[00:07] asac: fta: what was that about:confix to reeanble process separation ? [00:55] chrisccoulson: isnt libnss3-0d empty? [00:56] hmm. the .0d links are in there [02:00] Does google do anything different when it builds google-chrome compared to how chromium-browser is built? After trying both, (both the same -dev version number) I think that google-chrome is just a tiny bit faster. This is a completely subjective comparison, but I was wondering if google did anything special with it's builds, like how firefox uses profiling with it's windows builds. [02:13] ripps: only thing that is probably differentis that they use a different gcc version or something [03:13] asac, moments away from tests :) [03:13] it only took 24 hours longer [03:28] * cwillu_at_work waits for ff to install [04:10] * cwillu_at_work waits for ff to launch [04:23] seems to be working [04:24] asac, karmic's 3.5.3 built from source shows the extra scrollbars in google maps [04:26] asac, amusingly, the error console is _missing_ its scrollbar : [04:28] although that may be related to my build more than anything [04:54] performance with the patch seems quite a bit better; I'll do some proper a/b testing tomorrow [04:54] patched firefox gets the slower beagle :) [08:34] hi, I've some problems with sites made with silverlight. I tryed to install the moonlight plugin by synaptic but the version is too much old (1.x). So, I tryed to install the new version directly from novell site... So, does anybody know if exist any reason why the ff should not work? [08:35] lonejack: lots of reasons [08:35] lonejack: which Ubuntu version? [08:36] What make me wonder is that I've a friend that work with debian and his system works wery well. VErsion 9.10 [08:38] lonejack: it might work [08:40] micahg: but it doesn't. Can we investigate why? [08:41] lonejack: which version of Firefox? [08:41] 3.5.9 [08:41] 64 bit [08:42] ubuntu [08:42] Novell moonlight 2.2 [08:42] I'm not sure what the compatability status of moon 2.2 with Firefox 3.5 is [08:44] we've actually had some problems with 3.6 as well [08:45] what version was your friend using? [08:45] I understand but, after moonlight synaptic installation, what other choose did I have? [08:46] lonejack: idk, I'm just saying that I don't know offhand the compatability of the upstream moon [08:46] I know that 2.2 is in Lucid and we are having a few issues with Firefox 3.6 [08:47] Squeeze [08:48] ? [08:48] in beta version [08:48] Lucid != Squeeze [08:50] Do I told Lucid? [08:50] ? [08:53] Anyone know if the ubuntu-mozilla-daily packages are supposed to have the Lorentz crash-isolation feature? [08:53] micahg: why U wrote "Lucid != Squeeze"? I know that my friend works with debian in beta version and he can see a lot of sites make with silverlight.... [08:53] When I last tried the PPA (a few hours ago), it didn't. [08:53] lonejack: ah, ok, I thought you were saying Lucid is Squeeze [08:53] DanaG: FF 3.7 has had it for a while, but it hasn't worked (since i last tested) [08:54] DanaG: 3.6 and 3.7 should [08:54] The version of Mozilla's own releases is 3.6.3plugin1 [08:54] The version on PPA is 3.6.4pre.... yet it didn't isolate plugin crashes. [08:54] DanaG: yes, bit should [08:54] DanaG: what happens? [08:55] google for "flashcrash" -- that page about "bug hasn't been fixed in years". [08:55] Firefox just freezes. [08:55] Doesn't crash.... just freezes. [08:55] it's been doing that for a while.. [08:55] DanaG: yeah, seems like we have the same problem in 3.6 as 3.7 now [08:55] The official beta 3.6.3plugin1 does isolate those crashes. [08:56] when you use the source? [08:56] which isn't good because upstream thought it was a separate xulrunner -> firefox issue, but 3.6 has its own xul [08:56] DanaG: yeah, we need to track down the issue [08:56] Oh, and the thunderbird 3.0 dailies still don't allow non-integer layout.css.csspixelsperdevpx [08:56] 147 DPI display... means I need that to be 1.5. [08:57] DanaG: i don't remember when that was fixed [08:57] * micahg goes to look [08:58] * micahg thought upstream fixed that in trunk [08:58] DanaG: does it work in Firefox 3.6 or 3.7? [09:00] hmm, which issue now... dpi, or crash-isolation? [09:00] dpi [09:00] DanaG: dpi [09:01] It's fixed in Firefox 3.6; now only Thunderbird is left to fix. [09:01] DanaG: then it will get fixed in Thunderbird 3.1 [09:01] Cool. [09:11] but in general these problems with moonlight when will be solved? [09:12] lonejack: well, squeeze and karmic have the same version of moon [09:13] lonejack: maybe before release, maybe after [09:14] micahg, your advice is to remove the plugin and install the one that stays on repo? [09:15] lonejack: i don't know if I have any advice ATM [09:17] micahg, exsuse me what it means ATM? [09:17] lonejack: At The Moment [09:17] lonejack: at the moment [09:17] whoa [09:17] ddecator: exactly at the moment ;) [09:18] every time is the right time to learn something new... [09:18] lonejack: well said [09:23] ok, last question. is there a difference between install FF pluging from repo or directly from direct download? [09:24] repo has more testing and is generally more compatible [09:24] but generally a lot of them are more for ease of install [09:25] you shouldn't notice a difference [09:26] lonejack: we're dropping a lot of the addons from lucid [09:35] micahg, ddecator: there is a new(beta) version on moonlight. is it wasted time to try to install it? [09:36] lonejack: idk [09:38] lonejack: they say it works [09:38] lonejack: I'd say give it a try [09:39] this moonlight plugin has something mystic. [09:40] also the 3.0 doesn't work... [09:41] lonejack: try a new profile: firefox -ProfileManager [09:41] today it's sunday not christmas [09:42] ..what? [09:42] micahg, I launched that cmd from terminal. It has opened a new ff window [09:42] is it correct? [09:43] lonejack: no, you have to close firefox first, or use: MOZ_NO_REMOTE=1 firefox -ProfileManager [09:46] micahg, something has chaged I obtain a message from terminal: "(firefox:3887): GLib-WARNING **: g_set_prgname() called multiple times" [09:47] then a dialog choose user profile [09:47] lonejack: that's normal [09:47] create a new profile [09:50] i did. so I went on a page written in silverlight... It show me a message that inform me that is necessary to install moonlight... Click here??? Moonlight is already installed!!! [09:51] No I did a mistake. The moonlight disappered.. [09:52] that's what creating a new profile does [09:52] it's a clean version, no add-ons, default settings [09:53] at this point what I have to do? Install it doing click or synaptic? [09:53] (you launch the profile manager again later to switch back to your normal profile) [09:53] not sure what micahg wanted you to do [09:54] lonejack: wanted to see if a clean profile helped moon [09:54] micahg, this is a test, ok: what U prefer: Install it doing click or synaptic? [09:55] lonejack: either [09:55] * micahg needs to go to sleep though, so catch me later with the results [09:57] ok, direct download first. [09:59] micahg, it works [10:00] lonejack: that means that there is likely an add-on or a setting on your default profile causing the issue [10:00] * ddecator would guess an add-on [10:02] lonejack: if you run "firefox -ProfileManager" again, you can select your normal profile again, then disable your add-ons one at a time to see which one(s) is/are conflicting with moon [10:06] Ok. but before I prefer to try to install the plugin from synpatic... [10:07] ok. i'm not sure if the version in synaptic is the latest stable they have on their site.. [10:07] you said you're running 9.10? [10:08] yes [10:09] it doesn't work [10:09] karmic is still using 1.0.1, yes? [10:10] forthermor the moonlight(after synaptic installation) isn't shown in the FF extension... [10:10] the synaptic one should show up under plugins as silverlight [10:10] 1.0.1, right [10:11] so 2.x works, just not the 1.0.1 in the repo [10:12] from synaptic I installed libmoon,moonlight-pligin-core/mozilla [10:12] but on FF extension isn't shown [10:12] do you see silverlight under FF plugins? [10:14] or it might say moonlight, i can't remember for sure. but the synatpic one should be under FF plugins [10:15] ok, now I undestand what you mean. When I installed the 2.2 from downlonad the moonlight was appearing on the extensions, instead now it appear as Silverlight plugin... [10:15] right. from website = extension. from synaptic = plugin. [10:16] moonlight 2.2 is in lucid already [10:16] for karmic, you can continue to use the version from the site [10:16] Ok. Now I've SL plugin installed but on the site I'm looking "please install ML.." [10:17] right, the one from synaptic is outdated and doesn't seem to work right [10:17] So it seems that on the next UBUNTU version (probably) it should work... [10:18] -should-, but according to micah there are reports of it not working for some people with FF 3.6 and 3.7 [10:19] but that might be due to moonlight being behind silverlight, so it can't handle everything [10:19] ddecator: more likely an issue with the migration to xul192 [10:19] is a nightmare... [10:19] * micahg is out for real now [10:19] whoa [10:20] well there's your answer, haha [10:20] ddecator, thank you very much for your help!!! [10:20] lonejack: np =) [10:21] So have you a good sunday... [10:21] bye [10:21] you too, cya [14:04] * cwillu gets out a stick and prepares to pokes asac with it [14:05] cwillu_at_work: ho [14:05] whats up? [14:06] finally got a root image booting up [14:06] haven't tried the karmic's stock firefox yet, but my 3.5.3 rebuild shows scrollbars on gmaps [14:07] and also by virtue of not having tried karmic's stock firefox, I haven't done a proper a/b test between the rebuild with the bitdepth patch, but the rebuild sure as hell feels alot faster [14:08] i.e., it feels responsive while scrolling a page of text, [14:08] I'm heading off to work in an hour or so, and then I'll poke you for real :p [14:10] asac: good morning/afternoon [14:11] chrisccoulson_: good morning/afternoon [14:11] asac, chrisccoulson_: so, what's the verdict? [14:26] chrisccoulson_: there? [14:26] chrisccoulson_: whats your eval? is the P.*FromOrigin Function broken and doesnt check if a link exists? [14:38] chrisccoulson_: so what we should do before droppin the links is checkkng the binary extensions like traybiff and enigmail if they link against the .0d thing === skipper_ is now known as bdrung [15:11] asac, firebug shows the extra scrollbars in its ui [15:12] asac, overflow: hidden seems to trigger an separate scrollbar [15:18] asac, that's odd: in firebug on a bugger element: background: url(...image) no-repeat scroll 0 0; [15:18] if I remove scroll, it pops back in [15:18] cwillu_at_work: yeah-. [15:18] feels odd [15:19] hmm, does the same thing on my desktop [15:20] on your desktop you get a scrollbar? [15:20] no [15:20] but the same behaviour of it popping back in after I delete it [15:20] cwillu_at_work: you said you see the same in chrome? [15:20] or just html? [15:20] firebug had an extra scrollbar, yes [15:20] hmm. [15:21] cwillu_at_work: what element is that? [15:21] hmm. mxr is slow [15:21] the html tab [15:21] it went away at some point, not sure when [15:21] jdstrand: ok so ... did chris upload that to -security ppa? [15:21] otherwise let me do and then test [15:22] in hardy thats reproducible right? [15:22] cwillu_at_work: ok. the tab could be a valid one? [15:22] asac: there is no new tbird in ppa [15:22] don't believe so [15:22] asac: the version of nss is still 3.12.6-0ubuntu0.9.10.2 [15:23] asac: ie, the one with the fips cleanup [15:23] kk [15:23] i was talking about tbird [15:23] jdstrand: thats hardy to karmic? [15:23] asac: karmic. hardy-jaunty didn't get a new nss [15:24] kk [15:24] asac: and presumably lucid [15:24] asac: though it has tbird 3 so I don't know [15:25] jdstrand: do you have the tbird bug id at hand` [15:25] ? [15:25] asac: 559918 [15:25] jdstrand: lucid is a new branch. that hasnt the problem [15:25] cool [15:25] bug 559918 [15:25] Launchpad bug 559918 in nss "Thunderbird cannot initialize the security component when libnss3-0d 3.12.6 is installed" [High,Confirmed] https://launchpad.net/bugs/559918 [15:26] that the one [15:28] jdstrand: what heading line should i use in changelog? [15:29] "Regression upload for nss3 security/stability update - USN-910-2" ? [15:29] oh thats the wrong usn ;) [15:29] asac: you don't need a USN. [15:30] asac: how about "Regression upload for recent nss3 security/stability update" [15:30] asac: with the LP number [15:31] i just skipped it [15:31] oh i failed [15:31] so let me use that [15:32] ' * fix LP: #559918 - Thunderbird cannot initialize the security component when libnss3-0d > 3.12.6 is installed; we drop the dangling .so.0d links that became obsolete when we moved to non-versioned SONAMES for nss3 and nspr4 - update debian/thunderbird.links' --fixes 'lp:559918' [15:32] Launchpad bug 559918 in nss "Thunderbird cannot initialize the security component when libnss3-0d 3.12.6 is installed" [High,Confirmed] https://launchpad.net/bugs/559918 [15:32] thats what it is now ;) [15:32] asac: excellent [15:33] http://pastebin.com/7jzH2U0a [15:33] asac: huge thanks :) [15:33] thats the diff [15:33] heh. not too early [15:33] lets first get this properly tested ;) [15:33] (and built) [15:33] :) [15:33] * asac found he doesnt have the right orig ,) [15:34] * asac goes for launchpad [15:34] so ... killing my decades old irc server from my internal net made my normal bandwith go up ;) [15:35] get 1.2M/s now when downloading from lp [15:35] heheh [15:36] i replaced it with mini 10 with poulsbo chip ;) [15:37] let me screen the recent tbird bug noise quick [15:37] bug 559923 [15:37] Launchpad bug 559923 in thunderbird "package thunderbird 2.0.0.23 build1 nobinonly-0ubuntu1 failed to install/upgrade: trying to overwrite '/usr/bin/thunderbird', which is also in package thunderbird-mozilla-build 0:3.0.4-0ubuntu1" [Undecided,New] https://launchpad.net/bugs/559923 [15:37] hmm [15:37] is that a moz build? [15:37] or some random guy? [15:39] I've confirmed that the nss upload fixes fips. another guy confirmed performing the debian/rules stuff manually on libnssdbm3.so fixes it [15:39] I'm going to upload that [15:39] err... publish that [15:40] yeah. missing .chk should be safe [15:40] if we see and improvement somewhere its probably and improvement for everyone ;) [15:40] chrisccoulson_ also tested it and said it worked [15:41] goodie [15:41] * asac pushes now [15:41] well in two seconds ;) [15:52] [PPA ubuntu-mozilla-security] [ubuntu/karmic] thunderbird 2.0.0.24+build1+nobinonly-0ubuntu0.9.10.2 (Accepted) [15:56] asac, further testing will have to wait an hour or two, my sd card just self-destructed [15:56] protip: don't buy surplus intelligence agency storage devices, they're not a bargain [15:58] asac: looks like i386 is queued to start in an hour. amd64 is building now. I'm going to go afk for a while [15:59] nss is pocket copied, and I'm just waiting on the publisher to finish before I publish nss SUN-927-2 [15:59] s/SUN/USN/ [16:00] SUN ;) [16:00] enjoy [16:01] 09:25 < ogra> asac, same image as mentioned on friday ... http://people.canonical.com/~ogra/lucid-netbook-armel+omap.img [16:01] oops [16:01] ;) [16:01] cwillu_at_work: ^^^ [16:01] you can try that image ;) [16:01] the image isn't the problem :p [16:02] but tats lucid ;) [16:02] just for next time ... [16:02] asac: can you look at the pyxpcom bug to see if sugar really needs it [16:03] I'm writing two new cards right now [16:03] and it's hard for me to express how little javascript performance matters to me before I can scroll the screen at better than 5 fps :p [16:04] micahg: i am sure sugar needs it for some parts [16:04] asac: so the question is do we get sugar in archive for Lucid? [16:04] cwillu_at_work: well. do what you want. i would just suggest to use most recent stuff [16:05] duh :p [16:05] micahg: you mean: are we ok to drop it ? [16:05] i assume its in [16:05] but I'm only going to change one thing at a time [16:05] asac: no, it's out ATM because it can't be built [16:05] thats fine then imo [16:05] it's gonna be a few months before we release anything on top of lucid anyway [16:06] cwillu_at_work: kk. what are you working on if i might ask? [16:06] or secret sauce ;)? [16:07] asac: ScottK wanted to get some version of sugar in for Edubuntu [16:07] industrial hci [16:07] applied real-time-strategy games, basically :p [16:07] nice ;) [16:08] I'm actually planning an isometric system overview screen later this year :) [16:08] micahg: do you understand what pyxcpom is needed for exactly? [16:08] asac: I think they use it for their web browser [16:09] asac: not exactly, I can try to track down one of the maintainers [16:09] yeah [16:09] that would probably be ok [16:10] asac: does that mean we should try to get it in? [16:11] no [16:11] first understanding what the are using it for [16:11] asac: ok [16:11] and if they can stop doing that ;) [16:11] (if its risky) [16:13] lfaraone: are you familiar with Sugar's usage of python-xpcom? [16:14] asac: BTW, fennec and prism built for Hardy :) [16:15] asac: on another note, not really sure about IPC in 3.6.4 [16:15] micahg: yes, it's depended on by hulahop iirc. [16:15] lfaraone: is there a way around it? [16:15] micahg: well, the Browse activity is a Python + GTK wrapper around it. [16:16] lfaraone: can we use python-gtkmozembed? [16:17] micahg: I don't think so. Hulahop gives access to the whole xpcom API to its clients, and I think Browse uses more than gtkmozembed provides. [16:17] I'll check with upstream. [16:18] lfaraone: ok, thanks, please let us know [16:18] asac: I don't think we should have IPC on the release CD if we can avoid it [16:22] micahg: from a cursory reading of http://git.sugarlabs.org/projects/browse/repos/mainline/blobs/master/browser.py ( a small component of Browse :)), it looks like that is the case, unfortunately. [16:22] asac: ^^^ [16:24] micahg: if I had a month to rewrite it, and a few years more PyGTK experience, I might be able to write a reduced-functionality version which accomplishes the same thing, but final freeze is shortly approaching. [16:28] lfaraone: micahg: is that "just" one package requiring it? [16:29] asac: it seems to be one package that uses it and the otehrs depend on it [16:32] maybe that packag ecan build in-source pyxpcom or something? [16:32] usually upstreams already include that in their tree [16:32] micahg: lfaraone ? [16:35] will be back in an hour [16:44] asac: sorry. Uh, sure, we could embed pyxpcom in hulahop, but pyxpcom will need to be rebuilt (I think) whenever there's a change to xulrunner. [16:46] ++-+- [16:47] ~. [16:47] . [16:47] hunspell or myspell? [17:25] asac: is it not feasible to enable building pyxpcom in xulrunner? I think debian's xulrunner has it. === yofel_ is now known as yofel [18:20] asac, chrisccoulson_: fyi-- just tested tbird on amd64 and it works fine. I can connect using pop3s with libnss3-0d installed. with old version I could not. [18:21] asac, chrisccoulson_: see my comment in https://bugs.launchpad.net/ubuntu/+source/nss/+bug/559918/comments/13 [18:21] Launchpad bug 559918 in thunderbird "Thunderbird cannot initialize the security component when libnss3-0d 3.12.6 is installed" [High,Fix committed] [18:22] asac, chrisccoulson_: waiting on armel, i386 and ia64 to finish building (all have started) [18:22] * jdstrand -> afk for a bit === BUGabundo is now known as BUGabundo_snack [19:01] asac: just wondering why you used source format 3 for enigmail [19:01] micahg: because it's great ;) [19:02] bdrung: we're going to need to backport it though [19:02] not good [19:03] bdrung: which, source format 3 or the need to backport? [19:03] micahg: the combination: the need to backport a 3.0 (quilt) formatted package [19:09] bdrung: that's why we can [19:09] 't migrate the mozilla stack to 3.0 [19:09] micahg, fyi, per doko's request, i've suspended all the dailies [19:09] fta: oh, so his rebuild can finish? [19:09] yep [19:10] fta: how long? [19:10] until it's done :P [19:10] fta: would we be able to do just firefox once I figure out why IPC doesn't work? [19:11] should be ok [19:11] fta: ok, thanks, I'll let you know [19:15] lfaraone: pyxpcom is a different upstream source now === BUGabundo_snack is now known as BUGabundo [19:46] wow [19:46] fta: chromium is *eating* http:// from the URL bar [20:39] hey guys, would a text area input bug be filed under "form controls"? [20:40] are you asking about bugzilla.m.o ? [20:40] yah [20:40] it depends. what's the bug? [20:40] text areas only work 25% of the time with 3.7 nightlies [20:40] sometimes you need to click or type elsewhere first [20:41] ridiculously annoying [20:41] if you use message boards [20:43] does it happen for all textareas or only specific ones? [20:43] is it a recent regression? [20:44] it sounds familiar for some reason [20:44] gavin, i really don't know how to distinguish text areas [20:44] for example, ubuntuforums is affected [20:44] well, I mean does it happen on multiple sites, or just ubuntuforums? [20:44] a lot of sites [20:45] with message boards [20:45] do those sites use the same message board software? [20:45] doesn't matter if phpbb, smf, vb, etc [20:45] ok [20:45] micahg, i've brought it up before, never bothered to file a bug yet. [20:46] LLStarks: ah, that's why [20:46] LLStarks: does it happen with upstream builds? [20:46] latest nightlies [20:47] has been like this for a week or two [20:47] haven't tested outside of umd [20:47] go ahead and file it in Core : General [20:47] if yo can find a 1-day regression range using nightlies, that'd be perfect [20:47] LLStarks: that's why I asked if upstream builds were affected [20:48] Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.3a4pre) Gecko/20100407 Ubuntu/10.04 (lucid) Firefox/3.7a4pre - Build ID: 20100331033712 [20:48] feel free to CC me too (gavin.sharp) [20:48] what do i regress against? xulrunner 1.9.3? [20:49] or better yet, what's that site that lists changes for zeroing in for regression testing [20:49] not sure what you mean - I assumed by "latest nightlies" you meant "latest mozilla.org firefox nightlies" [20:49] i remember it being a mercurial site [20:49] https://hg.mozilla.org/mozilla-central/ [20:49] bingo [20:49] thanks [20:49] if you get the changeset IDs from the relevant builds using about:buildconfig, you can enter them into http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=1b944ebb5ca6&tochange=35059e8e8ce8 [20:49] to get a list of changes between the two builds [20:50] what would be some keywords to search for? [20:52] gavin, how can get older mozilla-central build? [20:52] *builds [20:52] *can i [20:52] http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/ [20:52] (the mozilla-central ones) [20:53] thanks again [21:01] window is shrinking [21:01] tedious and fun [21:11] damnit. this is harder than i though. as soon i think i've cornered the bug, it pops in the build i think is pre-regression [21:15] asac: ping [21:16] asac: ping (too) [21:16] * micahg waits for asac_the_second to pop in :) [21:17] let's clone him. then everyone can have his own asac ;) [21:22] gavin: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=2cc5ad2cf917&tochange=5108c4c2c043 [21:23] asac: https://code.launchpad.net/~bdrung/ubufox/m-d-0.21/+merge/22973 [21:23] do i have to go commit-by-commit for further drill down? [21:26] jeez. these tinderbox builds are huge. [21:26] 80mb+ [21:28] i have a hunch that the textarea focus commit is the culprit [21:28] LLStarks: there are several in that range [21:29] where can i get commit binaries? [21:29] the gzip i dl'ed is source [21:30] at least we have a good range === BUGabundo is now known as tellari === tellari is now known as BUGabundo === BUGabundo is now known as BUGabundo_dinner [22:04] LLStarks: yeah, I would suspect one of ehsan's changes [22:04] CC him on your bug? [22:04] :ehsan [22:06] ah [22:06] you suspect the focus one too? [22:06] cork and i figured as much [22:08]