[00:41] asac: for the bug 243064 news2000 has made a redirect on the site, so the bug is now override, but seems to be confirmed in Firefox 1.5 on Dapper (I think I don't have the Release Notes link in the Help menu on FF1.5 on Dapper... I don't have see it yesterday when trying) [00:41] Launchpad bug 243064 in ubuntu-website "page "releasenotes 604" not found" [Undecided,Fix released] https://launchpad.net/bugs/243064 [00:41] I don't know if you want to add the MozillaTeam and/or the Ubuntu Firefox package to the bug [06:32] bug 246694 [06:32] Launchpad bug 246694 in firefox-3.0 "Can Not Uninstall Add-Ons" [Undecided,Incomplete] https://launchpad.net/bugs/246694 [08:06] asac: you around? it seems i have to backport nss and nspr [08:06] here is why * bump minimum libnss3-dev build requirements to >= 3.12.0~1.9b3 [08:07] - update debian/control [08:07] * bump minimum libnspr4-dev build requirements to >= 4.7.0~1.9b3 [08:07] or i lower the minimum version of nss nspr but i think its easier to backport them [08:13] hi, [08:13] bye [08:13] :) === rzr is now known as rZr [08:18] gnomefreak: just drop the versions from the depends [08:19] i am [08:22] ok xulrunner started [08:48] * gnomefreak remembers nss builds being faster than this [09:03] if i dont fall asleep first i should be done with these by lunchtime like 7-8 hours [09:04] * gnomefreak going to lay down for an hour or so waiting for xulrunner [09:20] Hi ! [09:22] hi XioNoX [09:22] hi asac [09:23] Ready for another day of ubufox :D [09:23] hehe [09:23] What can I do today ? [09:23] XioNoX: ok. so today the idea is to cleanup your branches ;) [09:23] o_O [09:23] XioNoX: then get the diff reviewed by gandalf [09:23] It is already dirty ? [09:24] okay [09:24] XioNoX: look at your branches. the code style is quite dirty [09:24] :) [09:24] he is not yet here at works :D [09:24] i commented on your plugin-finder branch in the branch whiteboard [09:24] on launchpad [09:24] XioNoX: take a look at those comments there [09:24] oh [09:25] XioNoX: so basically: use same indentiation, same variable name formatting and _dont_ remove code by commenting it out (instead remove it completely) [09:26] so comments are only real comments [09:26] what do you call indentiation ? [09:26] XioNoX: indentation [09:26] to indent ;) [09:26] I'll gooogle translate this [09:27] I've try to respect the variable name formating [09:27] XioNoX: anyway. so half day code cleanup, then we will send you to new tutorials while drafting the next big task you can do ;) [09:27] XioNoX: try better ;) [09:27] XioNoX: its good to use bzr diff to see your changes [09:28] do that and you probably see what i mean [09:28] ok [09:28] XioNoX: maybe work in the plugin-finder comments i gave in whiteboard [09:28] i will give you comments on the other branch as soon as i get to it ;) [09:28] ok [09:28] today I'll left earlier [09:29] beacause I'm going back home (in britany) for a week [09:29] so you will maybe see me a little bit during the next week, but mostly the week after [09:31] XioNoX: ok. so we should certainly induce you what you should look at during that week ;) [09:31] the next tasks we were thinking of is more complex ;) [09:32] ok, i'll see if i can do it [09:32] "Remove the Uncomplete header", what do you mean in your whiteboard ? [09:33] and what I've call littledesc, will maybe be used for the homepage url [09:36] XioNoX: "uncomplete" comment [09:36] which is not an english word ;) [09:36] ok [09:36] littledesc == i am fine if you want to use that for homepage, name it "homepage" ;) [09:37] just remove the comment ? [09:37] XioNoX: yes. it doesnt have any benefit, so just drop it [09:37] XioNoX: then look at the diff: [09:37] bzr diff -r 94.. [09:37] it is to warn people that it point to nowhere [09:37] and see whatelse you want to improve [09:38] ok [09:38] XioNoX: ok. i don think that we need that information there [09:38] if we want to add that we should document that in a TODO file [09:38] e.g. TODO: implement homepage + description in plugin finder webservice ... or something [09:39] * asac out grabbing some food [09:56] [reed]: who is jemalloc reviewer? [09:56] fast! [09:56] <[reed]> stuart or jasone [09:56] [reed]: mozilla bug 435683 [09:56] Mozilla bug 435683 in jemalloc "libjemalloc deadlocks with aoss" [Major,Assigned] http://bugzilla.mozilla.org/show_bug.cgi?id=435683 [09:57] we (ubuntu) havent hit that bug, just saw that comment on debian bug tracker [09:58] <[reed]> hmm [09:58] armin76: so my static jemalloc patch works for you on gentoo, right? [09:58] armin76: do you have same libc as we do? [09:58] asac, code clean & pushed fot the 2 branches [09:59] * asac pulls [10:00] XioNoX: bzr diff -r94.. [10:00] (on plugin-finder branch) ... there is still some wierd indentation [10:00] for table.appendChild(cellDescUnder); [10:00] normally not [10:01] XioNoX: there is ;) ... further, please make two lines out of: "if (!pluginInfo.desc) pluginInfo.desc = "No description found in plugin database. Report a bug to ubuntu-mozillateam@lists.ubuntu.com";" [10:02] XioNoX: ok. and a not needed new-line addition below "// create the radio" [10:02] thats all from style from me [10:02] on plugin-finder [10:03] but that is peanuts, not ? [10:04] teacher always told me to comment as much as possible, put a lot of space, etc... :D [10:04] XioNoX: hehe. yes, but that doesnt mean to add a new line where there wasnt before ;) [10:04] patches should be minimal :) [10:05] with a lot of comment I would understand the code a lot faster, you don't thinks so ? [10:07] plugin.finder pushed [10:10] XioNoX: my comment was about removing an empty line [10:10] not about removing a comment [10:10] look at the bzr diff -r94.. [10:11] there is a new line added below // create the audio [10:11] XioNoX: ok. i commented on the other branch [10:11] in whiteboard [10:11] https://code.edge.launchpad.net/~xionox/ubufox/restart.notifier [10:12] i've remove the empty line under // create the radio [10:12] ok cool [10:13] (sorry, i dont want to be picky here ...) [10:13] but i am sure you would get those comments from any other mozilla dev when sending these patches for review to bugzilla ;) [10:13] so better get used to it :-D [10:14] ok [10:15] XioNoX: ok. once you have done those code clean-ups there is one more thing left to test :) [10:15] ubufox is ment to work with firefox-2 as well [10:16] so install firefox-2 package (backup your .mozilla directory in case it messes with it) :) [10:16] and see if it works there as well [10:21] 2 things [10:21] first, i dont get this : [10:21] + a=Components.interfaces.nsIAppStartup,Components.classes["@mozilla.org/toolkit/app-startup;1"].getService(a).quit(a.eRestart | a.eAttemptQuit); [10:21] get the service outside the function scope and remember it as a global variable. this will make thise line basically just read: gAppStartup.quit(gAppStartup.eRestart | gAppStartup.eAttemptQuit) [10:22] and about firefox 2, the script check the file : /var/lib/update-notifier/user.d/firefox-3.0-restart-required, which is about firefox 3 [10:22] XioNoX: make a global variable var gAppStartup = .... [10:23] just change the "a" by "gAppStartup" ? [10:24] i dont get it [10:24] no ... get the service as a global variable [10:24] and use that instead of a [10:24] global variable == what you did with var gRFPrefsService = C [10:25] XioNoX: ok. for ffox 2 i will think about what to best do. leave it alone for now [10:25] XioNoX: another indentiation thing: please use 2 spaces instead of one in content/updateRestart.js (and all the other .js files you did) [10:26] and introduce line breaks for: var ProDir = Components.classes["@mozilla.org/file/directory_service;1"].getService(Components.interfaces.nsIProperties).get("ProfD", Components.interfaces.nsIFile); [10:26] too [10:26] thats a ridiculous long line ;) [10:26] thats all from me i guess [10:26] haaaaaaaaaaaaaa [10:27] first i put : [10:27] var gAppStartup=Components.interfaces.nsIAppStartup,Components.classes["@mozilla.org/toolkit/app-startup;1"].getService(a).quit(a.eRestart | a.eAttemptQuit); [10:27] outside function, [10:27] and then i put gAppStartup in the restart() function ? [10:28] do you understand what the line is doing? [10:28] we want the service to be stored in global variable gAppStartup [10:28] and the quit in restart [10:29] asac: i have glibc-2.6.1 [10:29] armin76: can you try 2.7? [10:29] err, 2.8 ;) [10:29] will take a while but yes [10:30] asac, no i don't really understand wht the line is doing [10:30] armin76: my guess is that libc 2.7++ doesnt work with jemalloc [10:30] i've just find it by exploring xlr, ant it works :D [10:31] asac: that would be pretty stupid, wouldn't it? :D [10:31] armin76: i find the complete jemalloc thing stupid ;) [10:32] XioNoX: try to understand it .... it basically gets you a service from xpcom [10:32] and then it runs a function (quit) on it [10:34] asac: disable it then :P [10:34] armin76: i was about to enable it ... because people bitch around upstream [10:34] but now it crashes in 2.7/2.8 [10:35] ok, done [10:37] thx [10:38] hmm ... not yet pushed ;) [10:50] i have errors [10:50] no errors [10:50] but it don't work now [10:54] something around the lastmodified date check [10:54] resReqFile.exist return undefined [10:55] \o/ [10:55] i could update my key in debian :)( [10:55] XioNoX: check that you didnt forget to rename one or the other variablE [10:55] yay [10:56] (my ssh key) [10:56] nop, all variables correct [10:56] now i can log into people.debian.org again ;) [10:56] XioNoX: you will find i am sure ;) [10:58] asac: took a while to bumb :P [11:07] armin76: i think i understand why you are so focussed on bumping ;) ... most likely your machnie bumbs all the time [11:09] lol [11:14] asac: hi === jt1 is now known as jtv [11:14] (it's me, jtv! :) ) [11:15] asac, ok, pushed [11:17] jtv: i know that you are jtX ;) [11:18] hi [11:21] asac: Wow, I'm a wildcard [11:22] asac: I am also the bearer of good news: expect the workaround for those "//" to be in production very soon. [11:22] asac: the spec says they're end-of-line comments, but we've changed the parser a bit to treat it as regular text if it follows a colon. [11:23] jtv: ok. sounds like a feasible workaround then [11:24] what spec are you referring to? [11:24] asac: just a moment, I'll dig it up. [11:24] asac, what do you think about the last push ? [11:25] asac: http://www.mozilla.org/projects/l10n/mlp_chrome.html#text [11:25] XioNoX: which revision? [11:25] XioNoX: i didnt receive any update yet [11:25] n$ bzr push bzr+ssh://bazaar.launchpad.net/~xionox/ubufox/restart.notifier [11:25] Pushed up to revision 102. [11:26] https://code.launchpad.net/~xionox/ubufox/restart.notifier [11:26] hmm ... still at 97 here [11:26] lets wait a bit more [11:26] 97 is for the plugin finder [11:26] 102 is for the restart notifier [11:26] ah ;) [11:26] yeah [11:26] let me check [11:27] XioNoX: move the getService to the global location too [11:27] e.g. Components.classes["@mozilla.org/toolkit/app-startup;1"].getService(appStartup). [11:27] ? [11:27] how ? [11:28] err. i think hte problem is that its messed up [11:28] use var gAppStartup = Components.classes["@mozilla.org/toolkit/app-startup;1"].getService(Components.interfaces.nsIAppStartup); [11:28] and then just gAppStartup.quit... [11:28] thats the idea [11:29] ok [11:29] and remove the empty line in the restart function [11:29] that should be it [11:29] (from my side) [11:30] if it works run it through gandalf ;) [11:30] oh ... and since you renamed Notif to Notification (good) everywhere, rename the function too: [11:30] showRestartNotif [11:30] :)= [11:40] ok [11:40] but about firefox 2 [11:40] nothing work :D [11:40] it is like ubufox isn't loaded [11:41] even for the plugin funder [11:41] finder [11:52] XioNoX: you have to wipe your profile in order to get extensions reregistered imo [11:53] thats why i said: backup your profile ,) [11:53] XioNoX: try stop firefox, then remove extensions.* files in your profile [11:53] ok [11:53] and start again [11:53] maybe that helps ffox to re-register extensions on the lower versionb [11:56] http://img79.imageshack.us/my.php?image=captureservicederechercqv6.png [11:56] it is a little bit more ugly [11:59] XioNoX: what is "undefined" for? [11:59] the homepage link? [12:00] normally it show just a little "-" [12:00] XioNoX: how comes? [12:00] ? [12:02] XioNoX: why undefined instead of "-"? [12:02] can we fix that in ffox 2? [12:02] I don't know [12:02] i don't know why i see undefined [12:04] XioNoX: most likely you set a value to "null" [12:04] test for null and either hide the element or explicitly set "-" or whatever suites well [12:06] if (!pluginInfo.homepage){ [12:06] pluginInfo.homepage = "-"; [12:06] } [12:06] it is the same as : [12:06] if (!pluginInfo.desc){ [12:06] pluginInfo.desc = "No description found in plugin database. Report a bug to ubuntu-mozillateam@lists.ubuntu.com"; [12:06] } [12:06] but the 2nd works [12:07] and now firefox don't launch anymore :D [12:07] I restart [12:07] 2min [12:15] back [12:15] the update notifier works well [12:16] but there are this litte bug in the plugin finder [12:18] * gnomefreak likes when things work smoothly [12:23] XioNoX: i think the "undefined" comes from the datasource [12:24] I hope too, not my part :P [12:25] XioNoX: well you can check whether the "homepage" resource exists [12:25] but the desc field shod have the same thing normally [12:25] *should* [12:26] XioNoX: hmm maybe homepage is already defined in webservice. try to use homepage2 ;) [12:26] (just to test) [12:26] if that helps we have to think about a different name [12:27] I don"t want tu run ff2 again [12:27] it mess up my computer [12:27] :D [12:27] hehe [12:30] until when ff 2 is supported ? [12:30] XioNoX: depends on where? [12:31] upstream will abandon it in jun + 6 month [12:32] ok [12:35] so, what now ? [12:38] asac_: is xulrunner-1.9 source stripped as much as possible. its still huge and i thought at one time we were going to try to strip it down a bit to make it smaller [12:38] tarball is around 39386k [12:39] XioNoX: is gandalf in the office yet? [12:40] yep [12:40] but we are just going to eay [12:40] eat [12:40] gnomefreak: there is not much we can strip from xul ... we removed bits from ffox source [12:41] ah [12:41] XioNoX: ok. i think next step is to do some thorough XPCOM and RDF Template tutorials (throughout your free week) [12:46] asac_: any idea why im getting this warning/error i left the versions i have installed below the build failure http://pastebin.mozilla.org/484305 [12:47] gnomefreak: you messed up syntax [12:47] a missing , [12:47] or a missing tab in front of libnspr4-dev [12:47] look at the dependency lines (as the rror suggests) [12:47] i didnt change syntax in control but i will look [12:47] thansk [12:48] gnomefreak: hmm [12:48] it says that it cant parse that [12:48] gnomefreak: you dropped the version from libnspr4-dev [12:48] you probably did something wrong while doing that === asac_ is now known as asac [12:49] but that package built and installed [12:49] even xulrunner built and installed [12:49] then whats the problem? [12:50] well i dont see anything in firefox debian/control [12:51] i found it i think there was a missing , after xulrunner dep and binutils so i added one there lets see if it works now [12:52] how did yours build while missing the ,? [12:52] gnomefreak: i am sure my dont miss , [12:52] gnomefreak: thats your work ;) [12:52] since i didnt touch control before that error [12:52] asac: you were last one in control [12:52] debian/control and if i didnt touch it how did it change [12:53] gnomefreak: ok. while this is building, lets setup proper gutsy-backport branches in mozillateam [12:53] (based on hardy) [12:53] all i did was run dch -i [12:53] gnomefreak: so we go directly for 3.0.1 [12:53] wait a few minutes [12:53] i am preparing that for you now [12:53] for gutsy? [12:53] asac: i sure hope i dont need to redo nss nspr xulrunner [12:55] gnomefreak: you dont need to do nss/nspr at all :) [12:56] but i told you yesterday ;) [12:58] asac: yes i did [12:58] xulrunner needed them [12:58] or PPA would have failed to build it [12:59] asac: and i cant do 3.0.1 until its in Intrepid [12:59] or it wont be accepted [13:00] i gave you the section in xul about nss nspr [13:01] gnomefreak: ok. work on top of bzr branch lp:~mozillateam/xulrunner/xulrunner-1.9.gutsy-backports [13:01] (should be there in a few minutes) [13:01] i prepared the changelog for you [13:02] so you just have to do the changes [13:04] gnomefreak: and: lp:~mozillateam/firefox/firefox-3.0.gutsy-backports [13:09] when is it going to land in Intrepid? [13:15] gnomefreak: soon enough to do the backport now ;) [13:15] gnomefreak: right after alpha-2 freeze is lifted [13:16] gnomefreak: hmm. this makes me think if we should really base the branch on hardy [13:18] gnomefreak: hold back a few [13:19] gnomefreak: lets base this on intrepid ;) [13:19] gnomefreak: let me redo the branches for you [13:19] sorry [13:19] good because its alot simplier and supposed to be done with intrepids build (also makes life easy) [13:20] oh and devscripts doesnt have a target for 3.0.1 it has it for 3.0 and 3.1 [13:21] i keep hearing a2 has been released so the freeze should have been lifted if what i hear is correect [13:22] maybe ./debian/rules get-orig source would work instead of devscripts [13:23] asac: my backport needs to be exactly the same as in Intrepid with a ~gutsy0 added to it [13:26] gnomefreak: get the orig.tar.gz from my ppa [13:29] gnomefreak: ok lp:~mozillateam/xulrunner/xulrunner-1.9.gutsy-backports and lp:~mozillateam/firefox/firefox-3.0.gutsy-backports should be ready to go (based on intrepid release branch (e.g. .dev) [13:34] asac: is firefox-3.0_3.0.1+build1+nobinonly.orig.tar.gz the final name for intrepid? [13:35] including the +build1+ [13:40] hello [13:40] asac, I'm back [13:40] gandalf already pointed me to RDF Template tutorials [13:43] asac: no need to add anything from prevous changelogs all the change log needs is backported info. and the naming you used is wrong ill get yelled at for that [13:46] gnomefreak: what? [13:46] XioNoX: cool [13:47] XioNoX: you should certainly try to implement a javascript XPCOM component and play around using that from the xpcshell [13:47] XioNoX: as we found earlier today that your xpcom knowledge could deserve some moderate development in order to become effective ;) [13:48] ok [13:48] when you made changelog for gutsy you used the lp bug #xxxx but that was in intrepids changelog and the version that backports use is same exact as intrepid only adding ~gutsy0 for me so official backport will use ~gutsy1 not 7.10 [13:48] zbraniecki: ^^ ok? [13:49] gnomefreak: hmm [13:49] feel free to bump the package version if that helps you :) [13:49] i changed it [13:49] gnomefreak: oh no [13:49] gnomefreak: thats bull-shit [13:49] oh yes [13:49] we have to keep the version i choosed [13:49] cant [13:49] gnomefreak: you can [13:49] gnomefreak: just keep it [13:50] no i cant it wont get backported [13:50] otherwise users wont upgrade to hardy [13:50] gnomefreak: you do the backport [13:50] no i dont [13:50] gnomefreak: and we find an upload [13:50] i package it archive admins push it with a script [13:50] i get nothing for doing it [13:50] i do nothing after testing if it builds and runs [13:50] thats all i do [13:51] you need to fight it out with backporters to use your version i cant make that call [13:51] gnomefreak: well. thats "doing the backport" [13:51] as you will fix issues you encounter while doing so [13:51] gnomefreak: just keep the version i proposed for now [13:51] gnomefreak: ill talk to the right people before we upload to figure out [13:52] we can bump it easily later in case we need to [13:52] ok? [13:52] if i use 7.10 than gutsy is not going to upgrade to ~gutsy1 my PPA will be higher [13:52] gnomefreak: you can remove the package from your PPA [13:52] if you already uploaded it [13:52] asac: i dont care locally its going from my PPA to gutsy that is going to have problems upgrading [13:53] they still have it on thier system [13:53] people testing it [13:53] cant upgrade to your version [13:53] gnomefreak: thats collateral damage we can accept imo [13:53] gnomefreak: people that are subscribed to backport bug will get that info [13:54] asac: I think that RDF template is more urgent [13:54] asac: while I'm not convinced why XPCOM is that needed here [13:54] (maybe because my XPCOM knowledge is limited too ;)) [13:54] zbraniecki: well, he doesnt know a thing about how to use XPCOM components [13:55] asac: how to USE, ok. then MDC tutorial on using components is needed [13:55] but do you need to write any XPCOM components? [13:55] zbraniecki: yes. we need to make search and install pluggable [13:56] if not search, at least installing plugins should allow you to add new install methods [13:56] e.g. .xpi + .deb + .rpm [13:56] atm there is no bug report until im done this time and i dont need a bug if you are pushing instead of archive admins and im sure not going through hell with this its not worth it and once backporters (have to ack it) sees versioning i get hell (dont you remember the other day with flash versioning? same issue as there IMHO firefox isnt worth that much shit for anyone, maybe ill look at it again later but please talk to scottk ab [13:56] that's ok [13:57] dont know where that ended but talk to scottk about the versioning [13:57] asac: but you'll need to help XioNoX more than I can with writing XPCOM :) [13:57] and I want to be there [13:57] to learn too :) [13:57] zbraniecki: ill write the points down as soon as i have time. but i got to know that XioNoX is away for a week [13:57] so that probably gives us enough time to flash out the details for what should be done [13:57] and with what priority [13:58] asac: yes, we can use this time to plan better what's needed :) [13:58] good. thats helpfulö [13:59] zbraniecki: can you point XioNoX to some basic XPCOM tutorials? i havent looked for quite some time in MDC, not sure where they are :-D [13:59] yea [13:59] zbraniecki: i think he can learn this and templating during next week then :-D [13:59] do you want him to read XPCOM before templates? [13:59] I can, but I'll not have too much time [13:59] zbraniecki: i dont mind which order ;) both should be done the week after next week ;) [14:00] yes [14:00] I think templates are easier [14:00] :) [14:00] zbraniecki: yeah, but they can also be painful :-D [14:00] and I could refresh my XPCOM knopwledge a bit ;) [14:00] right [14:00] XioNoX: what do you prefer? [14:00] there are so many words here : http://developer.mozilla.org/en/docs/XUL:Template_Guide:Introduction [14:00] learning painful but easier stuff, or more C++ like but more logical stuff? :) [14:00] XioNoX: i think you just need a few hours and to try the basic tutorials... it would just be really helpful if you knew the basics ;) [14:01] give me the 2 tutos [14:01] And I will chose [14:01] XioNoX: you need both :-D [14:02] so give me both [14:02] asac: are you ok with those XPCOM components being written in JS? [14:02] http://developer.mozilla.org/en/docs/XPCOM this ? [14:02] XioNoX: yeah ;) zbraniecki will give you a few more links about basic use of XPCOM in javascript ;) [14:02] zbraniecki: yes. thats my idea [14:02] why did we go back to ~/.mozilla-thunderbird instead of ~/.thunderbird [14:03] XioNoX: http://developer.mozilla.org/en/docs/How_to_Build_an_XPCOM_Component_in_Javascript [14:03] gnomefreak: we never moved to .thunderbird [14:03] i had it [14:03] maybe tbird 3 [14:03] zbraniecki: does that also discuss how to add a new interface? [14:03] yes, tb3 is in .thunderbird [14:03] zbraniecki, ha, english only [14:04] it would be too easy with a french one [14:04] XioNoX: chinese and japanese are ready to use too [14:04] I've seen [14:04] XioNoX: do not hesitate to localize it :) [14:04] zbraniecki: mozilla.com always used .thunderbird ... its just that we and debian still are stuck in the custom profile folder :/ [14:04] asac: I know. [14:04] zbraniecki, it is not my job :) [14:04] and I'm tolerant so I never asked this tremendous question - WHY? [14:04] :) [14:04] asac: seee #ubuntu-bugs about ~/.firefox instead of ~/.mozilla/firefox [14:05] XioNoX: I know :) You CAN do this, but you can read english one, I don't mind [14:05] ;) [14:05] asac : shouldn't you use freedesktop recomandations ? : http://ploum.frimouvy.org/?184-cleaning-user-preferences-keeping-user-data [14:05] btw firefox meets SRU why are we backporting it? [14:05] zbraniecki: when i did this in tbird 0.1 i was used to mozilla trashing profiles if you downloaded a new version and tested that [14:05] http://standards.freedesktop.org/basedir-spec/latest/ar01s03.html [14:05] i didnt want to deal with users complaining about broken profiles when they installed tbird from mozilla.com [14:06] now i am just scared that i will break existing profiles if i auto migrate them (as there are still absolute paths in really, really old profiles) [14:07] but last time i seriously investigated this was in the beginning of 1.5 cycle ;) [14:07] tbird 3 should be back to normal for sure ;) [14:08] someone needs to take that over i was thinking about doing it but havent decided yet [14:08] same with ff4 [14:08] gnomefreak: no ffox will stay in that folder for sure [14:09] fta2: didnt want to update them. come to think of it i havent had any updates from his PPA [14:09] asac: its not that its why isnt it was the question [14:09] gnomefreak: he is not here atm. ithink he does a combined business + holiday trip this july [14:09] he asked why we dont use ~/.firefox i figure we dump everything else in there why not leave it [14:10] asac: figured that [14:12] asac: what is the datasurce for plugins? [14:12] is it RDF? [14:13] what is the source of data for those templates? [14:14] asac: ping [14:17] zbraniecki: the datasource is not a real rdf datasource [14:17] zbraniecki: it uses RDF and parses that to a structure [14:17] (e.g. pluginInfo) ... which is then used by the Wizard.js to construct the xul tree [14:17] ok. because next to us we have an owner of RDF module in Mozilla :) [14:17] and he has some questions ;) [14:18] zbraniecki: one task would be to make a _real_ RDFDataSource XPCom out of it [14:18] zbraniecki: invite him to get in here [14:19] soooooo [14:19] bumb [14:19] let me give zbraniecki: is it RDF? [14:19] [15:13] zbraniecki: what is the source of data for those templates? [14:19] [15:14] zbraniecki: asac: ping [14:19] [15:17] asac: zbraniecki: the datasource is not a real rdf datasource [14:19] [15:17] asac: zbraniecki: it uses RDF and parses that to a structure [14:19] [15:17] asac: (e.g. pluginInfo) ... which is then used by the Wizard.js to construct the xul tree [14:19] what's the plugin data in? [14:19] [15:17] zbraniecki: ok. because next to us we have an owner of RDF module in Mozilla [14:19] [15:17] zbraniecki: and he has some questions [14:19] [15:18] asac: zbraniecki: one task would be to make a _real_ RDFDataSource XPCom out of it [14:19] [15:18] asac: zbraniecki: invite him to get in here [14:20] asac: why would you add another RDF use case? y'know, RDF is bound to die :-) [14:21] I can perfectly see a case for using xul templates without RDF, though [14:24] PikeMac: ok fine. its just that the current webservice already ships RDF [14:24] pfs? [14:24] PikeMac: what dataformat would you suggest for requesting plugin database [14:24] PikeMac: pfs + ubuntu-pfs [14:24] yes [14:25] s/requesting/querying/ [14:26] now, I can see that poor pfs returns RDF/XML, like all our webservices have to [14:26] but you're more interested in what the user has currently installed, or are we talking about the real plugin installation experience? [14:28] PikeMac: we are talking about multiple points ;) [14:29] PikeMac: i want to start with plugin finder service as this is the most neglected thing the world has seen [14:29] PikeMac: in the end i want plugins to be installable, like extensions [14:29] don't we all :-/ [14:30] PikeMac: in addition i want the plugin finder wizard to be replaced with the addons dialog opening with a search pre-run [14:31] PikeMac: since doing all this right requires also work on the way plugins are manged, i want to do it in small steps [14:31] for instance you cannot really install multiple plugins and then select which plugin to use for which mime-type and such [14:32] PikeMac: so the first step is to move the current plugin finder wizard to a new tab in the addons dialog [14:32] PikeMac: for now that tab is hidden unless spawned by the "install missing plugins ..." button [14:33] Bug 246694 [14:33] Launchpad bug 246694 in firefox-3.0 "Can Not Uninstall Add-Ons" [Undecided,Incomplete] https://launchpad.net/bugs/246694 [14:34] PikeMac: once we manage to understand what is exactly required for managing installed plugins in a reasonable fashion we probably can include general search capabilities for plugins in the "get addons" tab [14:34] PikeMac: but thats later on [14:34] I wonder how we're locally representing installed plugins. I bet it'd be good to focus on that data format for plugin representation rather than RDF [14:34] PikeMac: no that sucks much more [14:34] look at pluginreg.dat [14:34] ;) [14:34] thats the "local" presentation of it [14:34] I do [14:35] well, that's just the file format [14:35] anyway. i dont really care which data format is used. i am more interested in other things. [14:35] you shouldn't be surprised to find that data in sqlite at one point [14:35] for instance: allow new "install methods" to be plugged in [14:35] so we can inject .deb packages in the results [14:35] and rehat can inject .rpm packages [14:35] smells like category manager to me [14:36] PikeMac: yes. thats the direction i think [14:36] PikeMac: same pluggability is required for Extensions too :) [14:36] so: the two main tasks here are: make install (or even search) extendable for _all_ addons [14:37] and 2. replace the plugin finder wizard by a usually hidden "wizard" tab in addons dialog [14:37] using richlistbox [14:37] and xul template ;) [14:37] e.g. not really a wizard, but just a result page where you can install one of the results you get [14:38] who's shepparding that code in on the mozilla side? [14:41] asac: so you just want to present results in the form extension brower in addons panel works like [14:41] what's the data source - ubuntu or AMO? [14:41] PikeMac: not sure. i probably will talk to people at firefox summit [14:42] PikeMac: if the patches make sense and people are _for_ improvement in this area, I doubt that its a problem [14:43] it seems to be a happy mix between bug fixes and UE decisions, which scares me [14:43] zbraniecki: the datasource can be either AMO or ubuntu. [14:43] cool! [14:43] PikeMac: thats true [14:43] I like that idea [14:43] PikeMac: but well. thats always the case for new features, isnt it? [14:43] sure [14:43] new features uncover bad decisions in the past and require some refactoring [14:44] ;) [14:44] PikeMac: ill try to split these things down into comprehensible steps/tasks we can then discuss with you or other mozilla developers [14:44] asac: i'm compiling... [14:44] glibc-2.8 and gcc-4.1 [14:45] luckily I won't have a lot more to contribute than "avoid RDF" ;-) [14:45] zbraniecki: in a perfect world you could add new repositories that get included in search ;) [14:45] PikeMac: whats the substitude now? DB? [14:45] say "sqlite" if you want to land stuff :-) [14:45] hehe [14:45] yeah ;) [14:46] its just a legacy thing and how we can get to where we want [14:46] keeping the RDF backend for the first steps seemed logical to me ;) [14:46] especially since the results usually come from web services [14:47] asac: yea, I agree with asac. Otherwise it'll be not deliverable in the time of XioNoX internship [14:47] actually, that might be a counter argument, if you're reading critical data from a webservice without validating the data [14:48] PikeMac: what do you mean? [14:49] What happens if some error on the server sends junk, incomplete data, whatever [14:49] Hi asac, tonight I have signaled to you a bug here, perhaps you have see it? [14:50] Volans: hmm not sure? [14:50] PikeMac: XML parser will choke [14:50] for the bug 243064 news2000 has made a redirect on the site, so the bug is now override, but seems to be confirmed in Firefox 1.5 on Dapper (I think I don't have the Release Notes link in the Help menu on FF1.5 on Dapper... I don't have see it yesterday when trying) [14:50] Launchpad bug 243064 in ubuntu-website "page "releasenotes 604" not found" [Undecided,Fix released] https://launchpad.net/bugs/243064 [14:50] I don't know if you want to add the MozillaTeam and/or the Ubuntu Firefox package to the bug [14:54] asac: not on malicious code :-/ [14:55] anyway, I need to catch a plane [14:55] PikeMac: well. the user (or most likely the distro) will add a trusted URL [14:55] which can use https [14:55] and so on [14:56] you cannot entirely prevent high-jacking of web-services, but if you assume that thats safe it should be fine imo [14:57] well, I'm just saying it's an argument, not necessarily a good one [14:57] gotta run [15:05] gah [15:05] xul-1.9 fails to build with libXrender-0.9.2 [15:13] armin76: bumb [15:13] 0.9.4 works [15:13] asac: looked at it? [15:14] * asac looking now [15:15] Volans: commented [15:15] ok [15:15] thanks [15:17] Volans: ok added firefox task ... which is now "low" ;) [15:18] asac: i know [15:18] but i didn't want to upgrade to xorg-1.4 [16:31] asac: how are you supposed to know if the patch did something? [16:32] hehe [16:32] armin76: set breakpoints? [16:32] well, doesn't crash [16:33] armin76: the remote client? [16:33] ah [16:33] i didn't try that :D [16:33] for me the main app is fine ... just the remote clients segfaults on exit [16:33] works [16:36] armin76: do you see a difference in performance? [17:57] ok off for a few [18:14] Hi all , ping FTA !! :) [18:14] fta2 it was... [18:15] asac: have you seen Fabien ? [18:59] hello guys [19:00] I've just seen this bug opened by asac 10 hours ago [19:00] https://bugs.launchpad.net/ubuntu/+source/firefox-3.0/+bug/247494 [19:00] Launchpad bug 247494 in xulrunner-1.9 "1.9.0.1 / 3.0.1 security stability update" [High,In progress] [19:00] somebody knows what's that? [19:01] from the description I can't figure it out [19:04] eh? [19:05] that means he's going to add xulrunner 1.9.0.1 and firefox 3.0.1 [19:16] hello asac! [19:16] I was asking just now some information on your bug #247494 [19:16] Launchpad bug 247494 in xulrunner-1.9 "1.9.0.1 / 3.0.1 security stability update" [High,In progress] https://launchpad.net/bugs/247494 [19:17] 247494 [19:17] #247494 [19:17] nite bot :) === asac_ is now known as asac [19:23] how can I take a look at the changes between current ubuntu ff3 and the next ff 3.0.1 ? === rZr is now known as RzR [19:47] I have some news about the gmail slowness problem [19:47] of firefox3 [19:47] two evidences so far [19:48] 1) if we use the old gmail UI (clicking on Old Version link) the scrolling gets pretty fast [19:49] 2) it seemed to me that the scrolling issue has been resolved some months ago, and then recently reappeared [19:49] well, speaking with other FF3 users I've found that I'm not the only one that discovered this fact [19:50] so I conclude that there must be a recent package upgrade that has slowed down firefox 3 scrolling [19:50] for example, I remember a recent libcairo upgrade [19:50] any ideas about that guys? === asac__ is now known as asac