[00:28] hi mozilla team :) [00:33] would i be correct in assuming that (almost) all firefox-* files are replaced with DEBIANNAME? [00:44] * Kamping_Kaiser attempts to do another build [00:45] i'm there, but i'm not :) *suspects hes doing it all wrong* [00:57] having looked i'm none the wiser, so my question is: i'm trying to build a modified (branding wise) FF3 package, and the tarball getting used isfirefox-3.0_3.0.1+build1+nobinonly.orig.tar.gz downloaded from a mirrror, not one of my locally generated ones. where can i look to try and debug this? [01:08] Kamping_Kaiser: Do you know of the existence of abrowser? It's an alternative set of branding for Firefox that was upload to Intrepid a couple of days ago. [01:09] wgrant, yes, i've been watching the discussion. and assuming it builds on hardy it may be worth me looking at. [09:51] bug 201655 [09:51] Launchpad bug 201655 in mozilla-thunderbird "send email with attachement from openoffice through thunderbird fails: "unable to open the temporary file"" [Low,Confirmed] https://launchpad.net/bugs/201655 [10:00] what is firefox-3.0-branding? [10:03] the description needs work for -branding [10:13] * gnomefreak is learning there are alot of people doing very stupid things :( [10:14] :/ [10:15] asac: this is bad [10:15] asac: why does abrowser conflict with firefox? you should beable to have both [10:16] Depends: firefox-3.0, abrowser-3.0-branding [10:16] The following packages will be REMOVED: firefox firefox-3.0-branding firefox-launchpad-plugin [10:17] hmmmmm why do we have both firefox and firefox-3.0 [10:17] but either way it shouldnt conflict [10:18] * gnomefreak is gonna regret trying to install 3D drivers [11:04] there bug is filed [11:04] ^^ for X [11:07] can i still run firefox? [11:09] i guess not [11:13] ok so firefox-3.0 is needed for abrowser but firefox is removed. i cant run firefox-3.0 from terminal since there is no icon for it. firefox-3.0 loads abrowser. shouldnt we make it so we can remove firefox-3.0 if installing abrowser [11:14] or atleast run both. === gnomefreak is now known as thunderstruck === thunderstruck is now known as gnomefreak === asac_ is now known as asac [17:29] hi [17:29] +1:30h mozillateam meeting ;) [18:39] asac: still around? [18:52] yes [18:52] meeting in a few minutes ;) [18:53] asac, I didn't put it on the agenda, but I would like to see what we're gonna do about automatic update process for extensions... [18:53] asac: hi. I've been looking at trying to remove the startpage localisation script stuff from *ubuntu-docs and mozilla-firefox-locale-all. You may remember we discussed doing it last cycle on the basis that it's not used now because ubufox handles the startpage [18:53] Jazzva: yeah. add it ;) [18:53] mdke: what info do you need? [18:54] asac: if you are happy for that stuff to be removed, I'll remove it from the -docs packages, but it should also be removed from mozilla-firefox-locale-all too (I think that has to be done first) [18:54] mdke: hmm [18:54] mdke: thats firefox-2 right? [18:54] asac: I don't know, I have to confess I don't know anything about firefox [18:54] yeah right [18:55] ok so that package isn't even used by firefox 3? [18:55] mdke: i'll add an agenda item for "remove firefox 2" for the meeting [18:55] if we remove it, the langpacks can go too [18:55] mdke: no its not used by firefox [18:55] asac: fine. Can you let me know how it goes? :) [18:55] 3 [18:55] asac: I won't do anything with ubuntu-docs until that's sorted, because mozilla-firefox-locale-all depends on ubuntu-docs [18:55] mdke: meeting starts in 5 minutes in -meeting [18:56] mdke: we can put that agenda point on top if you have to go ;) [18:56] asac: ok, I'll try and follow [18:56] yes, if you could do it first, that would be great, I have to disappear shortly [18:57] mdke: sure [18:57] thanks [19:08] hi [19:08] fta2: in #ubuntu-meeting ;) [19:08] sorry, i'm late for the meeting (traffic jam, 2h stuck) [19:08] only 8 minutes, if you want I can pastebin the log [19:17] asac: as a matter of interest, is something special required to remove the alternatives, like in the postinst, or can I just remove the script that provides them? [19:17] (no need to reply now if you're busy with the meeting) [19:28] mdke: the removal code should be in the current ubunt-docs package that gets replaced [19:28] we should ensure that that works properly for all cases [19:28] but lets talk later about that [20:30] asac: ah, damn - I thought it would be simpler! [20:31] mdke: so it doesnt work= [20:31] ? [20:31] mdke: at best let me check the scripts tomorrow. is that a bzr branch? [20:32] asac: no, I mean - I thought that it would be enough to remove the existing scripts; I now understand from your message above that we need to add some new scripts to remove the alternatives? [20:32] mdke: that depends on whether the previous scripts were correct [20:32] if they were they should remove the alternatives [20:32] if not we need to hack/fix it in the new scripts [20:33] so ... what we need to do is to try ... and see if any clutter is left over [20:33] what the hell id https://edge.launchpad.net/~deadubuntuteam ? [20:33] is [20:33] if yes -> look into it ... otherwise all fine [20:34] asac: ah... well they were done by Ian Jackson, so I hope they'd be correct :) you can see them in the ubuntu-docs source [20:34] mdke: ok. i'd say: just remove them from current scripts. then upload to PPA and lets test [20:34] is that ok? [20:35] asac: sure. What I intend to do is this - http://doc.ubuntu.com/~mdke/ubuntu-docs.remove_startpage_translations_script.diff [20:35] asac: unless you can see anything already wrong from that patch :) [20:38] urgh. thats bloody perl ;) [20:38] fta: ^^ [20:39] yes, it is [20:39] mdke: looks good from what i can tell without looking at the complete sources [20:41] asac: okey. This will be my first use of PPA so lemme check how it works [20:41] i can easily understand it but it looks more cryptic than mine ;) [20:42] fta: q [20:42] yeah. .. thats the code from ian jackson [20:42] the author of dpkg and all the tools [20:44] i already patched some of his scripts, i find his perl old stylish [20:44] fta: really? [20:45] what is "old style" perl? [20:45] like perl4 [20:45] that particular one is two years old, fwiw [20:46] the code above is definitely perl 5 but with some old perl 4.0.36 style in it ;) [20:46] perl 5 is ~15y old [20:47] ah [20:47] fta: interesting [20:50] some of my (old) code is public: http://www.cpan.org/modules/by-authors/id/FTASSIN/ [20:51] * Volans back, finished now to read the log [20:53] gasp, seems i haven't pushed anything to CPAN since 1999 [20:54] fta: did your patches to iwj's scripts get merged into the official branches= [20:54] iwj ? [20:54] fta: or is there something we should merge? [20:54] ian jackson [20:54] sorry [20:55] asac: I have just read the last part of the meeting and I'll be happy to help in some way for the automatization of the extentions [20:55] so feel free to ask [20:56] Volans: cool [20:57] Volans: feel free to work on whatever you want. we tried to split this up in multiple scripts [20:57] i'm concerned by my last cairo update.. i introduced a regression with that lcdfilter stuff but i can't clearly identify what's wrong :( [20:57] needed is download-amd-xpi [20:57] people are complaining [20:57] fta: have you asked keybuk? [20:59] asac, he said he'll have a lot, no answer since [21:01] -lot+look, wtf, i can't type today [21:02] asac: what do you mean exactly with download-amd-xpi? [21:02] in particular "amd" [21:07] Volans: sorry: amo [21:07] ah ok :) [21:07] the script should download a particular version of an xpi from amo [21:08] the last one is not the right one always? [21:08] public, obviously [21:08] Volans: no. we need a particular version i think [21:08] Volans: we want to connect that on a higher level with fta's check extension script [21:08] which basically gives you the latest version [21:09] Volans: i think probably should be [21:09] because if you remember some times ago I have found 2 different manners to automatically download file from AMO [21:09] one that after 3 automatic redirects take the last public versione of an extension [21:10] the second is to watch a watchable url that have all the public versions of an extension [21:11] Volans: right. you think any of those approaches would allow us to download specific versions? [21:11] the first one no, it download always the latest public version in AMO [21:12] the second sure, is a public folder [21:12] we can download the version we want [21:12] fta: how do you find/choose the extension's version with your script? [21:13] I parse the page as my goal is to find the last one [21:13] another nice to have would be to commit all intermediate releases to bzr too [21:13] but thats not important to start with [21:14] and only gives a bit finer granularity - e.g. to track down regressions [21:14] sincerly I think is not necessary fta [21:14] Volans: well. in that way we probably get the latest officially released version, dont we? [21:15] when I say public I was meaning "officially", those that have been public in AMO, not in sandbox [21:15] my script was a quick hack i wrote when we first discussed about that topic here, i plan to rewrite it when i have time to add hooks to trigger actions [21:16] fta: trigger actions? [21:17] such as: if amo vers != ubuntu vers => start dl_from_amo_and_commit_in_upstream_branch [21:17] hmm [21:17] ok [21:18] actions like that being external scripts with a stable api (list of arguments) [21:18] ok. but thats probably far away right? [21:18] so we can ignore that approach for what we are doing now? [21:19] no, not really, it's a progressive approach [21:19] or wouldnt that kind of trigger logic be implemented on a higher level? [21:19] e.g on top of check-extensions [21:19] fta: why not to watch a simple folder instead of parsing the web page? [21:20] asac: probably after the download of the xpi you want some "report" from the script before continue with branches [21:20] Volans, this is a detail in the implementation so it's isolated and could be changed easily.. sure not in the current shell script, but the idea is there [21:21] Volans: yes. the script probably should echo the downloaded file and the version [21:29] asac: Alexander, why abrowser? O_o [21:30] sebner: webbrowser didnt make it through the archive admins [21:31] asac: wouldn't it be easier to sync iceweasel or do we have technical problems then? [21:31] no ... we maintain things anyway. so it doesnt add much extra work [21:31] and iceweasel is a political problem [21:31] also we dont have all patches against xulrunner that debian has [21:32] so it would become difficult mostl ikely [21:32] asac, ff3.1 a2 as it was or as my b1pre is today ? [21:32] asac: I see but in generel what is the difference between abrowser and iceweasel? [21:32] fta: please the official tag [21:33] at least thats QAed by upstream somehow [21:33] asac, for the src sure but the packaging ? my b1pre has abrowser-3.1, not the a2 i had a few weeks ago [21:33] fta: ah. the latest packaging :) [21:34] ok [21:34] not sure how to best to id [21:34] it [21:34] but most likely we need to create a .dev branch as the intrepid release branch [21:34] and that version lowering could be done there= [21:34] ? [21:34] i guess you will find a way ;) [21:35] sure [21:37] does it matter if the 1.9.1 changelog contains the most recent UNRELEASED 1.9 entry ? [21:41] hmm ... cant that be converted to intrepid? [21:41] e.g. bump down -> push to release branch -> merge up -> bump version to 3.0.2bpre? [21:43] what i mean it i currently have 2 "UNRELEASED": http://paste.ubuntu.com/46980/ [21:44] -it+is [21:47] RFC: Any ideas why xulrunner-1.9 c(onfigured with --enable-debug) would get a SIGILL (Illegal Instruction) when firefox is started from a terminal, but not when started via a Gnome launcher icon? [21:48] IntuitiveNipple: you sure both use that build? [21:48] how do you use it? [21:49] asac: what do you mean by 'both' ? Firefox and xulrunner? [21:50] IntuitiveNipple: a) firefox started from launcher b) terminal [21:51] asac: I'm working on that issue we talked of last week, and for some reason something in my xulrunner-1.9 build/install configuration seems to have changed because recent builds of xulrunner-1.9 cause this... if I drop the "--enable-debug" from configure, it doesn't SIGILL. [21:51] asac: Oh... I see what you mean... let me work it out :) [21:52] fta, afaik that is not a problem. But the release entry should have a proper release name. [21:52] well, the launcher is "firefox %U" and I've tried both "firefox" and "export NSPR_LOG_MODULES=HelperAppService:5; export NSPR_LOG_FILE=/home/tj/Desktop/xulrunner-1.9.log; /usr/lib/firefox-3.0.1/firefox" and get the same thing. Let me recheck this, maybe I've got myself mixed up :) [21:54] Jazzva, i know it's not a technical problem, just best practice. here, i try to keep the branches as close as possible so this is a side effect [21:54] fta, I think I've seen that somewhere before... so, I suppose it should be ok. but not 100% sure :) [21:55] fta, check in #ubuntu-motu [21:56] either i pretend the last 1.9 is released, or it's a lot of work to rebase everything on the last 1.9 really released [21:57] or we close and push that 1.9 now ;) even better [22:01] asac: According to strace both end up doing execve("/usr/lib/firefox-3.0.1/firefox", ... [22:02] does it SIGILL inside a debug code ? [22:03] asac: Also, both using "/usr/lib/xulrunner-1.9.0.1" [22:03] can you get a trace ? [22:04] fta: I haven't looked too hard as yet, but it looks like XPCOM / javascript ... let me pastebin the stacktrace [22:06] Here we go: http://paste.ubuntu.com/46989/ [22:07] It must be caused by something I've done doing this debugging, but I can't figure out what! [22:07] The only changes to the configure options I've got are "--disable-optimize \ --enable-debug \" [22:19] can you run it inside gdb and show us 'bt' after the crash ? [22:20] Could not write out perisistant registry! [22:20] ###!!! ASSERTION: Default pref file not parsed successfully.: 'Error', file nsPrefService.cpp, line 723 [22:20] WARNING: Error parsing GRE default preferences. Is this an old-style embedding app?: file nsPrefService.cpp, line 791 [22:26] Yes, I'm rebuilding it atm just-in-case it's some weird transient. I don't get why it starts fine if not from the xterminal [22:30] asac, i branched 1.9.1.head at my a2 commit and cherry picked the improvements on top, leaving all the b1pre out [22:32] fta: There a lot of ASSERTION failures so is that one significant? I don't see anything obvious in the profile's prefs.js [22:33] difficult to say from here without a proper backtrace of the crash [22:33] yeah... if it still happens after this build (with dbgsyms) it might be easier to work out [22:34] asac, http://paste.ubuntu.com/46997/ [22:36] fta: I have a proposal: add the AMO watchable dir in the debian/watch file on the *.ubuntu-dev branches and the use the standard packaging tool uscan and uupdate (just rename the downloaded .xpi file to .zip before running uupdate) [22:37] s/and the/and then/ [22:57] fta asac: OK, figured it out. One of my lines of code was using PromiseFlatCString() without calling the .get() method in a LOG() statement :p [22:59] I guess if the console is detached it doesn't evaluate some part of that - can't think of any other rational explanation anyhow :) [23:05] i wish people wouldn't post unrelated crap to a bug just because it's being looked at by people or sabdfl [23:25] IntuitiveNipple: haha [23:25] ok [23:25] pwnguin: which bug? [23:34] asac: the EULA bug hit slashdot [23:49] bug 269656 [23:49] Launchpad bug 269656 in firefox-3.0 "AN IRRELEVANT LICENSE IS PRESENTED TO YOU FREE-OF-CHARGE ON STARTUP" [High,Confirmed] https://launchpad.net/bugs/269656 [23:49] asac, i thought i blocked that a long time ago [23:49] did you revert it ? [23:50] hmm... what's in the eula? and why do we need it now? [23:52] strange, i blocked it in commit #262 [23:53] timestamp: Sat 2008-05-10 23:21:49 +0200 [23:53] pwnguin: EULA bug? [23:53] asac, look at bug 269656 [23:53] Launchpad bug 269656 in firefox-3.0 "AN IRRELEVANT LICENSE IS PRESENTED TO YOU FREE-OF-CHARGE ON STARTUP" [High,Confirmed] https://launchpad.net/bugs/269656 [23:54] also slashdot mention it ;) [23:54] fta: what do you think about my previous proposal? [23:56] asac: everyone else knows which bug I meant ;) [23:56] Volans, i need to think about it, my brain is not working correctly today. at 1st glance, it could be ok but then the script can't dl without beeing in a branch and at the proper location [23:57] pwnguin, i just happened to read http://ubuntuforums.org/showthread.php?t=919444 in my rss feeds [23:59] fta: yes, this is a limitation, but if you plan to have a long-time maintenance of extensions probably you will have a place whith all the *.ubuntu-dev branches and a master script can cycle through every branch, searching for a new upstream version and made all the stuff explained in the LargeScaleMaintenance [23:59] another way is to make a script that only check the latest version in the watchable dir and download only the extensions that have a new version