/srv/irclogs.ubuntu.com/2010/08/19/#ubuntu-mozillateam.txt

fta2hi08:30
fta2micahg, ff is blocked on firefox.postinst08:30
fta2Setting up firefox-branding (3.6.9~hg20100817r34537+nobinonly-0ubuntu1~umd2) ...08:30
fta2Setting up firefox (3.6.9~hg20100817r34537+nobinonly-0ubuntu1~umd2) ...08:30
fta2Installing new version of config file /etc/apparmor.d/usr.bin.firefox ...08:30
fta2update-alternatives: using /usr/bin/firefox to provide /usr/bin/x-www-browser (x-www-browser) in auto mode.08:30
fta2stuck for ~1h doing nothing08:30
micahgfta2: on your system?08:31
micahgfta2: can you attach gdb or strace to it and see where it's stuck?08:31
fta2there's no crash08:35
fta2apt-get(13297)───dpkg(15432)───firefox.postins(15525)───cut(15529)08:36
fta2so the cut is expecting something08:36
micahgfta2: yes, but you said it's stuck, so it must be waiting for something08:36
* micahg checks08:36
fta2        ubuntu_major=`cut -d '.' -f 1`08:37
fta2        ubuntu_minor=`cut -d '.' -f 2`08:37
fta2needs something to cut08:37
fta2a filename or something on stdin08:38
* micahg and bzr blame jdstrand :)08:38
micahgfta2: agreed08:38
micahgjdstrand: can you look at the ff3.6.head postinst in the morning and make sure there's something for it to cut :)08:38
* micahg is supposed to be sleeping08:39
=== maxb_ is now known as Guest7510
=== Guest7510 is now known as maxb
=== maxb is now known as Guest10006
=== maxb_ is now known as maxb
jdstrandmicahg, fta2: ooooppps!13:40
jdstrandI'll fix13:40
jdstrandI was trying to test yesterday but couldn't get firefox to build on maverick and was going to look at why this morning13:40
fta2jdstrand, hi, did you get the last part?13:41
fta2<fta2> jdstrand, http://paste.ubuntu.com/480310/13:41
fta2<fta2> also, i'm not sure this -ge test is wise. isn't lucid now 9101?13:41
jdstrandfta2: lucid is 10.04.1, but the cut is on the '.', so I'll only take the 1st and 2nd anyway13:42
jdstrandfta2: not to mention, lsb_release doesn't show 10.04.113:42
jdstrandfta2: I thought the same thing yesterday and looked at it pretty hard13:42
jdstrandand decided it was ok13:43
jdstrandof course, if I actually used lsb_release! :)13:43
fta2i've been told to use /etc/lsb-release instead of lsb_release as the latter involves a lot of python and is too slow, hence my patch13:43
jdstrandI stole that from debian/rules btw13:43
jdstrandfta2: oh, your patch in chromium? I'll look at it13:44
jdstrandthanks13:44
fta2the pastebin is for firefox13:44
fta2inspired from my patch in chromium13:44
* jdstrand nods13:44
jdstrandoh I see13:44
jdstrandcool13:44
jdstrandoh I get it now13:45
jdstrandright, so if yours did 10.04.1 we'd be in trouble13:45
jdstrandecho DISTRIB_RELEASE=10.04.1 | cut -d= -f2 | sed 's/\([0-9]*\.[0-9]*\)\..*/\1/g' works13:49
jdstrandI don't like it as much though13:49
* jdstrand keeps fiddling13:49
jdstrandecho DISTRIB_RELEASE=10.04.1 | cut -d= -f2 | cut -d '.' -f1,2 | tr -d '.'13:50
jdstrandthat last cut works, but is weird13:51
jdstrandI think I'll use this:13:52
jdstrandecho DISTRIB_RELEASE=10.04.1 | cut -d= -f2 | cut -d '.' -f-2 | tr -d '.'13:52
jdstrand(obviously with grep and not echo)13:56
HurricaneHarryI have a dpkg --configure problem with firefox (3.6.9~hg20100817r34537+nobinonly-0ubuntu1~umd2~lucid)13:56
jdstrandwell, maybe not obviously with my last lame commit13:57
jdstrandHurricaneHarry: is it just hanging?13:57
HurricaneHarryyes13:57
chrisccoulsonhi jsdstrand13:57
chrisccoulsonso it was you who broke the dailies :P13:57
jdstrandHurricaneHarry: yeah, that would be me and what I am talking about13:57
jdstrandchrisccoulson: I think you mean I just fixed the dailies13:57
chrisccoulsonheh :)13:58
jdstrandchrisccoulson: it is all a matter of perspective13:58
jdstrandseriously, sorry13:58
chrisccoulsonjdstrand - i also noticed that you're committing to lp:~mozillateam/firefox/firefox-3.7.head13:58
chrisccoulsonwe're not using that branch anymore ;)13:58
jdstrandchrisccoulson: ah, ok, well I'll get that fixed13:58
HurricaneHarrynice, nice and nice, but how do I recover ?13:58
chrisccoulsonjdstrand, thanks13:59
jdstrandHurricaneHarry: ps auxww|grep cut13:59
jdstrandHurricaneHarry: then kill the offending cut13:59
jdstrandHurricaneHarry: you can then update /var/lib/dpkg/info/firefox.postinst to have:14:00
jdstrandubuntu_major=`lsb_release -rs | cut -d '.' -f 1`14:00
jdstrandubuntu_minor=`lsb_release -rs | cut -d '.' -f 2`14:00
jdstrandHurricaneHarry: then run 'sudo dpkg --configure -a ; sudo apt-get -f install'14:01
jdstrandchrisccoulson: so can we trigger a new dailies build manually?14:01
jdstrandchrisccoulson: otherwise we should probably delete those from the ppa14:02
chrisccoulsonjdstrand, fta probably can ;)14:02
jdstrandwell, we can do the latter regardless14:02
jdstrandchrisccoulson: I can tell you what happened. I made the commit knowing that the dailies wouldn't be built for a while. then I went to test and couldn't build maverick :( then I said 'I'll look at why it isn't building in the morning' forgeting that I made a change I didn't test :(14:04
* jdstrand puts brown bag securely around face and neck14:04
HurricaneHarryjdstrand: thanks, worked like a charm.14:05
jdstrandHurricaneHarry: glad to hear14:06
jdstrandfta2: can we trigger a dailies rebuild and/or delete the offending firefox packages? I don't have power to delete in there14:07
gnomefreakfirefox is taking forever to setup14:16
gnomefreak> 5mins14:16
fta2kill the cut14:20
fta2jdstrand, respining umd14:21
jdstrandgnomefreak: see my discussion with HurricaneHarry ^14:25
jdstrandfta2: thanks14:25
gnomefreakjdstrand: thanks, i just have to remember the command i need :)14:25
gnomefreak!lock14:26
ubot2Factoid 'lock' not found14:26
gnomefreak!dpkglock14:26
ubot2Factoid 'dpkglock' not found14:26
gnomefreakjdstrand: thanks that worked14:35
jdstrandgnomefreak: glad to hear. sorry for the problem14:35
gnomefreakjdstrand: no problem14:35
micahgchrisccoulson: so about that screensaver bug, we already have 2 bugs for that are linked upstream, I was going to dupe it against the non-full screen one which upstream isn't sure if they want to do yet15:38
chrisccoulsonmicahg - ok, no worries.15:38
chrisccoulsonfor the non full-screen case, i'm going to add it to my power-management extension15:38
chrisccoulsonit should be quite trivial to do15:38
chrisccoulsonalthough it won't ever work for flash15:38
micahgchrisccoulson: full screen makes more sense actually than non-full screen15:47
micahgchrisccoulson: BTW, I'll upload Thunderbird later tonight, I think the issue is just with my system for the upgrade15:48
micahgthe live CD for 10.04 had no issue when I added my PPA15:48
chrisccoulsonmicahg - cool, that's good :)15:51
micahgchrisccoulson: I found out that I need the symlink patch from ff36 and I respun the builds in my PPA15:52
micahgchrisccoulson: upstream was supposed to have committed the fix, but they backed it out15:52
chrisccoulsonmicahg - i'm going to start fixing the lp:firefox branch tonight16:27
chrisccoulsonit might be a good idea for us to not commit to it whilst i'm fixing it, so i don't have to keep rebasing my local branch16:28
chrisccoulsonjdstrand, are you still pushing changes to lp:firefox?16:29
jdstrandchrisccoulson: I'm done16:29
chrisccoulsonjdstrand, cool, thanks16:29
micahgchrisccoulson: k, I'll leave it to you then to manage daily breakage until you tell me17:08
asacInstalling new version of config file /etc/apparmor.d/usr.bin.firefox ...17:52
asacupdate-alternatives: using /usr/bin/firefox to provide /usr/bin/x-www-browser (x-www-browser) in auto mode.17:52
asac:(((17:52
asacmy upgrade hangs there for 5 hours now17:52
micahgasac: should be fixed in umd217:52
asacoh that was a bug on our side?17:52
asachell17:52
micahgasac: yes17:53
asaci cannot even kill it using ctrl+c17:53
jdstrandasac: yes, so backscroll at 07:59 < jdstrand>17:53
jdstrands/so/see/17:53
* micahg will bbiab17:53
jdstrandasac: assuming you are using a daily for firefox (3.6)17:53
jdstrandasac: it should be fixed in the dailies now (after apt-get update)17:54
asacjdstrand: ah it was you ;)17:54
jdstrandit was :(17:54
asacthats the result from rejecting qws :-P17:54
jdstrandhehe17:54
asacit always comes back ;)17:54
=== almaisan-away is now known as al-maisan
chrisccoulsonhmmmm, why have i suddenly started noticing triagers coming along and marking bugs incomplete without asking for any information?18:35
chrisccoulsonweird :/18:35
=== gavin__ is now known as gavin
asacumd218:56
asacis still hanging for me18:56
asacdo i need to do some cleanup from previous upgrade?18:56
asacjdstrand: ?18:56
micahgasac: is it that bad if I upload enigmail tonight with -fshort-wchar and figure out later why it's needed?18:57
micahgchrisccoulson: BTW, it's still up in the air if we actually need the new version of  NSPR for the next update18:59
micahgchrisccoulson: NSS is no longer needed18:59
micahgchrisccoulson: but they're updating18:59
jdstrandasac: you need umd319:00
jdstrandhttps://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa/+packages19:00
chrisccoulsonmicahg - oh, they reverted the change?19:01
micahgchrisccoulson: no, we're just not sure why we need it since the issue was there before, I'm going to post one of the UMD logs upstream and maybe they can tell us19:01
micahgs/issue/variable/19:01
asacjdstrand: do i need to remove umd2 first ?19:04
asacor do some cleanup?19:04
jdstrandasac: well, if you manually edited var/lib/dpkg/info/firefox.postinst, no. otherwise I'm not sure19:05
jdstrandasac: let me rephrase. as long as dpkg is happy, you don't have to do anything19:05
micahgchrisccoulson: still might be a good idea to update maverick since NSS includes new root certs I think19:07
asacjdstrand: ok i can confirm that it recovered by just dist-upgrade19:17
jdstrandcool, thanks19:18
chrisccoulsongrrrr, hate nspluginwrapper19:44
asacmicahg: back to your question. if its the case that you need to add short-wchar, go ahead. but i really think its a) auto added to the .pc files ... and b) auto added by the confiugre.in ... so something must have changed and we should check why was that etc.19:44
asacchrisccoulson: right ;)19:44
chrisccoulsoni should try and write something that actually works19:45
asacmicahg: e.g. what change made that flag go away on amd64 ... maybe upstream think toolchain was fixed to not need it etc.19:45
asacchrisccoulson: heh. you should make it so that the OOPP process can be 32-bit ;)19:45
micahgasac: it was added to the xul 192 build system, but I don't see it in the nspr.oc19:45
micahg*.pc19:45
asacmicahg: it was never in nspr.pc i think ... only in libxul.pc etc.19:45
asacand in the mozilla main build system19:46
chrisccoulsonasac - i wonder if that's possible already?19:46
asaclike confiugre.ac or so19:46
micahgasac: ah, maybe it never got into our tarball then, I'll check later19:46
* chrisccoulson will try the 32-bit plugin-container on my box19:46
micahgasac: I just want to get it uploaded so people can start testing19:46
asacchrisccoulson: i think it cant be that bad. though i dont know how they rpc is done. if its just binary datastructures it might not work, but if its a real protocol i dont see why not19:46
chrisccoulsonasac - i'll try it in a bit19:46
asacchrisccoulson: yeah ;) ... then we do a two run build and fix firefox to choose the right plugin container ;)19:47
chrisccoulsonif it works then we should just offer a solution based around that and get rid of nspluginwrapper entirely19:47
micahghas multiarch landed yet?19:47
asacbased on the plugin needed ... though i am not sure what happens with two different containers ;)19:47
asacchrisccoulson: yeah. thats what i am thinking. kill nspluginwrapper by using OOPP for that19:47
asacmicahg: no. but that shouldnt stop us i would hope ;)19:48
asacmozilla always spans their own micro cosmos ... most likely we would just also go for in-source gtk+ ;)19:48
asacwhich probably would be embraced upstream instantaneously ;)19:48
asacj.k19:48
micahgyeah, but we would need both 32 bit and 64 bit and have too choose the right one based on the plugin19:48
asacsince when is mconner gone from our channel?19:49
asac;)19:49
asacseems the war is over :-P19:49
asacmicahg: right. thats what i ment above with "choose the right plugin container"19:49
micahgok :)19:49
micahgif that could be upstreamed, they could have 64 bit FF with 32 bit flash too :)19:50
micahgon windows I mean19:50
asacyeah. would be a major contribution ;)19:50
asacgood stuff19:50
asacbut i dont see us getting that done :-P19:50
asacthough now with mighty chrisccoulson it might really happen19:50
chrisccoulsonheh :)19:50
chrisccoulsonbbiab, dinner time for me19:50
asackk i am off too soon19:51
rsavoyegnash runs great 64 bit, don't know why Adobe can't figure this out :-)21:29
Dimmuxxthe 64bit alpha flash works great too21:33
micahgDimmuxx: no longer developed21:33
DimmuxxI know but still, it works great :P21:34
rsavoyeGnash is your only hope :-)21:34
micahgrsavoye: well, chrisccoulson might get a 32 but plugin container working ;)21:34
rsavoyewe even still support 64bit ppc21:35
micahg*bit21:35
chrisccoulsonheh, i'll certainly have a go ;)21:35
=== BUGabundo is now known as BUGa_OngBak3
chrisccoulsonhmmm, ubufox upgrades from lucid -> maverick are broken21:47
chrisccoulsoni've just seen 2 identical bug reports now21:47
rsavoyeI'd think ubufox is unnecessary these days since Firefox lets you switch now21:49
Dimmuxxwhat's ubufox?22:02
chrisccoulsonubufox is the extension we ship which provides the plugin finder integration with apt, restart notifications and dynamic homepage URL22:06
chrisccoulsoni can't remember what else it does ;)22:06
Dimmuxxah22:07
chrisccoulsonjdstrand, not sure if you want to sponsor nss/nspr when you get some time :)22:32
=== yofel_ is now known as yofel
=== BUGa_OngBak3 is now known as BUGabundo

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!