[00:10] asac, lp:~jazzva/firefox-extensions/mozilla-ctxextensions.ubuntu are ready for merge with ubuntu-dev's branch. Rev 15 is the one. If it's still not showing up, wait a minute, it will :) [00:11] asac, and I'll start using "open tree for development", UNRELEASED as dist and similar in the next cycle of updates :) [00:57] asac, lp:~jazzva/firefox-extensions/mozilla-livehttpheaders.ubuntu are ready for merge. Rev 17 is the one... [00:57] s/are/is/ [01:02] ok off. have a dentist appointment tomorrow morning and dont want to be completely wasted there ;) [01:02] uh... dentist [01:02] yeah [01:02] umm... good luck :) [01:02] hopefully can help me without much pain ;) [01:02] ill try to do those two first thing in the morning [01:02] take your time [01:02] well. things should go up ;) [01:02] after dentist is fine too :) [01:02] hehe yeah [01:03] not before for sure [01:03] quite early [01:03] cu [01:03] cu [01:03] asac: do we have vimperator packaged? ;-) === asac_ is now known as asac [08:02] Nafallo: vimperator? [09:01] Jazzva: did we ever file a bug for the webkit MIR? [09:04] bug 250738 [09:04] Launchpad bug 250738 in webkit "[MIR] promote webkit to main" [Undecided,Confirmed] https://launchpad.net/bugs/250738 [09:21] hi [09:22] hi XioNoX [09:24] Back at work [09:25] I've miss my flight yesterday, so it is why i wass not connected [09:29] XioNoX: oh. so you have to fly in every week? [09:30] no, i've just left 1 week [09:30] now i'm up to work 5 weeks [09:31] how did you manage to miss your flight? ;) [09:36] I was working at a music festival, from the 16 to the 20 (helping the computer team), and I've left earlier (midnight) the sunday (after seeing the Hive) to go back home (150km). And my flight was at 6 am. I've decided to sleep 2h at home, but i was so tired that I haven't hear my alarm clock. [09:37] but the plane company change my flight for free, so I've arrived yesterday afternoon [09:41] XioNoX: argh. sorry for that. i feel with you [09:41] XioNoX: (sorry for missing flight) ... great to hear that they changed for free [09:41] :D [09:42] it was this festival : http://www.vieillescharrues.asso.fr/festival/programmation_2008.php [09:43] motorhead still exists ;) ... fun. [09:45] it was cool [09:45] asac: http://vimperator.mozdev.org/ [09:45] :-) [09:45] lol [09:46] Nafallo: reads like good crack ;) [09:50] asac, i've read all the xpcom tutorial, and i don't realy know how we use it (and why) [09:54] XioNoX: what is unclear? [09:55] what does an XPCOM component? [09:56] what it is used for ? [09:58] XioNoX: thats a tough question ;) [09:58] i know [09:59] but maybe just an exemple will be enough [09:59] XioNoX: ok. components implement interfaces that define a service in a cross-language-fashion ;) [10:00] implement interfaces ? [10:00] XioNoX: do you know OO programming? [10:00] Object-Oriented [10:00] a little bit [10:00] with C++ [10:00] then you should know what an interface is ;) [10:01] ok in C++ its similar to the class definition [10:01] e.g. just the contract, not the implementation [10:01] ok [10:01] class declaration is the right word, sorry [10:01] so now instead of writing [10:01] class A { [10:01] public: [10:01] int x; [10:01] int func(); [10:01] }; [10:01] you have a syntax that is language independent [10:02] namely IDL [10:02] Interface Definition Language [10:02] you can find plenty of .idl files in your mozilla source tree [10:03] so what xpcom provides is that you can call and implement such classes/objects in a across-language boundaries [10:04] ok [10:04] e.g. you can implement a Car class in javascript and call it from C++ or vv [10:05] now there are two use cases of xpcom interface implementations: [10:06] 1. service - only one instance exists per process (like a singleton) [10:06] 1. instances - you create new instances of a certain interface implementation as you like [10:06] 2. ;) [10:07] does that help you already? [10:07] yep [10:08] XioNoX: so one thing you can do with interfaces is that you can have multiple implementations of the same [10:08] it is more clear now [10:08] XioNoX: which comes handy in our case as we want to make something extensible in a pluggable fashion [10:09] ok [10:09] and I'm reading the rdf stuff [10:10] ok [10:10] the dor is bigger but it is easier to understand [10:10] dor? [10:11] sorry, doc [10:47] Nafallo: https://www.mozdev.org/bugs/show_bug.cgi?id=19598 [10:47] www.mozdev.org bug 19598 in General "Please include License.txt file in your released .xpi" [Normal,New] [10:47] Nafallo: could you add that extension to the Firefox3Extensions page in wiki [10:47] so it doesnt get lost? [10:47] sure [10:47] http://wiki.ubuntu.com/MozillaTeam/Firefox3Extensions [10:48] Nafallo: i think it should go into the "Missing Detail" table ... at best with the bug link above somehow attached so we know where discussion is going on [10:48] maybe put the bug link in the "extension developer contact" column ;) [11:00] added :-) [11:08] buuuuuuuuuumb! [11:08] * armin76 files 30 bugs for bumbing [11:13] lolz [11:13] asac: you broke firefox on ia64! [11:13] ah [11:13] its the dependencies stuff [11:14] still, bets that it segfaults on ppc and sparc? :D [11:15] on my way [11:19] armin76: BUMMMMMBED! [11:19] deep impact [11:19] lies [11:19] on hardy-updates [11:19] what a breathtaking moment [11:19] armin76: [11:19] 12:18 < asac> so ffox and xul sources+ bits can go [11:19] 12:18 < pitti> asac: awesome; copying [11:21] bumbing! [11:58] asac, thanks for opening the bug for webkit. I asked you a few days ago should we open a bug now, or when we get the review. You probably missed it, and the I forgot to re-ask [12:00] Jazzva: no problem [12:01] https://wiki.ubuntu.com/MainInclusionProcess [12:01] thats the process [12:01] wow... sage is also updated. I thought they stopped its development. Guess I'll update it now :) [12:01] asac, thanks :) [12:02] good [12:04] Jazzva: when you done with that ping me so i can upload all three ;) [12:04] * asac going lunch ;) [12:04] k :) [12:04] have fun [12:04] it's gonna be four of them. foxyproxy needs update, too :) [12:09] thats [12:25] asac: omggggggggg! [12:25] looks like sparc intrepid doesn't sigbus! :D [12:43] uh, wtf [12:43] why doesn't ports.ubuntu.com have hardy-updates and proposed on Release? [12:43] asac: fix! [12:44] bah, stupid! [12:55] uh...nah, it gives bus error anyway [12:55] takes a lot more, though [12:55] on hardy it just gives bus error at the start [12:56] armin76: ok lets be focussed [12:56] what patches shall i apply [12:58] no idea :P [12:58] i don't apply any patch to fix the sigbus [13:00] asac, can mozilla-devscripts use patching system? [13:01] *package using mozilla-devscripts [13:01] Jazzva: it could, but wouldnt make much sense [13:01] except when we cherry-pick to stable release branches [13:01] but for that we can also merge individual commits imo [13:01] so, to change the source directly? [13:01] sorry... I meant can package using mozilla-devscripts use patching system? [13:01] Jazzva: mozilla-devscripts is a native package . there is no orig.tar.gz ... so you develop _inside_ the package [13:02] Jazzva: ah [13:02] havent tried ;) [13:02] but why not [13:02] can I just use quilt and include in the rules [13:02] ok, i'll try [13:02] as long sa the .xpi is produced by build command [13:02] maybe we need to tweak a hook or something [13:03] Jazzva: Jazzva but in general the idea is t change _in-source_ [13:03] in the .ubuntu branch [13:03] aha... ok [13:03] i dont expect lots of patches to accumulate in extensions [13:03] usually just a maxVersion bump [13:03] and maybe a quick fix or cherry-pick for some stable updates [13:04] well, sage package has some patch that should fix two CVEs. Now I'm wondering if that still applies... The code didn't change, so that's why I wonder [13:04] armin76: so is it a sigbus in libc? [13:05] Jazzva: how intrusive are those changes? [13:05] can you post them? [13:06] if they are not packaging related we should push it to upstream i guess [13:06] not too big [13:06] http://paste.ubuntu.com/29282/ [13:06] but look reasonable [13:07] looking at the upstream bts [13:08] asac, who have decided to keep the old skin on firefox 3 whereas the beautiful new one ? [13:14] asac, it's marked as fixed here [13:15] https://www.mozdev.org/bugs/show_bug.cgi?id=15101 [13:15] www.mozdev.org bug 15101 in Javascript "sage fails many of the tests in the feed reader XSS test suite" [Critical,Resolved: fixed] [13:22] XioNoX: he? [13:22] XioNoX: linux uses gnomestripe [13:22] er have that as default [13:23] why they don't have the big back button ? [13:23] XioNoX: thats windows [13:23] and mac [13:23] it is the "default" firefox theme [13:23] no ? [13:23] most likely moz devs didnt care _enough_ [13:24] for linux [13:24] asac, I'll test sage without the patch on some security tests which are reported in bug report. [13:24] to do that or thought that it wouldnt fit nicely into gnome desktop [13:24] Jazzva: the bug is fixed [13:24] according to bug tracker ;) [13:25] yes pleaes test then [13:25] asac, umm... that's what I saw :) [13:25] maybe we can drop them ;) [13:25] asac, but I would like to [13:27] asac, for exemple, this theme will fit perfectly in gnome : https://addons.mozilla.org/fr/firefox/addon/7379 [13:27] and he have the firefox 3 looks [13:27] XioNoX, I think it's because FF will use icons from current gnome theme [13:28] s/will use/uses/ [13:28] and who manage that ? [13:28] umm... what do you mean? [13:29] FF picks up the icons from your current gnome theme and uses them. If they used only one icons for back and forward, it would fit well with the rest of the gnome desktop. [13:29] because i've heared a lot of people complaining about the firefox 3 theme on ubuntu [13:30] and why we don't install by default a firefox skin [13:30] ? [13:30] As far as I can see, if they wanted to use big back/small fwd button, they would have to hardcode the icons. That way, it would fit with the rest... [13:31] XioNoX, we do have default FF3 theme... It's the one we use. It's just that it's a bit different than from win/mac skin. It's adapting to your current theme. [13:31] XioNoX: you are the first i hear complaining ...everone else i talked to cheered ;) [13:31] but we can install a theme like the one i've show you easily through amo [13:32] XioNoX: we can certainly package skins as xpi in ubuntu, but that wont become the default for sure [13:32] another exemple here : http://brainstorm.ubuntu.com/idea/5754/ [13:32] XioNoX, it's true. But that would break the uniformity of the look of the desktop. [13:32] we are happy that we are finally integrated with gnome theme [13:32] +1 on that :) [13:33] ok [13:33] "Linux has the best theme for Firefox 3. [13:33] Just look at the Windows theme for Firefox 3 it looks absolutely horrible. " [13:34] "Proto is hated by Mac users, because of the roundness and the big back-forward button, they like more this: http://www.takebacktheweb.org/ . [13:34] Themes for Windows look completely amateur. Both. The colors, the shapes... Hahaha... " [13:34] XioNoX, that's the way firefox developers designed it. Windows and Mac have already built themes. But, afaik, windows' and mac's desktop environment are not that easily skinable. [13:34] i think this is something you cannot find a solution where everyone agrees [13:34] XioNoX, that's why the fixed theme in win and mac won't break the look of the rest of the desktop [13:34] as a matter of fact the default gnome theme isnt that bad [13:35] On the other hand, gnome is designed to be easily skinable. If you would use fixed theme for Fx, it wouldn't fit well with the rest of the desktop. I suppose that's the reasoning of FF developers. [13:36] ok [13:36] Jazzva: yeah. but the size of the back button could probably not be tuned [13:36] thats a design devision i guess [13:36] decision [13:36] that explaination it enough for me [13:36] asac, that too... If they wanted to use big button, I suppose all gnome themes should add one big button... [13:36] XioNoX, I'm glad we could help :) [13:36] anyway, we can package themes up if there is really demand [13:36] not sure about that though [13:38] ok [13:41] I didn't know that the firefox theme was linked to the gnome theme [13:43] asac: i have no clue [13:44] asac: as i said, it takes more time to sigbus than in previous versions [13:44] i don't know if thats due to firefox or libc [13:44] need to test on ppc [14:02] asac, this is the feed that has the tests [14:02] http://www.gnucitizen.org/blog/cross-context-scripting-with-sage/feed.xml [14:03] I have tested it. It does interpret it as script, as no text is shown, but it doesn't give me the alert window... [14:04] I'll test with the patch now [14:04] if it shows that script as text, I'll use the patch [14:04] Jazzva try to alert in the exception case for the unix paths [14:05] e.g. catch (e) {} => catch (e) { alert('here goes the rdv'); } [14:05] mhm... ok [14:05] at best the exception [14:05] whats the CVE id? [14:05] (for that test?) [14:06] CVE-2006-4711 and -4712 [14:07] test is located at mozdev bug [14:07] asac: hey there! [14:07] mozdev bug 15101 [14:07] Launchpad bug 15101 in linux-meta "linux-image-2.6-686 depends on linux-image-2.6.8.1-3-686 for Hoary" [Medium,Invalid] https://launchpad.net/bugs/15101 [14:07] hmm [14:07] https://www.mozdev.org/bugs/show_bug.cgi?id=15101 [14:07] www.mozdev.org bug 15101 in Javascript "sage fails many of the tests in the feed reader XSS test suite" [Critical,Resolved: fixed] [14:09] ubottu: CVE-2006-4711 [14:09] Sorry, I don't know anything about cve-2006-4711 [14:09] hi jt1 === jt1 is now known as jtv [14:10] ha, that's better [14:10] jtv: you should start a _real_ irc client ;) [14:10] asac: still haven't finished our discussion of the file_references issue! [14:10] we havent? [14:10] ;) [14:10] asac: I used to, and people told me to use this one instead. :-) [14:10] not that i thought otherwise :-D [14:11] Jazzva: i think the feed.xml is to show CVE-2006-4712 [14:12] asac, if you are not busy, can we talk about what do you want that I do ? because xul templates tuto is very boring, it would be cool to know on what i'll apply it, and using what already exist to learn. [14:12] for CVE-2006-4711 executing javascript without cross context appears to be enough [14:12] same for xpcom [14:12] XioNoX: sure [14:13] XioNoX: lets start with a simple modification of the "Get Addons" tab in addons dialog [14:13] so you get used to that huge code ;) [14:13] ok [14:13] asac: anyway, the question was this: if we start using proper XPI paths for file references in new XPI imports (as opposed to the pseudo-XPI paths we have now), would that be a big problem for you? [14:14] XioNoX: do you have a hg trunk tree? [14:14] XioNoX: do you have a place where you can push your "topic" branches? [14:14] nop [14:14] asac, tested... not getting any alerts :/ [14:14] not yet [14:14] XioNoX: ok so for now we need patches then i guess ;) [14:14] (to show me) [14:14] ? [14:14] Jazzva: ok. so the script is not executed at all? [14:15] XioNoX: well ... since you cannot publish your private hg branches you will be working on you need to show me patches when you want feedback on your changes [14:15] asac, it looks like. But I'm still confused why it hides the script code. If it's not executed, it should be visible, since it's marked as text [14:16] XioNoX: ok. you need a hg clone that you can build [14:16] XioNoX: doy ou have that already? [14:16] nop [14:16] i've nothing yet [14:17] XioNoX: ok, get such a tree then ;) [14:17] http://developer.mozilla.org/en/docs/Mercurial [14:17] and make it build [14:17] i've already used mercurial [14:17] but not for firefox [14:17] XioNoX: yes, but you have to build the tree ;) [14:17] i need to get all the sources of firefox ? [14:17] otherwise you cannot test your changes [14:17] XioNoX: yes [14:18] and what is the url of the trunk ? [14:19] XioNoX: so first thing is to add a new tab like the "Get-Addons Tab" to the addons dialo [14:19] ok [14:19] http://developer.mozilla.org/en/docs/Mozilla_Source_Code_(Mercurial) [14:20] add a tab like the extention, theme, laguage, etc... ? [14:20] the code we will be mostly working in is toolkit/mozapps/extensions/ [14:21] XioNoX: yes. thats the idea. dont spend too much time on detail. we have to think how it should actually look like and probably need a few prototypes later :) [14:21] XioNoX: call it Plugin-Wizard ;) [14:21] or whatever [14:22] and then we will try to move the plugin finder service there ;) [14:22] http://hg.mozilla.org/ ? [14:22] which one should I get ? [14:22] read the urls i gave you :) [14:23] mozilla-central ? [14:23] yes [14:23] its all on the page ;) ... http://developer.mozilla.org/en/docs/Mozilla_Source_Code_(Mercurial) [14:23] The integration repository for Mozilla 2 development is mozilla-central [14:23] ha [14:23] Mozilla 2 [14:23] not firefox 2 [14:24] mozilla 2 ... but current focus should be 1.9.1 [14:27] done [14:28] 34564 files updated [14:28] 260,0 Mio [14:29] :~/Stage/Plugin-Wizard/src/toolkit/mozapps/extensions$ ls [14:29] content jar.mn Makefile.in public service src test [14:29] there are a lot of files [14:29] XioNoX: do a build first [14:29] want to be sure you can test what you code ;) [14:30] run allmakefiles.sh ? [14:30] its on that website ;) [14:31] i try to go too fast [14:31] most likely [14:34] /bin/sh: autoconf-2.13: not found [14:35] but autoconf2.13 exist [14:35] most likely AUTOCONF is wrong then [14:35] in .mozconfig [14:36] okay it work [14:36] I have all the -dev package to install [14:36] asac, applying the patch produces the same result [14:37] guess we can drop it then :/ [14:41] asac, configure: error: Could not compile basic X program. [14:41] now there's another problem... sage and sage-too replace each-other in Tools menu... they both use the same shortcut. Maybe using Conflicts field in debian/control? [14:41] XioNoX: you lack build dependencies [14:41] install the ones we use for xulrunner-1.9 and firefox-3.0 package [14:42] Jazzva: are sage and sage-too independent projects? [14:42] sage-too is a fork of sage [14:42] but both are active? [14:42] yes [14:42] I thought sage is inactive until yesterday... [14:43] but then I found out it has update dated june 23rd on amo [14:43] why cant we fix the overlay? [14:43] so they dont conflict? [14:43] hmm... good :) [14:43] I'll push sage unchanged, since sage-too is a fork [14:43] I'll also see with the upstream to fix this, if they can [14:43] ok. do we have sage-too alredy in archive? [14:44] not atm... waiting to see about copyright issues [14:44] sage also misses LICENSE file, but copyright mentiones it's licensed under trilicense, so I suppose it was sorted out before. Anyway, I sent an e-mail to the upstream to see if they can add license... [14:44] it is running [14:45] s/also misses/is also missing/ [14:45] XioNoX: good [14:45] XioNoX: so in public/ there are the .idl files [14:45] look at them ;) [14:46] nsIAddonRepository.idl [14:46] defines abstract contract to implement backend that gets result from addons databases (like amo) [14:46] we could use that to search for extensions available in ubuntu [14:47] but thats not the start here ;) [14:47] the xul pieces are in content/ [14:47] asac, mozilla-ctxextensions, foxyproxy, mozilla-livehttpheaders and firefox-sage are ready for upload [14:47] do you need branch links? [14:47] modules/libpr0n [14:47] o_O [14:48] XioNoX: yeah. you dont want to touch that unless you really want to get dirty [14:48] lol [14:48] haha :) [14:49] XioNoX: so the idea is to add a nsIPluginRepository.idl and use that to implement the plugin-by-mime-type-search that we want to put into the new tab [14:50] ok [14:51] asac, sorry... I forgot to add "New upstream release" to changelog for firefox-sage. I pushed new revision [15:04] asac, will I have to build firefox at each modification ? [15:13] XioNoX: yes and no [15:13] usually just the subdirectory needs to be rebuild [15:13] ok [15:13] and i don't really know how to start [15:14] XioNoX: yeah. you probably should try to match what you see with the code you read to get a basic idea [15:14] and then look more in detail [15:14] ok [15:49] asac, the problem might also be that they're using sage instead of sage-too name in chrome.manifest [15:50] thats not so great ;) [15:50] yeah... [15:50] we can change that here, but should tell upstream about thta [15:50] it needs a lot of changes [15:50] well... there is one neat (maybe not so clean way) [15:50] s/sage/sageToo/ [15:51] s/Sage/SageToo/ [15:51] :) [15:53] hmm... and both sage and sage-too have minVersion set to 3.0b1 [15:53] asac, please don't upload firefox-sage for now. I would like to see if it's compat with FF2 [15:58] asac, i don't know what i have to put in the idl file [15:58] i've put the default lines, but it is all [16:11] XioNoX: thats a good start [16:11] XioNoX: you can now go to the xul part ;) [16:12] and add a new item to the addons dialog [16:12]