/srv/irclogs.ubuntu.com/2009/09/02/#ubuntu-mozillateam.txt

ftai wish gwibber had a way to show URLs fully resolved, like the Attachments in the identi.ca web pages00:00
BUGabundome too00:02
BUGabundoI hate short links00:02
BUGabundoI think I have a bug for that00:03
ftahm, sometimes, when i want to submit a dent, gwibber just adds a carriage return in the textfield but doesn't send anything00:08
BUGabundoeheh00:12
BUGabundoI'm back on 1.200:12
BUGabundoI couldn't take it anymore00:12
ftahttp://arstechnica.com/tech-policy/news/2009/09/does-less-evening-internet-mean-europeans-lead-better-lives.ars00:38
andvasac, all-in didnt appear on NEW01:31
andvasac, plus I received no NEW mail01:31
andvasac, something went wrong then01:32
andvgoing to sleep01:32
andvcya tomorrow01:32
* asac goes and gets some coffee08:46
=== zbraniecki is now known as gandi
asacejat: there?09:45
asacbdrung: hi!09:45
asacbdrung: saw you got added to the team ... congrats ;)!09:45
asacbdrung: did your MOTU application work out well?09:53
fta_asac, aren't MOTU nearly dead? i mean, with the archive reorg in progress10:08
fta_d'oh! http://www.youtube.com/watch?v=0qqxjO5nr8k&feature=featured10:14
eagles0513875morning gnomefreak11:00
andvasac, hi11:08
andvasac, had a problem with the upload yesterday? :)11:09
asacandv: no orig11:26
asacits now done11:26
andvasac, yeah :)11:27
andvsaw it on new11:27
eagles0513875hey asac :)11:32
asachi eagles051387511:45
eagles0513875how goes it asac11:46
asacpretty good ... still catching up a bit on vacation stuff ;)11:48
gnomefreakok how the hell do i get normal gwibber back?12:23
bdrungasac: thanks. the MOTU application was deferred because they had no quorum12:27
asacbdrung: during meeting?12:28
asachmm12:28
asacgnomefreak: no way ... just keep using it ;)12:28
gnomefreakasac: i cant figure out how to send message12:29
gnomefreak<enter> brings you down to next line to type12:29
bdrungasac: no, before :)12:29
asacgnomefreak: i dented about that and got an answer12:29
asacgnomefreak: http://identi.ca/notice/927738912:30
gnomefreakthanks lookinhg12:30
asachttp://identi.ca/conversation/9276118#notice-927738912:30
gnomefreakI DONT LIKE IT12:30
gnomefreakasac: i cant move it up12:31
fta2there's a pref in the menu12:31
gnomefreaknot here there isnt12:32
gnomefreakat least not in pref12:32
gnomefreaki have same ui for new as was for old versions12:32
gnomefreakfor pref12:33
fta2not in pref, just View / Editor12:33
gnomefreakfta2: i can typw but i cant send it12:33
gnomefreaks/typw/type12:33
fta2it's enter, as before12:33
gnomefreakfta2: brings me down a line in editor12:33
fta2but sometimes, it just adds a carriage return12:33
fta2it's a bug12:33
fta2scroll up, and enter12:34
gnomefreaksame thing12:34
asacgnomefreak: you can drag it up ... just try a bit more ... its  tiny grab thing like where the separator line is right at bottom12:35
gnomefreakasac: not here it just brings the text back to editor as if it is not permitted12:36
asacwell then you already have the editor12:36
asacwhich was what i didnt have12:36
asacsending worked for me12:36
gnomefreaki have editor since i typed in it but no way to send it12:36
fta2iirc, View / Editor is disabled by default, which is bad12:37
asacgnomefreak: i hit enter and it works12:37
gnomefreaki have the editor that part i knew from other releases its sending it that isnt working12:37
asacfta2: for the searchplugins we actually should add replaces to all lower firefox versions on all firefox branches i think12:38
asacor we should get rid of the common firefox-addons/searchplugins dir12:39
fta2asac, or make a dedicated src package12:40
asacyeah12:40
gnomefreakcant clear window either let me see if there is an update for it12:40
gnomefreakbug in apturl too :(12:42
fta2asac, i like the dedicated source package, so we can add/remove/update searchplugins independently of ff12:45
asacfta2: technically i agree. problem is that it bears the risk that we miss if upstream changes their default searchplugins12:46
asacand we are obliged to have them unless we explicitly applied/communicated for a change12:47
asacwonder if there is a way to add a safety net12:47
asaclike "if we are package that ship default package, verify that the searchplugins in the other package are all the same"12:47
fta2we can just pull that from trunk, like today12:48
asacthats not what we do today12:51
asacwe pull them from the branch of the package installed12:51
asacthe package with highest version is supposed to win12:51
asacbut in default install we have the plugins of firefox-3.5 as shipped by upstream unless someone installs firefox-3.7 et al12:51
asacso i guess we should need to pull them from current default branch12:52
fta23.7 doesnt provide those searchplugins, just a link12:52
asacas thats where mozilla wants trademark being inforced12:52
asacfta2: yes. so i think what we should do is provide a firefox-searchplugins package which is produced from the same package where we have the meta package12:53
asacdoes that sound about right?12:53
asacand everything depends on firefox-searchplugins in turn12:53
fta2yes12:53
asacok so no standalone source package12:54
asacjust an unversioned package for the current METAPACKAGE thing12:54
asacat some point we probably want to assemble debian/control on-the-fly ;)12:54
asachehe12:54
asacjdstrand: so at best the apparmore profile feature would become independent from the firefox version by some template magic12:58
asaci would like to able to just land it in firefox-(3.0?)/3.5/3.6/3.712:58
asacjdstrand: ok ... so i think the general 3.* and 1.9* things can be kept that way12:59
asacbut the firefox-3.5.* would probably be better replaced during build with @DEB_SOURCE_PACKAGE@.* or something12:59
asacif it makes things cleaner (as we use templating anyway) we could also replace the 1.9* and 3.* with proper template things13:00
jdstrandasac: I'm cool with that conceptually, but will it make it too easy to not verify it? I intentionally did it this way so someone would need to look at it. ie, 3.0's profile didn't work with 3.513:01
jdstrandit mostly worked, but needed tweaking13:01
asacjdstrand: yes. but we usually have those branches right from day zero where 3.5 == 3.6 for instance13:04
asacjdstrand: so its more a continous process anyway13:04
asacso assume we have it for firefxo-3.6 and then 3.7 got created the files would be the same anyway without verification needed13:04
asaconly while upstream changes we need to adjust13:04
asacso i dont think it provides a safety net for us13:04
jdstrandasac: if that works better with your process then I'll update it13:05
asacwould be precious. let me check if i can see something else13:05
asacjdstrand: not sure if i am just confused, but isnt13:12
asac+        if [ ! -e "$APP_CONFFILE" ]; then13:12
asac+            ln -sf $APP_CONFFILE $APP_DISABLE13:12
asacwrong?13:12
asace.g. APP_CONFFILE does not exist => then we link it?13:13
jdstrandit looks wrong but isn't13:13
asacshouldnt it be the other way around?13:13
jdstrandit is based on https://wiki.ubuntu.com/ApparmorProfileMigration13:13
jdstrandthe idea is this happens in preinst, and we know our package supplies the file after unpacking13:13
jdstrandso yes it dangles for a second, but then is resolved at unpack13:14
jdstrandkeep in mind ApparmorProfileMigration is mostly geard for files moving from apparmor-profiles to a package which provides its own profile, but many of the ideas are the same13:15
asacyou think we can use $0 in the pre/post stuff to prevent templating there?13:16
asace.g. for firefox-3.5 etc.13:16
asac?13:16
asachmm postinst.in is alrady a template. so i guess we are fine with making .in out of everything13:18
jdstrandasac: I should be able to do something with preinst13:19
jdstrandasac: I'll work on making it all generalized13:19
asacjdstrand: thanks. i dont like all the maintainer script stuff, but if its really needed then i am fine with the changes once they are generic13:20
jdstrandasac: the idea is to protect the user. if they upgrade from jaunty to karmic, the profile is disabled. if they already have a profile they wrote or they upgrade from karmic to karmic-security we don't want to automatically disable it again13:22
jdstrandthat said, I'm confident I can generalize it13:23
asacjdstrand: yes. i figured that and accepted it ;)13:25
asacall fine13:25
asacjdstrand: you might want to investigate at some point if you can do some debhelper magic to automatically add the right snippets13:26
asacso you dont need to duplicate all the things in all maintainer scripts ;)13:26
asacbut not karmic topic of course13:26
asacdh_installapparmor ;)13:26
asacor something13:26
jdstrandyeah, the problem is they tend to be different enough for each package... but it has been something I've thought about13:27
jdstrandthanks :)13:27
asacno problem ;) /me likes giving easy advices for not-so-easy things :)13:28
gnomefreak!grub213:30
ubottuGRUB2 is the default Ubuntu boot manager in Karmic. For more information on GRUB2 please refer to https://wiki.ubuntu.com/Grub213:30
asacjdstrand: you think i should wait for the karmic ppa security upload for your new merge request?13:46
asaci think release is still 1-2 weeks away so we can wait a bit if you would like to get that into the first .14 upload13:47
jdstrandasac: well, I'd personally like it sooner than later, but I'll leave ff release management up to you :)13:48
asacjdstrand: we only ship profile in firefox package right? not xulrunner?13:48
jdstrandasac: correct13:48
asacjdstrand: when do you think can you do the generalization?13:48
jdstrandasac: I'm building a package with my changes right now13:48
jdstrandso... soon?13:48
jdstrand:)13:48
asacgreat13:50
asaci will wait a bit then13:50
jdstrandasac: do you have any objections to me adding something to the apport hook?13:50
jdstrandasac: I did it with evince and it has worked out fantastically13:51
jdstrand(is that a word?)13:51
asacjdstrand: i am happy to retrieve apport hook love13:53
asac;)13:53
asacof course depends on what you want to append ;)13:53
jdstrandcool, I get that going then too13:53
asacno root passwords for instance :)13:53
jdstrandheh13:53
asacin general who last touched the hook owns it until someone else touches it ;)13:54
jdstrandeek13:54
jdstrandthat is a big committment!13:54
asacnot so big13:55
asacjust a joke ;)13:55
jdstrand:)13:55
asacfta2: i think there are localized search plugins in the locale hg tree ... which are not in the .xpi translation packages we ship14:18
=== asac_ is now known as asac
jdstrandasac: ok, I committed r459 for apparmor/firefox16:09
jdstrandasac: it should be generalized completely16:10
jdstrandasac: it also will add apparmor stuff to the apport report if the profile is not disabled16:10
jdstrandasac: https://code.launchpad.net/~jdstrand/firefox/firefox-3.5-apparmor/+merge/1106116:11
asaclet me check16:12
jdstrandasac: I think I did all the LP right this time too, so the 'Review Diff' should now be accurate16:12
asacyeah saw that the other is properly marked as superseded ... well done ;)16:12
jdstrandheh16:12
asacjdstrand: have you tested the apport hook?16:12
asacthis also smeels like something that could be factored into apport general package i think16:13
jdstrandasac: yes, works great (both when disabled and enabled)16:13
asacwe have similar generic functions for wifi etc in there now16:13
asacjdstrand: thx for confirming16:13
jdstrandasac: yeah, I thought of that too, but too late for FF16:13
jdstrand(on my todo)16:13
asacimo thats not FF relevant ;)16:13
asacbut thats just me ;)16:14
jdstrandI'll talk to pitti when I have more time, and if I can get to it, I will and will update this packaging16:14
asacthe feature is there ... just not in the central apport hook library ;)16:14
* jdstrand nods16:14
asacjdstrand: no problem16:15
asac[ Jamie Strandboge <jamie@ubuntu.com ]  -> wrong (i will fix it during merge)16:16
asac;)16:16
jdstrandasac: fyi, I compared all the old maintainer scripts with the newly generated ones and they're the same16:17
jdstrandasac: re changelog> oops! :)16:17
jdstrandthanks16:17
jdstrandasac: fyi, I fixed the typo in debian/changelog in r46016:19
asacjdstrand: there is also DEBIAN_NAME_OTHER ... for abrowser ... not sure if we can make that work easily ... without copying the full block16:20
jdstrandasac: oh, I forgot-- you probably should wait to upload until the next kernel is uploaded16:20
asacjdstrand: what will happen with old kernels?16:20
jdstrandasac: as it is disabled by default, nothing, but if someone enables it, it won't load cause I use PUx for evince16:21
asacjdstrand: there is no new dependency added16:22
asacis that because its installed by default?16:22
asacsince when are the binaries used available?16:22
jdstrandasac: not a huge issue-- the patches are committed to the kernel and just awaiting upload16:22
asaclike aa-... ?16:22
jdstrandasac: this will work fine without apparmor installed, so no Depends or Recommends. sometimes I add a Suggests, but didn't here16:23
jdstrand(tried to keep the packaging changes to a minimum)16:23
asacjdstrand: but it calls apparmor_parser and aa-status without checking that the binaries actually exist16:24
asachmm ... but only if APP_PROFILE exists for postinst16:24
asacjdstrand: point is that we build the .head branch everywhere: hardy - karmic16:24
asacso if possible it would not break there ;)16:25
asacok seems ok16:25
jdstrandasac: well, 'aa-status' will fail if it isn't around16:25
asacdo you see anything that might break?16:25
asacyeah true16:25
jdstrandand I 2>/dev/null it16:25
micahgasac: I'm reminding you about the fixed-3.5.3 and fixed-3.0.14 tags16:26
jdstrandasac: this is the same recipe I use all over the place16:26
micahgI saw the branches committed for the next ff release16:26
asacmicahg: huh ;)16:26
jdstrandasac: I have tested it here and am comfortable with it16:26
asacthanks16:26
asacmicahg: that was last minute ;916:26
jdstrandasac: I'll of course fix anything that breaks16:26
asacmicahg: though i already uploaded xulrunner ;)16:26
asacbut thats ok i guess16:26
asacnext time i should remember it better16:26
asacjdstrand: ok. i think its ok. lets merge it and see if the daily mob shouts ;)16:27
jdstrandheh16:27
asacjdstrand: but we need to fix abrowser later16:28
asacdoes the name need to match the binary? or just this: /usr/lib/@APPNAME@.*/firefox { ?16:28
asacoh its referring to firefox16:28
jdstrand/usr/lib/@APPNAME@.*/firefox16:28
asacso i guess only question is if usr.bin.firefox.apparmor16:28
asachas any meaing16:28
asacbut i guess not16:28
asacok great16:29
asaclets go for it16:29
jdstrandwhatever uses that binary is confined16:29
asacjdstrand: ok merged ... i will check if its easy to just pull that over to 3.7 and 3.6 branch and if not i will poke you ;M)16:30
jdstrandasac: if it makes sense to rename the profile to simply apparmor.profile, we can do that.16:30
jdstrandasac: cool, I will be more than happy to fix anything wrt this16:30
jdstrandasac: thanks for the review and merge! :)16:30
asacmicahg: bug 236853 ... is that really fixed by ffox? not by nss?16:31
ubottuLaunchpad bug 236853 in firefox "firefox crashed with SIGSEGV in NSSRWLock_LockRead_Util()" [Unknown,Fix released] https://launchpad.net/bugs/23685316:31
* asac checks upstream bug16:31
asacok its fixed in ffox/xul16:32
asacthx16:32
micahgI just followed whatever they said upstream16:33
asacmicahg: documented. thanks a bunch for that ;)16:36
asacmicahg: you were right.16:36
micahgare they actually releasing 3.0.14 or did it go t oRC?16:43
micahgasac: I made my patch better, but now it fails to build for bug 6601516:48
micahghttp://launchpadlibrarian.net/31218684/buildlog_ubuntu-jaunty-amd64.xulrunner-1.9.1_1.9.1.2%2Bnobinonly-0ubuntu0.9.04.1~ppa10_FAILEDTOBUILD.txt.gz16:48
ubottuLaunchpad bug 66015 in hundredpapercuts "Duplicate spell checking dictionaries for every entry" [Low,Confirmed] https://launchpad.net/bugs/6601516:48
asacmicahg: beta16:48
micahgok16:48
asacusually it goes 1-2 weeks before release to beta channels16:48
asacthats when i try to upload stuff to PPA16:48
asacif all goes well they dont respin and the same binary will be the release16:48
asacjdstrand: remember to mark your branch as merged (if launchpad didnt do that yet)16:49
jdstrandasac: done16:51
micahgasac, here's the new patch http://pastebin.com/f3982e87716:52
micahgApparently, I don't have PRUniChar loaded16:53
asacmicahg: but thats still the same approach16:55
asachow about the idea replacing _ with - (or vice versa)16:55
asacand then checking whether a dictionary with that name is alread in the mDictionaries map?16:55
asacbtw its PRUnichar16:55
micahgah16:55
micahgwhy not stop it at the source though16:56
micahgprevent the file from loading16:56
asacthe file isnt loaded if you dont put it in the dictionarie16:56
micahgmy original patch actually worked16:56
micahgok16:57
micahgwhy not use a simpler patch though?16:57
asacbecause its fragile16:57
asacnobody knows what is wanted16:57
micahgwhy?16:57
asac_ or -16:57
micahgah16:57
asacso better safe than sorry16:57
asacand allow both16:57
micahgI can add a commet16:57
micahgoh16:57
micahgthat's what you mean16:58
micahgI looked at the upstream binaries16:58
micahgand they ship - in the files16:58
micahgit seems to be a mozilla thing16:58
asacjust allow both variants, but just normalize the key16:58
asac(e.g. dict)16:58
asacthe other alternative is to resolve an eventual link first16:58
micahgwell16:58
asaclike if file == link -> resolve ... then use leafName, normalize like discussed (e.g. replace _ by -)16:59
micahgwhat I'm wondering is why it doesn't display the whole name for the _ dicts but does for -16:59
asacmicahg: that logic was at the other place16:59
micahgah16:59
asacwhat bug id was it again?16:59
micahgbug 6601516:59
ubottuLaunchpad bug 66015 in hundredpapercuts "Duplicate spell checking dictionaries for every entry" [Low,Confirmed] https://launchpad.net/bugs/6601516:59
asacthe other patch touched the source that assembles the name16:59
micahgwould that type of patch be worthy of upstreaming then?17:00
asaccontent/global/inlineSpellCheckUI.js17:00
asacmicahg: http://launchpadlibrarian.net/31020349/inlineSpellCheckUI.patch shows that it only shows the full name for "-" most likely17:00
asacat least thats what is processed properly17:00
asac(without that patch)17:00
micahgok17:01
asacmicahg: what are the links for us?17:01
micahgso I'll work on it tongiht17:01
asacthe _ or the -?17:01
micahg-17:01
micahg_ is the normal file17:01
asac- are links to the _ files?17:01
asachmm17:01
micahgapparently the ISO standard is _17:01
micahgbut mozilla likes -17:01
asacok so what we want is to use the dict key normalization to use "-" (because thats what mozilla uses)17:01
micahgok17:01
micahgI have to leave for work soon17:02
asacand then we want to first split with "-"17:02
micahgso I'll work on this tonight17:02
asacso i think in best case we dont need to touch the .js17:02
asacyeah17:02
asaci think you dont need to resolve the link17:02
asacthough upstream might want us to do that17:02
asacbut lets first try to just normalize to use "-" as the key17:03
asacand check that for (var i = 0; i < list.length; i ++) {17:03
micahgok17:03
asaclist here is that key17:03
asacbut i think its right then17:03
asacas long as we normalize to use -17:03
micahgok17:03
micahgBTW, is ABrowser the official branded name of the unbranded browser?17:03
micahgwe have bug 41307617:04
asacits our unbranded browser name17:04
ubottuLaunchpad bug 413076 in firefox-3.5 "abrowser: Change menu label to just "Web Browser"" [Undecided,New] https://launchpad.net/bugs/41307617:04
asacthere is no official unbranded branding ;)17:04
micahgso, can we change the menu to ABrowser vs A Web Browser so it doesn't seem so weird on other languages?17:04
asacabrowser was ment to be short for "A Web Browser"17:06
asacit was considered rougue to just use "Web Browser"17:06
asacbecause other browsers would then want that name too17:06
micahgyes, but if it's like that, it would seem that it should be localized in other langauges17:06
asaci will think about it17:06
asacnot if "A Web Browser" is considered a brand name17:07
asacbut well17:07
micahgoh17:07
micahgoops17:07
asac;)17:07
micahggot it mixed up17:07
micahgit is translating it17:07
asacyeah17:07
micahgbut only patially17:07
asacso technically its right17:07
micahg*partially17:07
asacquestion is if the bug has a point about confusion. i dont think so for now17:07
asacbut have to think more17:07
micahgok17:08
micahgok to mark as triaged then?17:08
asacyes assign it to me and triaged stating this is pending decision17:08
asacyou can also project that we probably won't go for just "WebBrowser"17:09
asacRather "A Browser" ;)17:09
asacgiven the rational against taking Web Browser from above17:09
micahgI was thinking without the space and have it be an actually name - ABrowser17:09
asacYeah17:09
asacbut firefox == "Firefox Web Browser" .... abrowser = "A Web Browser" or "ABrowser Web Browser"17:10
asacwhich would be odd too ;)17:10
* micahg likes the last one actually17:10
micahgABrowser web browser17:10
micahgthen the description can be localized17:10
micahgbut the name remains17:10
asacwe intentionally picked the most generic name so we wont need to enforce any trademarks on it as its probably not trademarkable on its own17:10
asacyeah17:11
asacbut ABrowser Web Browser is double ;)17:11
micahgdo you think it's worth throwing up on brainstorm?17:12
asacdefinitly not17:12
asacthtough branding discussion in the wild and you get a fully raged rant17:12
asaceverybody in the end being unhappy17:12
micahgah, right17:12
* micahg remembers the Shiretoko rants17:12
asacmicahg: fyi: https://developer.mozilla.org/en/XPCOM/Strings17:34
asacMozilla internal string guide17:34
asachttps://developer.mozilla.org/en/String_Quick_Reference17:35
* mac_v seems nothing rogue about using "Web Browser"17:47
mac_v*sees17:47
asacmac_v: we would claim a generic name that other browsers would want to use if it was open for discussion18:12
asacso everyone agrees to not use such a generic name18:12
mac_vhmm... too bad , now we have to confuse the user ;p18:39
micahgmac_v: ??18:49
mac_vmicahg: the "AWeb Browser" , i was replying to asac ;)18:50
micahgsorry, I must've caught the last line of that18:50
ftaasac, http://launchpadlibrarian.net/31252914/buildlog_ubuntu-intrepid-i386.firefox-3.7_3.7~a1~hg20090902r32157%2Bnobinonly-0ubuntu1~umd1~intrepid_FAILEDTOBUILD.txt.gz19:49
andvasac, all-in-one accepted19:54
andvasac, I pinged a friend, who is ftp-assistant19:55
andv;)19:55
sveinungandv and asac: thanks a lot for your help and sponsorship getting all-in-one-sidebar into Debian21:55
andv;)21:55
andvsveinung, np21:56
sveinung:)21:56
andvsveinung, let me know when there will be a new upstream release21:56
sveinungandv: sure21:57
andvsveinung, I gonna sponsor the package from now on21:57
andvsveinung, do you know how Debian BTS work?22:07
sveinungandv: yes, at least part of it22:08
andvok, both me and you will receive bug mails22:08
sveinungok22:08
andvso if you don't know what to do, ping me22:08
sveinungsure22:08
=== ripps_ is now known as ripps

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