ok im going out while this damn thing builds if it fails anymore ill look at it tomorrow [20:17] fta: you're the crack master! [20:17] jcastro, i don't think so, but i sure like fresh food [20:18] * jcastro hugs [20:21] hm, too much http://is.gd/j51Q [20:22] i'm seeing this http://is.gd/j528 now [20:23] do you want miro2 ? [20:23] jcastro, ^^ [20:24] I think it'd be great to have in the distro [20:27] jcastro, well, that i can't do, so for some reason i don't really understand, new stuff takes forever to get in [20:38] dtchen, do you guys have a daily ppa for pulseaudio ? or should I setup one (or try to do so) [20:38] i mean trunk of course [21:23] sigh ... /me too dumb to do a non-debug build ;) [21:24] respin [21:27] i think i should trash my main chroot, too much stuff in there [21:29] 256 upgraded, 0 newly installed, 0 to remove and 37 not upgraded. [21:29] Need to get 217MB/217MB of archives. [21:29] this is not good in a chroot :P [21:47] heh [21:47] take care that you dont trash your bind mounted home ;) [21:47] i don't do that [21:48] no risk then [21:48] thats bad [21:48] firefox-bin from browser build doesnt crash [21:48] on exit [21:48] whats going on with the xul stub [21:51] http://paste.ubuntu.com/116603/ [21:51] asac, grrr, still no build-system.tar.gz in the xul 1.9 you last pushed [21:52] fta: here is what i did: i produced a orig.tar.gz in jaunty; then bumped version and spun package [21:52] fucking patch is starting to piss me off now [21:52] fta: what is missing on xulrunner-1.9.head branch? [21:52] its needed but wont apply [21:53] gnomefreak: which patch? [21:53] 35_theme_switch.patch [21:54] everythin with # less than 35 is trash they applied all upsttream [21:54] let me look [21:54] souds wierd that many patches get applied upstream on 1.8 [21:54] i didnt push to branch yet [21:54] so all old patches are still there [21:54] cant believe they applied all. really [21:55] gnomefreak: i think you messed up your orig ;) [21:55] yep checked each by hand [21:55] with find in geidt [21:55] how did you check? [21:55] hard to do [21:55] open patch open file that was patched used find in gedit located and compared [21:55] yes. i think you made a mistake .... please paste first 100 lines or so of tar tzf seamonkey-.....orig.tar.gz [21:56] if the + was added that hunk was applied [21:56] of my build failure? [21:57] no the command i wrote ;) [21:57] tar tzf ... [21:57] append: [21:57] | head -n 80 [22:00] damnit that is my fault hold on a sec [22:01] \http://pastebin.mozilla.org/620955 [22:01] http://pastebin.mozilla.org/620955 [22:01] that didnt look like 80 lines [22:03] http://pastebin.mozilla.org/620957 were applied upstream http://pastebin.mozilla.org/620957 and yes each hunk :( [22:03] whats with the double shit [22:05] gnomefreak: did you try to use quilt push to apply them? [22:06] yes each time i used totle of 5 qyuilt commands starting with quilt applied than push -f than autoconf than refrexh than pop and push -a [22:07] until all were done [22:07] all were " cannot bew removed cleanly" as i recall thats when i brought them up and made sure they were applied [22:08] and i copied the patches dir to our debian dir [22:09] gnomefreak: na. first do a single quilt push [22:09] if that doesnt apply check what the error is [22:09] i did each one at a time [22:09] took me 3 fucking hours [22:09] give or take an hour [22:11] thats odd a3 is < a2 :( [22:11] where? [22:12] i fixedit [22:12] my PPA wasnt in sources.list [22:13] SM2 is done finally [22:14] btw is there a faster way than looking at each patch and each file(hunk) to make sure they are applied? [22:15] diff not very handy in that since some of these files are 9000 lines [22:15] in general no. usually if the exact same patches are applied, quilt will tell you that they could be reverse-applied [22:15] well. thats how it is [22:15] anyway, i still doubt that all got applied [22:15] asac: only 17 out of what 30? [22:16] _only_ lol [22:16] i mean. thats like a major upstream happening on a EndOfLife branch [22:16] fta: what do you think ;) [22:16] ? [22:16] yeah i figured that too but i'm going by what i saw :) [22:17] gnomefreak: my guess is you somewhat applied all patches and then tried to reapply them ;) [22:17] i didnt push so they are all there in the branch still [22:17] at least one possible story [22:17] or you didnt create proper patches directory link [22:17] and quilt messed up [22:17] fta: can dquilt help gnomefreak to apply quilt patches in a safe manner? [22:17] i ran all quilt commands than copied over than did again and so on [22:18] yeah that sounds wrong ;) [22:18] dquilt? [22:18] even though i dont understand it [22:18] gnomefreak: for me all patches apply cleanly ;) [22:18] all i did was: [22:18] every time i ran autoconf-xx i copied over [22:19] bzr branch lp:~mozillateam/seamonkey/seamonkey-1.1.dev [22:19] asac, dquilt is not perfect, but it's not an embedded tarball, it should work just fine [22:19] +if [22:19] cd seamonkey-1.1.dev [22:19] bzr bd --merge --dont-purge --builder='debuild -b' [22:19] gnomefreak: ^^ [22:19] asac: looking [22:20] well ... i also did ./debian/rules get-orig-source; mv seamon*tar.gz ../tarballs/ [22:20] before spinning [22:20] bzr bd --merge --dont-purge --builder='dpkg-buildpackage -rfakeroot -k764D5E13 -i.bzr' . should not make a diff. between yours and mine [22:20] yeah i did that already ") [22:20] gnomefreak: ok so i was wrong. all patches apply except 35_zip_cache.patch -> that can be dropped [22:21] that makes sense because i remember that this landed upstream [22:21] let me check if something else fails [22:21] that was the second one first was 30_killall.patch as i recall [22:21] gnomefreak: you said: _all_ [22:21] but no ... 30_killall.patch worked [22:21] i said 17 [22:21] gnomefreak: thats it [22:22] let me commit it [22:22] everything else applies [22:22] ok will try again [22:22] i also have a shitload of CVE's [22:23] gnomefreak: ok committed [22:23] rev 158 [22:23] ok looking [22:24] yep i got it. will let you know when im done [22:25] asac, the problem of dquilt is to get DEB_SRCDIR for embedded tarballs, it's broken when DEB_SRCDIR is not an already expended value, i.e., when it refers to another variable or to a shell command [22:26] fta: hmm. ok [22:26] fta: so make a "makefile out od dquilt" that includes the debian/rules ;) [22:27] asac, i was tempted to add an echo-deb-srcdir target to debian/rules, or to something always included [22:29] like DEB_SRCDIR=`make -f debian/rules echo-deb-srcdir` [22:33] fta: maybe like: echo-config CONFIGVARS=DEB_SRCDIR,MOZ_SOMETHING,... [22:36] gnomefreak: does it build? === thunders1ruck is now known as gnomefreak [22:37] asac, having that wont help spread dquilt outside of the mozillateam. it would need to be in cdbs [22:39] getting 14K mail headers from gmail imap probably takes a bit ;) [22:50] hmm ... now i have a bug that should go into an FAQ [22:50] not sure how to tag that properly without reopening [22:51] https://bugs.edge.launchpad.net/ubuntu/+source/firefox/+bug/295099 [22:51] Ubuntu bug 295099 in firefox "Firefox Problem, Letters with huge space between them" [Undecided,Invalid] [22:56] OOm [22:57] good that i have disabled swap ... otherwise tbird wouldnt have been killed [22:57] someone really needs to implement memory pressure [23:43] <[reed]> asac: Gecko supports memory pressure already... [23:43] <[reed]> Fennec is using it