/srv/irclogs.ubuntu.com/2009/12/17/#ubuntu-mozillateam.txt

asacfta: http://launchpadlibrarian.net/36885088/buildlog_ubuntu-lucid-armel.chromium-browser_4.0.272.0~svn20091215r34555-0ubuntu1~ucd1_FAILEDTOBUILD.txt.gz :(00:20
asacfta: oh ... do you plan to uploda gyp to archive?00:24
asacwe need it in order to use porter machines00:25
asac(can only install whats in the rachive)00:25
asacthanks for considering :)00:25
asacand and ... enable verbose CC on armel too for now ;)00:29
micahgasac: looks like we get at least 1 more month with FF300:32
micahgsupport till end of Jan 201000:32
=== ripps|sleep is now known as ripps
copproNo Tunderbird 3.0 stable PPA or backport yet?01:00
micahgcoppro: closest thing is my beta PPA01:01
micahgcoppro: it's the release version but unbranded01:01
copprook, thanks01:01
Turlfta: ping02:14
=== asac_ is now known as asac
=== mac_v is now known as _
=== _ is now known as _7hills
=== _7hills is now known as mac_v
BUGabundo_workmorning11:02
=== c0p3rn1c is now known as copernic
ftaone french guy compared browsers on windows: http://www.n1fo.fr/2009/12/benchmark-navigateurs-internet-1ere-partie-sous-windows-le-16-decembre/11:22
ftaBUGabundo_work, ^^11:22
asacfrench guys shouldnt publish french content anymore ...11:29
asacthere is already enough french stuff out there11:30
asacnative content just means duplication of efforts and encourages stealing ;)11:30
asacrecently i tried to read a book in german ... that was a translated book ... i couldnt read it anymore after two pages11:31
asacso many things were odd and i saw that it would have been nice untranslated11:31
asacnever felt that way before11:31
asacso i assume it must have been a very bad translation11:31
asac(though it was from grisham)11:31
asac(though it was from grisham11:31
BUGabundo_workeheh11:36
BUGabundo_workgo watch anotehr TBBT episode11:36
BUGabundo_work1st time ch beats Safari11:38
BUGabundo_workeh11:38
ftaasac, seems that guy tested that by himself, it's not a translation, and he promises to test on ubuntu in a few days11:50
BUGabundo_workba bye Flash evil cookies rm -rf .macromedia/Flash_Player/11:52
ftabuy otherwise, i agree, native is better than translated. that's why i'd like to understand as many langs as possible :P11:53
BUGabundo_workasac: [reed] did Firefox "fixed" that bug where emptying Browser cookies would not remove the Flash ones?11:53
ftaasac, i see it failed on arm, strange, it worked everywhere else. was it a copy or did you tweak it?11:55
ftad'oh! http://paste.ubuntu.com/343502/12:08
fta 13:12:38 up 7 days, 20:51,  5 users,  load average: 14.70, 11.13, 8.1312:13
ftaoh my12:13
ftait's putting my quad core 3GHz on its knees for more than 30 min, clearly not acceptable12:27
fta50 min to link 4 binaries :P12:45
=== ikonia_ is now known as ikonia
asacfta: copy13:42
asacfta: you can actually testbuild for arm locally13:42
ftacan?13:42
ftacan I?13:42
asacqemu-arm-static13:42
asacits a package that allows you to setup an arm chroot13:42
asacthere you can do everything ... build, start apps etc13:42
asacspeed should be between native and cross-compile13:43
asacbut is not that bad i was told13:43
asaci will check that out too ;)13:43
asacaccording to ogra that package is really trivial to use13:44
asac/usr/bin/qemu-arm-static13:45
asac/usr/bin/build-arm-chroot13:45
asacthose are the binaries13:45
asac /usr/bin/build-arm-chroot13:45
asacI: usage: [OPTION]... <suite> <target> [<mirror> [<script>]]13:45
asacI: Try `debootstrap --help' for more information.13:45
asacE: You must specify a suite and a target.13:45
asacasac@hector:/srv/chroots$ sudo build-arm-chroot lucid lucid113:46
asacI: Retrieving Release13:46
asacI: Retrieving Packages13:46
asacI: Validating Packages13:46
asacI: Resolving dependencies of required packages...13:46
asacI: Resolving dependencies of base packages...13:46
asacI: Checking component main on http://ports.ubuntu.com/ubuntu-ports...13:46
asaccool. so works oob indeed13:46
asaclets see how well it works after its done ;)13:46
asacok starting over with sudo build-arm-chroot lucid lucid-arm113:47
asac;)13:47
ftadoing another test build with no more than one ld at a time, with make13:59
asack14:02
asacfta: is it committed?14:02
asaci would try this qemu spin then ;)14:03
ftanot yet14:03
ftahttp://paste.ubuntu.com/343555/14:03
ftaoops14:04
ftabad patch14:05
asacwill we get a new gyp too?14:05
ftahttp://paste.ubuntu.com/343557/14:06
ftayes, it's already in the ppa14:06
asacfta: too bad to keep scons working and just a flip/switch?14:06
asacat least until we are sure make is good?14:06
ftathat's why i didn't commit it yet, i want to verify that it works14:07
asachmm14:07
ftait will work for you too, don't worry14:08
* asac installs build-deps in chroot14:09
asackk14:09
asac;)14:09
asacwe need to get gyp in the archive. are there any license issues with that?14:09
asacor binaries in source etc.?14:09
ftanope, just python, 1 license, should be easy14:10
ftafor once ;)14:10
asacbug 49466714:15
ubottuLaunchpad bug 494667 in squashfs "[armel] non-ISO-C misaligned pointer punning causes slowness and SIGILLs" [Critical,In progress] https://launchpad.net/bugs/49466714:15
asacfta: http://paste.ubuntu.com/343593/ \o/14:47
asacdpkg-buildpackage: host architecture armel14:47
asaccrazy stuff14:47
asachot stuff ;)14:47
fta?14:48
asacfta: thats on mx x86 system ... just setup a chroot with build-arm-chroot14:48
asacdone14:48
asacworks as expected14:48
asaca few syscalls are not emulated as i seems, but didnt get any issues yet14:49
asaceven lzma is unpacking :)14:49
ftahm, make[1]: warning: -jN forced in submake: disabling jobserver mode.14:49
asacand python is indeed running at almost full hostspeed14:51
asac 7588 asac      20   0  137m  18m 2492 R  100  0.9   1:34.05 python14:51
ftawho's using google talk here? in pidgin?14:52
ftaBUGabundo_work, ^^ ?14:52
BUGabundo_worknot at this moment14:54
BUGabundo_workwhy ?14:54
ftanm, i figured it out14:55
asacfta: folks are moving to empathy ... which also supports various talk stuff14:56
ftai just used what is by default in karmic..14:56
asacempathy is default in karmic ;)14:56
asacwe dumped pidgin afaik14:57
ftai mean, the thing in indicator-applet14:57
asacright. ... both: pidgin and empathy support that indicator14:57
asacso it just shows what you have installed (probably from old times)14:57
ftahm i jsut have evo, gwibber and pidgin in there14:57
asac"Generating bindings ..." -> works :)14:57
asacfta: yeah. its a recommends of ubuntu-desktop14:58
asacguess you didnt get it since you have an old install14:58
asacGenerating binding from /var/builddir/asac/builds/chromium-browser-4.0.272.0~svn20091216r34708/build-tree/src/third_party/WebKit/WebCore/plugins/MimeType.idl15:01
asacis it more or less ok that one binding is about 1s ?15:01
asacwell. .. probably between 0.5 and 1s15:01
* asac waits for a .o being produced before dorpping out for housekeeping15:02
ftaasac, did anything happened wrt the action item to put chromium into the archives?15:04
fta-ed15:04
ftagood, 1 link at a time is better, ~1min per binary on my box15:08
asacfta: well. one thing i wanted to have was a working arm version ;)15:08
asacwe could do that in parallel ... maybe the beta build? do you have a branch where we could do licensing improvements on it for upload?15:09
ftathe upsteam bug move a bit since the beta15:13
ftawhich is ~1 month old now15:13
asacdidnt they push something new to the beta yet?15:16
asaclike an update?15:16
asaci would think we should push the beta with just i386 and amd64 and once arm is fixed on trunk we can upload that too15:16
asacand hope for another beta before rlease15:17
asacwould be kind of the _perfect_ scenario for me ;)15:17
jcastrothey pushed a beta yesterday iirc15:17
asaccool15:17
ftaok, 30 min to build everything, vs 50 min just to link binaries 4 by 415:18
asac30 min?15:18
asac:(15:18
ftathat's good, it was nearly 2h in my previous attempt15:18
asacare you sure that its all ok now?15:19
asacfeels a bit like a big chunk wasnt built at all ;)15:19
asaci mean ... its not like most energy was spend on scons before15:19
ftai'm letting it run15:19
asacthat would be really really good and bad for scons ;)15:19
asaci see the rampup time .. but then it shouldnt do much15:20
ftaupstream wiki reads: "Scons Tricks: Avoid Scons entirely"15:23
ftaand in the build instructions: "Make too fast? Wanna go slower? See LinuxSconsBuild for how to use the old Scons build."15:24
asaclol15:32
asaccool15:32
asacso if your build succeeds, please commit15:32
asaci want to abort this build and test ;)15:32
fta1234     19179 99.6  9.3 394348 379504 pts/2   R+   16:29   3:22 lzma -9c15:33
fta-9, there are crazy, it's useless, -7 is enough15:34
ftai wonder if there's a way to trick it15:34
asac*I* wonder if anyone actually tested usb-creator-gtk ;)15:37
ftai used something like this to install unr on my netbook15:40
asacyes15:40
asacbut current state is nasty15:40
asacdoesnt go well with udev for usb i guess15:40
BUGabundo_workasac: i've filed 3 bugs on lucid15:44
BUGabundo_workits broken even in karmic15:45
BUGabundo_work:(15:45
asacyes i am using karmic ;)15:45
asacthats why i asked15:45
BUGabundo_workkeeps failing to see the device15:45
BUGabundo_workor ask to format, and wont work *after* u format it15:45
BUGabundo_workbahh15:45
BUGabundo_workits as if it broke around release15:45
asacyes. usb device handling is really broken15:45
asacnow it works, but took me 5 attemps15:45
BUGabundo_worki used to use it before beta15:46
BUGabundo_workand it was ok15:46
BUGabundo_worklucid seems a bit better15:46
asaci used it in < karmic and it was ok ;)15:46
BUGabundo_workbut still depends on PC or pen15:46
BUGabundo_workyeah jaunty was nice15:46
asacif i remember i will poke whoever does that ;)15:46
ftanot sure i used that tool, it was a pygtk wrapper around dd15:47
asaci think its evand15:47
ftafull build was ok for me on x64, wondering if i commit now15:51
ftadailies in 10 min15:51
asacfta: thought we are doing those at 4am now?15:52
ftanot ucd15:52
asacucd are the builds that take most time ;)15:52
ftahopefully less now15:54
ftahttp://www.mail-archive.com/debian-lint-maint@lists.debian.org/msg10217.html  i wonder if it is still true15:56
ftacommitted15:59
asacok lets hope this flies then ;)16:05
asacaborted16:05
asacstarting over16:05
asac389.16:06
BUGabundo_worklol16:06
asacfta: no new enough gyp16:07
asac:(16:07
asacdarn ;)16:07
ftaeh?16:08
ftagyp (>= 0.1~svn769),16:08
fta0.1~svn769-0ubuntu1~ucd116:08
ftawfm16:08
ftaif you have 767, i told you yesterday it wasn't good enough16:09
asachmm16:09
asaclet me copy16:10
asacok copied16:11
ftawhere? in your chroot or the ppa?16:18
ftadon't use the patch i gave you, use the branch16:19
ftachromium-codecs-ffmpeg (0.5+svn20091210r34297+34315+34632 -> 0.5+svn20091210r34297+34315+34829) [3.23MB (+0kB, +0.00%)]16:20
ftaas expected, i have to rebuild because they touched build/ which is not ffmpeg related :(16:20
asacfta: copied to native ppa16:21
asacand waited till it was there16:21
ftahttp://identi.ca/notice/1691999516:21
asacinteresting16:21
micahgasac: what's our policy for FF extension suppport?16:25
asacTBD16:26
asacaccording to new support model spec16:26
asacwe try to go to a more slink approach16:27
asace.g. only extensions that are essential ... or have native components (e.g. benefit from our multi-archs)16:27
micahgasac: what do I do with a bug now where the user has installed from amo but we package an older version?16:27
asacuser install wins over global install16:27
asacso its not a bug of our version if he sees it16:28
asacwishlist bugs are probably wont fix for extensions that have no natie component16:28
asacnative component16:28
micahgasac: I know, but do we just point the user to amo or reassign to our package or both?16:28
asacmicahg: point user to amo and reassign to our package as wont fix16:31
asacyes16:31
asaci dont know where folks are supposed to file bugs16:31
asaci think there is no single way of filing bugs16:31
asacsome extensions probably dont even have a contact ;)16:31
micahgasac: right16:32
asac[reed] can find out more about amo policy and requirements and escalation paths16:32
asacor knows it right away (as usual)16:32
micahgooh, they have an actual bug tracker for torbutton :)16:33
asachaha16:34
ftaEmpathy, no protocol installed, you have to 1st install a backend for each protocol16:35
ftawtf?16:36
asacmaybe16:36
asacdid you disable automatic recommends?16:36
asacmaybe thats the prob16:36
ftadid i install am empty shell? :)16:36
asacarent there even basic selection of protocols ?16:36
ftanothing16:36
asacRecommends: telepathy-gabble, telepathy-salut, telepathy-haze, telepathy-butterfly, gvfs-backends16:36
asacnot that i would know what is in there ;)16:37
asacexcept for gabble which afaik is a protocol/service16:37
ftai just have gvfs-backends16:37
asaclol16:38
asacempathy-megaphone-applet16:38
asacempathy-megaphone-applet - High-level library and user-interface for Telepathy (megaphone applet)16:38
asactelepathy-gnome ?16:38
asacmaybe that?16:38
ftatelepathy-haze16:38
asachmm thats i nuniverse16:38
asacok16:38
asacso you found it ;)16:39
asachmmm ... seeing a bunch of gcc <defunct> popping up16:39
asacwhile building16:39
asacthey go away16:39
asacbut still i can see them from time to time in top16:39
ftadoes it look faster?16:40
asacit feels faster16:40
asacmaybe scons took 0.25s for each .o16:41
ftalet's hope there's no arm specific death trap16:41
asacfta: hmm ... is out/Release/obj.target/webcore_bindings/third_party/WebKit/WebCore/bindings/v8/DerivedSourcesAllInOne.o the first file to biuld?16:58
asaci only had RULE output before ... though that was .o production16:58
asacbut seems like it just started now16:58
ftauhh16:58
asacgood. so i think the make doesnt make a big difference ;)17:00
asacthe first that .o took like a minute to build17:00
ftapastebinit broken17:04
ftathe new paste.ubuntu.com needs openid :P17:05
micahgasac: I just want to verify that bug 367827 is either a won't fix or needs to be reassigned17:08
ubottuLaunchpad bug 367827 in firefox-3.5 "Raise Firefox update-alternative priority for x-www-browser in order to overrule Epiphany" [Low,Triaged] https://launchpad.net/bugs/36782717:08
ftaasac, the builders are much slower than my serv :P "1 hour, 43 minutes" still building what took me only 30min18:00
asacwell. the testsuite is really slow on its own18:02
asaci mean ... i dont think you can get down to 30 minutes ;) .... maybe do another clean build and see if its really just 30 ;)18:02
ftaasac, the build part only18:14
asack18:25
ftaasac, you seem doubtful18:38
ftaasac, http://www.sofaraway.org/ubuntu/tmp/build.png18:39
ftaless than 35 min in that run, but i'm also doing something else on that box18:39
ftathat includes setting up the chroot, unpackaging, etc18:40
ftadpkg-buildpackage: full upload (original source is included)18:55
ftadone18:55
ftathat's 59min total18:55
ftaincluding lintian18:56
ftaasac, i see, the builders are all stuck on ld. not enough memory then, even just one link at a time19:11
ftanothing i can do19:11
ftamaybe use gold on karmic/lucid19:11
ftanot sure about the status of gold on arm19:12
bdrung_asac: can you review https://code.launchpad.net/~mozillateam/mozilla-devscripts/dh_xul-ext ?19:49
and`asac: hey, how do you manage multiple alias with mutt? e.g your Ubuntu mail, your Debian mail etc, all in one account?20:23
and`I hope there is a way to do it without having to set up multiple single accounts20:24
ftamultiple from addresses? or just recognize your multiple emails as you?20:38
ftafor the 1st, i set "my_hdr From: foo bar <foo@bar>" as default, and i use send-hook to change it + hostname based on the recipients of my message20:40
ftafor the 2nd, i have a giant reqex in "alternates" matching all my emails20:41
ftaand`, ^^20:41
and`fta: like recognizing my multiple mail aliases20:41
and`of course they aren't real mailboxes20:42
and`but just aliases20:42
ftaaliases for recipients?20:42
ftai have hundreds20:42
and`nono, my personal aliases, I mean my Debian mail, my Ubuntu mail etc20:43
and`was having a look at the alternates option20:43
ftaalternates is just to identify the email as from you / to you / Cc you in the index pane20:44
and`ah damn, so that's not what I really want20:45
and`I just want to be able to send mail from my mail aliases20:45
ftasend-hook is what you need to automatically change the from depending on who you're writing to20:45
ftathen it's send-hook20:45
and`send-hook should be the answer then20:46
and`need to read some docs20:46
and`about how to set it up properly20:46
ftaand`, http://paste.ubuntu.com/343675/  something like this20:47
ftasetting hostname changes the message-id too20:48
ftayou can set as many send-hooks as you want, i have ~30 for mailing lists, work, etc20:50
and`fta: thanks for the examples, one question, if I want to use one of my mails to send a mail to friend or to anyone else which is not into the send-hook table, it won't work?20:50
ftadefault is first, so the last to match will be used20:50
and`ok, great, thanks for the hints20:51
and`ok, great, thanks for the hints --> setting up now :)20:51
ftai use "set edit_hdrs" so i can change the from whenever i want while i write the email20:52
and`sweet, that's what I was searching for20:54
ftathe hooks are really handy, especially with a dozen of emails20:55
ftai've use mutt since 1996 as my unique mailer, up to earlier this year (i moved to evolution)20:57
and`why such change?20:58
and`fta: ^^21:03
ftawork mostly, where everyone uses outlook, mutt breaks the crazy "inline" style21:03
ftaevo has an inline style21:04
ftaa lot of people i'm working with don't understand the default quoted style21:04
and`ah, you are unlucky then : /21:05
ftaand they are lost when i answer in the text and drop the unneeded parts21:05
and`explain them how it works, it's not that hard :)21:06
ftai work for a multinational company21:06
ftaso the short answer is: no way21:07
and`you could use both evo and mutt21:07
and`one at work, other at home21:07
ftai kind of like evo now21:12
ftai had to patch a few things but it's mostly ok now21:13
ftai just miss a way to have the contacts & agenda synced21:14
ftaand gnome 60261221:15
ubottuGnome bug 602612 in Mailer "Reply to As.. (Quoted / Inline) from the menu" [Enhancement,Unconfirmed] http://bugzilla.gnome.org/show_bug.cgi?id=60261221:15
and`fta: since I'm registered with my main address on quite all mailings lists21:17
and`I set edit_hdrs21:17
and`and it did the work21:17
and`e.g I manually changed the From address21:17
and`and it sent me the mail using my alias21:18
and`fta: thanks for the right hint :)21:20
micahgfta: can you make the mozilla team an admin for the prism project?21:48
ftamicahg, done22:04
micahgfta: thanks22:08
ftaasac, just changed lzma for the debs from -9 to -7, it now just takes a few sec for the huge -dbg22:14
ftaasac, and i hooked up some perf lines in the logs: http://paste.ubuntu.com/343687/22:15
asacfta: good. also i pushed todays daily to the ppa again ;)22:38
asacafter not seeing much progress on this qemu thing ;)=22:38
asacwould be interesting to see how well it performs on your sys with lots of mem etc.22:38
asachttps://edge.launchpad.net/~ubuntu-mozilla-security/+archive/nss3.12.3/+build/140280322:40
* asac off watching movie22:41
fta  9013262824292842194 "Chromium requires Windows Vista or Windo..." el,gu,zh-TW,sw,ca,am and 45 more22:46
ftalol22:46
ftaasac, http://code.google.com/p/chromium/issues/detail?id=28291  patches for the license bug22:49
ftatiny22:50
micahgfta: do you see any problems renaming the prism branch prism.head?23:14
ftanone, but i need to update the bot23:17
ftanot today, i'm off23:17
micahgfta: ok, let me know when a good time is23:18
micahgfta: I also wanted to bind it to the trunk development series in LP, that's why I asked for the team to be admins23:18
micahgbbiab23:19
ftaasac, more than a few sec actually, http://paste.ubuntu.com/343704/  (that's the -dbg deb)23:40
ftamicahg, http://www.workswithu.com/2009/12/15/pushing-prism-on-ubuntu/23:53
ftanow i'm really off23:54

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