/srv/irclogs.ubuntu.com/2008/03/11/#ubuntu-mozillateam.txt

dmbhello02:31
dmbfirefox 3 b4 is out :D:D!02:33
asachmm b4 as a bad regression for me08:50
asaci cannot close ephy ... nor can i close tabs in ephy08:50
asacUbulette: ^^^08:56
asac[reed]: i have a bad regression here :) ... epiphany cannot getService the login manager anymore. any idea which commit might have introduced this?10:28
asachmm ... might be my fault10:29
asaclets check10:29
asacno luck10:41
OurgonUbulette: empty white pages in seamonkey-2.0 from your PPA - does that ring a bell?12:42
asacdamn epiphany regression :(13:27
asaceverything else looks  good13:27
asacwhats going on13:27
=== Greenery_ is now known as Greenery
Greeneryyou guys have ppa for beta 4? cos I'm still on Gutsy15:14
[reed]asac: yes15:28
[reed]that sounds familiar15:28
asac[reed]: it looks that the regression appears to be a freed contentviewer still living in docshell15:29
asacon closing tab15:29
asacnot yet sure whats going on15:29
asactried to backout a patch against docshell/ tree that sounded familiar, but no result15:30
[reed]is it only the login manager?15:30
asacno the login manager was my problem15:30
[reed]ah, k15:30
asaci had a hacked version and did a bad conflict resolvation15:30
[reed]k15:30
asachttp://paste.ubuntu.com/558015:31
asacthats the crash that i see if i close an ephy tab15:32
asacin beta 3 it worked15:32
asachtmlDoc is freeded apparently15:32
asac[reed]: do you remember a patch that tried to resolve some memory/leakage issue?15:39
[reed]uh, there's been a lot of them15:40
asacno ... i mean between b3 and b4 :)15:40
asacmost likely in the content/docshell/view area :)15:40
[reed]yeah15:42
[reed]I'll ask15:42
[reed]and look15:42
[reed]let me catch up on bugmail15:43
asac[reed]: when did trunk open for b4 development? (so i can checkout a version in the middle :))15:44
[reed]looking15:44
[reed]tagged at 2008-03-03 11:15 PST15:44
[reed]oh15:45
[reed]b4 development15:45
[reed]one sec15:45
[reed]tagged at 2008-02-04 17:20 PST15:45
[reed]so, sometime between 2008-02-04 17:20 PST and 2008-03-03 11:15 PST15:45
[reed]checkout in the middle?15:45
asacyeah15:46
asac[reed]: is there anything special i have to take care of ?15:46
[reed]special?15:46
[reed]such as what?15:46
asacor just MOZ_CO_DATE="..." checkout ?15:46
[reed]that should work15:46
asacno ... to release any tag i have in my tree :)15:46
[reed]ah15:47
[reed]-A releases tags15:47
asaci know that thats true for cvs15:47
[reed]but MOZ_CO_DATE should work15:47
asacbut how to tell client.mk15:47
asacok15:47
asacwhats the middle15:47
asac?15:47
asac19th deb?15:47
[reed]deb?15:49
asachmm15:49
asacfeb15:49
[reed]ah15:49
[reed]17th or so?15:49
[reed]something like that15:50
[reed]binary search, woo!15:50
asacyeah15:52
asacbut pretty slow15:52
asacunfortunately there are no daily xul builds15:52
asac[reed]: cvs up -A client.mk didn't help15:52
asacjust MOZ_CO_DATE ... didn't help (as it still carried the release tag)15:52
[reed]hmm15:52
asac[reed]: oh sorry -A worked :)15:53
[reed]http://developer.mozilla.org/en/docs/Mozilla_Source_Code_(CVS)#Specific_Time15:53
asaci am c15:53
asacgetting the 20th now (hopefully)15:53
asaci don't think binary-search-compiling mozilla trees is a really fruitful approach :(15:57
asacbut well ... here we go. first spin :)15:58
asachow many spins do i need at max to find the proper date?15:58
asacand how many to find the proper checkin (how many checkins are average per day?)15:58
armin76500016:15
asac[reed]: can you please commit a bunch of approved1.8.0.15 patches? at best all distributed over a few days ... so we can find regressions in the daily builds ;) ?16:49
[reed]asac: I committed a ton last week16:56
[reed]er16:56
[reed]Friday night16:56
[reed]I'll take a look later16:56
asac[reed]: ok thanks17:35
asac[reed]: its good to distribute them over multiple days i get17:36
asacwe approved a bunch more today17:36
asacso there should be work todo ;)17:36
asac[reed]: any idea if i can get email notification if bonsai burns for a certain branch?17:36
asacaeh tinderbox i mean ;)17:37
[reed]no such thing, no... you could set up buildbot yourself to do that :p17:37
[reed]or write a script to parse tinderbox17:37
armin76lolz17:38
asac[reed]: yeah. i think at some point we need our own buildbot anyway17:38
asac(i think we have ~4 weeks time until they go offline)17:39
[reed]you better hurry ;)17:39
asacin worst case i will do daily PPA builds17:39
asacbut that is just debs17:39
asacwould love to build proper tarballs17:40
shirishasac: any possibility of ffb4 hitting hardy?17:51
armin76no17:51
shirish:(17:51
shirisharmin76: don'te tell me u asked the same question/query :)17:52
armin76he is working for microsoft now17:52
armin76so he doesn't care17:52
Greenery>.<17:52
shirisharmin76: are u serious, asac is working for MS ?17:52
armin76hahaha17:53
armin76no :)17:53
shirishthank god, that would be the day17:53
shirisharmin76: what about ffb4 hitting the official repos?17:53
armin76no idea17:54
armin76i don't use ubuntu, so don't ask me :)17:54
shirisharmin76: u're a BSD user?17:54
shirishor a gentoo one17:54
armin76lol17:54
armin76so, if i don't use ubuntu, i use bsd? great :P17:55
armin76being a gentoo dev what do you think? :)17:55
asacshirish: the package is ready17:55
asaci just have a serious regression for epiphany17:55
shirishasac: cool, is it in queue17:55
asacthats why i haven't uploaded yet17:55
shirishah ok17:55
asaconce thats sorted out i will upload17:55
asacbut no ETA yet17:55
shirishasac: sure that would be nice.17:55
asaci can upload the package to PPA if you want17:55
asaceverything except ephy works well17:56
shirishI love ephy as well :)17:56
asac[reed]: ok looks like 20th feb already had that regression17:56
asacnext build ;)17:57
asac12th? or 10th=17:57
[reed]:)17:57
asac?17:57
asacyou decide17:57
asacgive me a good guess ;)17:57
shirishasac: I went to the epiphany web-site but curiously the site is pretty parse17:57
shirishdoesn't have much content.17:57
armin761117:58
[reed]yeah17:58
[reed]11!17:58
asacok ... 11 is a reasonable day17:58
armin76am i being helpful? :D17:58
asacarmin76: no idea ... does gentoo have epiphany on top of xul yet?17:58
shirishah so you guys are working backwards to see if u find a build that works with that.17:58
asacif so you could test other dates ... so we can narrow down the regressoin window quicker17:59
armin76does it even work? :P17:59
asacshirish: yes ... finding regression window17:59
asacand finall finding the commit that introduced this is the primary objective right now17:59
armin76hrm...but what version of xul? 1.8 or 1.9?17:59
shirishok cool, I'll be away for a while but will be back17:59
asacarmin76: 1.9 obviously17:59
shirishok that's cool enough.17:59
armin76what ephy ver?17:59
asacarmin76: epiphany upstream has committed my xul patch yesterday17:59
asacso it should build out of the box18:00
armin76oh...then no, it doesn't build on top of xul, only with 1.818:00
asacarmin76: yeah. you should move ahead now that ephy trunk supports 1.918:00
asacotherwise gentoo will suffer from out-of-dateness18:00
asac :)18:00
asacdon't you think?18:01
armin76that's up to the gnome team :P18:01
asacget them in here18:01
armin76already pinged the guy :P18:01
asac[reed]: why doesn't export GRE_PATH=/path/to/my/xul/ ... not work?18:02
asacis it GRE_DIR?18:02
[reed]sorry, in Firefox 3/Gecko 1.9 call18:02
asack18:03
asacraise the regression thing there :) ... <rant>this is completely unacceptable </rant>18:03
armin76asac: leio, leio: asac18:06
asachey leio  :)18:06
asacleio: latest epiphany head now supports xulrunner 1.918:07
asacstill some bugs to flesh out, but definitly usable18:07
asacwonder if you plan to switch to xul 1.9 soonish18:08
armin76he fears you :P18:10
asacwhy?18:10
asacam i too verbose?18:10
armin76haha, no idea, i was kidding :)18:10
leioasac: it doesn't seem like we specify the version, so doesn't it just pick up 1.9?18:12
asacleio: no it won't18:19
asacyou need to specify another gecko engine:18:19
asac--with-engine=mozilla --with-gecko=libxul-embedding18:19
asaci think it will use that in case ... but only as last option18:19
leioepiphany is about to enter the official tree, while xulrunner isn't there yet, so it's a hard sell right now, but once it can be done, sure, I'm at least interested in doing that. And also interested in making webkit a choice for the user18:21
leiohaving an epiphany using xulrunner-1.9 in the same overlay and/or visibility might be a good idea though18:22
asacwhats needed to get xulrunner in the official tree?18:23
asacshould be a reasonable decisoin to build firefox on top of xulrunner18:23
asacarmin76: ??18:23
asacdidn't you do that already?18:23
armin76a woot18:23
armin76no18:23
armin76well18:23
leioit'll break epiphany-2.18 that has to stick around18:23
armin76i didn't put it yet because it broke a lot of stuff18:23
leiobut something might be possible to work out, such as appropriate blockers and ensuring that ephy-2.22 and xul-1.9 will stay at the same visibility level at all times18:24
leio(an ephy revision that is made to use xulrunner-1.9 mind you, a xulrunner-1.8 and 1.9 user can co-exist, just at different visibilities - what you'd call testing and stable tree)18:26
asacsorry ...  i don't know how and if you do releases at all?18:26
leiowe have a moving tree. There is no dist-upgrade alike thing18:26
asachmm. ok18:27
asacmust be hard to do a "complete gnome" migration for instance18:27
asace.g. from 2.18 -> 2.2218:27
leioand then approx. twice a year a snapshot of the package tree (the ebuild instructions of building them) is taken, and shaped into a package tree that will get used to build installation media18:27
leioyou get 2.18 -> 2.22 if you haven't updated stuff for over a year18:28
leioas soon as a GNOME release comes out and all bad kinks are worked out, we add it to the tree with ~arch keywords (all arches in an equivalent of your testing tree), then in about 30 days architecture teams upgrade the keyword to stable by GNOME teams request. Then stable users can upgrade immediately, and not when a new debian stable or ubuntu release comes out and they go dist-upgrading18:30
leioso that's the principles :)18:30
asacok18:30
asacmakes sense18:30
leioso in the same package tree we typically have about 2-3 epiphany versions18:30
leioand xulrunner lives in the same tree, and they shouldn't break eachother18:31
armin76yikes 39 pkgs reverse rdepend on mozilla-firefox!18:31
leioit'd be perfect if xulrunner-1.8 and 1.9 could simply coexist side by side18:31
asacah ok18:31
armin76and stuff still depend on firefox and not on xul18:31
asacleio: they can coexist18:31
leiothen armin76 should make that happen in gentoo and there are no problems18:31
asac(both binary + source)18:31
asacwe use a different pkglibdir for xulrunner-1.918:32
leiojust a different SLOT, we slot depend on the one we build against and that's it18:32
asace.g. /usr/lib/xulrunner (1.8)18:32
armin76haha18:32
asac/usr/lib/xulrunner-1.9b3 (1.9b3)18:32
leiohe'll have fun with all the USE=firefox stuff though18:32
asacUSE?18:32
armin76http://tinderbox.dev.gentoo.org/misc/rindex/www-client/mozilla-firefox18:33
armin76:(18:33
asacfirefox can also coexist18:33
asacwe have firefox-3.0 and firefox package ... both can be installed in parallel18:33
leioprobably won't be a pretty sight in the menus and such18:33
asacpretty short list :) ... we have patches for all of them in ubuntu.18:34
asacarmin76: could help to clean those patches up so they don't break builds against 1.818:34
asac:)18:34
asacmostly breakage of 1.8 builds with our patches is considered a bug/glitch18:35
asacUbulette: we have a reporter.jar in xulrunner and firefox?18:45
asacthought we don't build with reporter at all18:45
asacwhats going on?18:45
asacUbulette: btw, i think nspr.head and nspr are _unrelated_ branches18:47
asaci couldn't diff them ... and ofc ourse couldn't merge them so i had to duplicate what you already did18:48
Ubulettehi18:53
asachi18:55
Ubulettenspr is the one i used for the last hardy upload, head is HEAD (a bit behind as I don't really have time those days)18:55
asacUbulette: ok ... yes. i think i updated the correct one then18:55
asacplease take a look18:55
asac(nss + nspr)18:55
asaci didn't try to merge down from nss.head ... but i guess merging down after the release should be fine as well18:56
asac(as long as we merge soon)18:56
asacUbulette: anyway ... shouldn't nspr.head be based on nspr?18:56
asaccurrently they don't have a common ancestor at all from what i see18:56
Ubulettethey only diverge because of the naming ~cvs vs ~1.9b18:57
Ubuletteare you sure nspr for b4 is still 4.7.0 and not 4.7.1 beta ?18:57
asachmm.18:57
asacmost likely wrong yes.18:57
asacbut lower version isn't really a disaster18:58
Ubulettebut pleeeeeese, don't up 4.7.1~1.9*18:58
asacno i won't18:58
asacso its fortunate :)18:58
asacwe can use 4.7.1~beta1~1.9* i guess18:58
Ubuletteeither use what we discussed yesterday or we need to discuss more18:59
asacis the above version ok?18:59
Ubulettethere's no such thing as beta118:59
asacthere is beta right :O)19:00
Ubulette<Ubulette> NSS_HEAD_BEFORE_RFC4507BIS19:00
Ubulette<Ubulette>  ** The format of the version string is19:00
Ubulette<Ubulette>  **     "<major version>.<minor version>[.<patch level>] [<Beta>]"19:00
Ubulettejust the last 219:00
asacwhat does that tell me?19:01
Ubulettebut as i said, mozclient is doomed with that19:01
Ubulettejust read my new nspr.mk.in19:01
asac_reconnect19:03
asac_reconnect19:03
asac_Ubulette: do you see epiphany-browser regression?19:04
Ubulette<asac> what does that tell me?19:04
Ubulette<Ubulette> but as i said, mozclient is doomed with that19:04
Ubulette<Ubulette> just read my new nspr.mk.in19:04
asac_ok19:04
Ubuletteyes, i told you twice earlier19:05
asac_really19:05
asac_hmm ... when did it start19:05
asac_my regression window is currently down to 2nd feb - 11th feb19:05
asac_(e.g. the 11th feb build still has that)19:06
asac_does that match with your experience?19:06
UbuletteFeb 04 22:00:44 <Ubulette>      epi doesn't quit on "File / Close"19:06
asac_e.g. that it happened quite early in the b4 release cycle19:06
asac_oh cool19:06
asac_so it happened between feb 2 and feb 4 :)19:06
UbuletteMar 01 02:13:30 <Ubulette>      asac, File / Close in Epiphany doesn't do anything for me.19:06
asac_for me it takes some time ... then it crashes19:07
asac_thanks19:07
asac_i should listen more carefully i gues19:07
Ubulettei guess so19:07
asac_ok lets see if its really not in 2nd feb19:08
asac_Bug 414894 - Remove content arena. r=smaug, sr=sicking, a=schrep19:12
=== asac_ is now known as asac
* RainCT just uploaded adblock-plus :)19:23
* asac hugs RainCT 19:29
asacany other extensions you are using? :)19:30
asacif 2nd feb build really fixes this, there shouldn't be much commits that might have caused the ephy extension.19:34
asacs/extension/regression/19:34
* asac should gracefully exit other contexts before writing to a channel ;)19:35
* RainCT hugs asac back :) (was away)19:47
UbuletteI hate when scrollkeeper-update starts during an upgrade20:00
asacyeah ... takes ages20:01
asacuninstall ;)20:01
Ubuletteit's within libcupsimage220:02
asacanyone has ipw2200 or ipw2100 ?20:02
asacor iwl3945?20:02
Ubuletteasac, xulrunner-1.9-1.9~b4~cvs20080215t0338+nobinonly/debian/tmp/usr/lib/xulrunner-1.9b4pre/chrome/reporter.jar20:04
Ubuletteseamonkey-1.1.8+nobinonly/debian/tmp/usr/lib/seamonkey/chrome/reporter.jar20:04
Ubulettefirefox-3.0-3.0~b4~rc1+nobinonly/debian/firefox-3.0/usr/lib/firefox-3.0b4/chrome/reporter.jar20:04
RainCTasac: btw, you probably already know, but just to be sure.. there are some extension packages for review on the firefox-launchpad project20:06
asacRainCT: yes. i planned to upgrade a few more in the archive and do them all in a row20:08
asacif you want to upload and file the appropriate bugs, we can certainly accelerate this :)20:08
Ubulettedamn, python-pysqlite2 also runs scrollkeeper-update20:10
RainCThm.. but they wouldn't need a mozillateam ack anyways. or?20:10
asacacking the report if the extension is pre-tested should be easy to do20:11
Ubuletteasac, why didn't you merge nss for nss.head ?20:50
Ubulettewell, nm. this versionning is a mess20:51
asacUbulette: one reason. the other i explained above (iirc)20:59
Ubulettenss was mergeable easily20:59
Ubulettenspr too20:59
asacnspr doesn't have a common ancestor (at least the branch i was looking at)21:00
Ubulettereally ?21:00
asacyes21:00
asacthats what i complained about initially21:00
asacnot that they diverge, but that they are not related at all21:00
Ubulettehere, they are21:01
Ubulettehttp://paste.ubuntu.com/5595/21:02
asaci looked at nspr.head21:02
asac(out of my mind)21:03
Ubulettethat's my nspr.dev locally21:03
Ubuletteold mess21:03
Ubulettenow it's even more difficult21:04
[reed]asac: CC me to all bugs from this query, please -- https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&keywords_type=nowords&keywords=fixed1.8.0.15%2C+verified1.8.0.15&chfieldto=Now&field0-0-0=flagtypes.name&type0-0-0=equals&value0-0-0=approval1.8.0.15%2B&field0-1-0=bug_group&type0-1-0=equals&value0-1-0=security21:45
[reed]security bugs that have approval1.8.0.15+ but haven't been checked-in21:46
armin76why don't cc yourself? :P21:47
armin76oh21:48
armin76its empty21:48
[reed]armin76: I'm not in the security group? :)21:51
[reed]I'm in two of the three security groups21:52
asac[reed]: can you give me a tinyurl?21:57
[reed]you can't use the url I posted above? :)21:57
asacterminal is too dump to alow to copy that long link21:57
[reed]lol21:57
[reed]k21:57
asacwell ... i could. but not easily21:57
[reed]http://tinyurl.com/29kell21:58
asacreed@reedloden.com?21:58
[reed]yes, or you can use :reed21:59
[reed]that matches me21:59
[reed]since my address will change when the summer starts21:59
armin76yay, why?21:59
[reed]reed@mozilla.com -- summer internship21:59
armin76about time :P22:00
[reed]heh22:00
[reed]I was an intern last summer, too :)22:00
asac[reed]: done22:00
asac4 bugs22:01
[reed]asac: thx22:01
asacthx 2 u :)(22:01
asac[reed]: i see 19 with approval+22:03
asacnot fixed22:03
Ubuletteasac, does the dell inspiron have (wired) ethernet ? I can't see it on their web site22:29
asachaven't heard of any laptop that doesnt22:30
asacthe ubuntu inspiron almost certainly has22:30
Ubuletteok, found it22:34
=== Ubulette_ is now known as Ubulette
asacdebian bug 47043823:44
ubotuDebian bug 470438 in iceape "iceape: FTBFS: (.text+0x526): undefined reference to `SECITEM_CopyItem_Util'" [Serious,Fixed] http://bugs.debian.org/47043823:44
Ubulettelibnss3-1d.symbols: SECITEM_CopyItem_Util@NSSUTIL_3.12 3.12.0~1.9b323:44
Ubulettein libnssutil3.so.1d23:45
Ubuletteasac, ^^23:45
asacyes23:45
asacjust wanted to see the title23:46
asacsomeone said icedove might be hit by that :)23:46
=== Ubulette is now known as fta
=== fta is now known as Ubulette
Ubulettedamned, taken23:48
asachmm ... try to figure how long not active23:49
asacif its longer than 2 month or so you can high-jack23:49
Ubulettehow could I know?23:49
asacnot sure .... you can ask an #op on freenode to high jack.23:50
asacthey can probably tell you inactivity as well23:50
asacbut i think there is a service running that you can use to take a look23:50
Ubuletteit's a registered nick23:51
asacyes23:51
Ubulettewell, maybe another day23:51
asacsure23:51
Ubulettemozilla bug 42212223:53

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