[01:33] does anyone have a karmic installation available and can spare a couple of minutes testing something? [04:55] micahg: btw, i'm working on sb. upstream moved the xulrunner branch on their server and made some changes, so i'm working on getting everything setup. i updated the patches, but i hit a failure at 4am last night. not sure how long it'll take, just wanted to let you know why it's taking so long [04:57] ddecator: k, np [04:57] good learning experience so far.. [06:25] micahg: i feel like this may have come up in other xr192-related builds. when building, it looks for gnomeui libs and it can't find libnotify. have you ever run into that before? [06:27] ddecator: you might need to add a dependency [06:28] * micahg doesn't remember that [06:28] micahg: just to the control file? [06:28] ddecator: yeah, look at the source for xulrunner-1.9.2 to see if you're missing something [06:29] micahg: alright, thanks, i'll look at that in a little bit, i'm introducing a student to the bugday first :) [06:31] * micahg needs to work on midbrowse and it's late [06:32] *midbrowser [06:32] yah, i hope i don't end up staying up until 5am again.. [06:32] * micahg needs to get somewhere by 8 [06:33] yaaaaaah... [07:16] micahg: yup, i had to add some dependencies, we'll see how things go now === almaisan-away is now known as al-maisan [08:38] dang, it's so close to building.. === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan [12:22] http://arstechnica.com/software/news/2010/06/chromium-based-flock-beta-offers-unobtrusive-social-features.ars [15:00] does anyone else have to type in your keyring password when starting Gwibber? [15:01] gnomefreak: makes sense if you're storing passwords [15:02] i didnt realize i was. anyway to change this? and it doesnt even start. ps aux shows it but its not showing [15:04] this damn app is so unrelable not sure how it made into default [15:05] i lost all my accounts [15:05] wtf [15:06] gnomefreak: couchdb upgrade? [15:06] gnomefreak: maybe check in +1 [15:06] micahg: shouldnt have dropped all my accounts [15:06] shouldn't is true, but this is the devel release [15:06] * gnomefreak working on fixing. that i am sure i can [15:07] ERROR:dbus.proxies:Introspect error on com.Gwibber.Accounts:/com/gwibber/Accounts: [15:08] more of those [15:08] ok lets see if i can do this [15:12] it seems that backing up .gconf no longer keeps accounts [15:14] gnomefreak: should be stored in you keyring, maybe new gwibber feature [15:14] ? [15:15] actually, I shouldn't say that for sure, ideally, it would be stored in the keyring, don't know what gwibber does :) [15:27] micahg: ah ok [15:27] i guess i re-set it up [15:28] * gnomefreak not sure if there is a faile to back up for keyring [15:29] s/faile/file [15:43] cant add facebook account no "add" button after filling all info needed [16:02] every time i open firefox, it loads all my previous pages, of which like 10 are protected via htaccess. i always have to click 10 times "ok" (since it remembered username/password). is there an addon or such which automatically uses the remembered data without prompting me? [16:07] DASPRiD: maybe but you would have to look on mozilla addons site. we dont support anything like that [16:07] * gnomefreak thinks it is a bad idea to do that [16:07] gnomefreak, why do you think that? [16:07] username and password are prefilled anyway [16:07] so what's the point of displaying the popup at start [16:08] DASPRiD: they are? and it is security risk [16:08] just on shared computers [16:08] tho there you have a master password [16:08] all the sites i use pre-fill user name but not password [16:09] well its specifically about htaccess passwords [16:10] DASPRiD: the add-on may not do just htaccess it is likely to do it with everything (maybe a setting to customize it [16:10] )* [16:14] anyone try lubuntu [16:14] i guess ill install it while im at lunch :) [16:40] chrisccoulson: if 3.6.4 is released today or tomorrow, I feel pretty strongly that the update should wait until monday-- unless there is a devastating vulnerability [16:40] jdstrand - i agree [16:40] ok cool [16:41] 3.6.5 wasnt released? [16:41] daily is on pre 3.6.6 [16:41] yeah, 3.6.5 was skipped [16:42] oh [16:42] well have fun im going to lunch [16:42] gnomefreak, "1.9.2.5 was used by Fennec, so next platform version is 1.9.2.6. We will keep the version #'s coherent by naming it Firefox 3.6.6" === al-maisan is now known as almaisan-away [18:54] Anyone else find 3.7a6pre epically faster, or is it just me? === JanC_ is now known as JanC === yofel_ is now known as yofel [20:10] chrisccoulson: hmmm. there are a number of test cases in http://mozilla.qa.ubuntu.com/qatracker/test/4224 that should not be there: seamonkey, thunderbird, sqlite3, and nss [20:11] chrisccoulson: having them there suggests they need to be tested, and I think the sqlite3 one may not be appropriate (though I'm not sure, asac should remember) [20:16] yeah, that shouldn't be there. those must just be packages that were already in the PPA [20:24] jdstrand - the totem plugin doesn't work in the webkit port of epiphany on hardy [20:24] :( [20:25] chrisccoulson: why would it? [20:25] it works with the gecko version, but the webkit version needs a newer version of totem [20:26] chrisccoulson: right, do they use the same plugin or they build a different one? [20:26] it's the same plugin, but the one in hardy is using xpcom and only compatible with gecko browsers [20:27] so we'd need to backport the totem plugin too if we wanted it to work [20:27] ah [20:27] ugh [20:27] chrisccoulson: looking more like it'd be better to fwd port epiphany :)( [20:27] chrisccoulson: how much code is actually in epiphany-gecko? [20:28] i'm not too sure [20:29] k [20:38] chrisccoulson: epiphany has all kinds of problems. I considered it blocked atm [20:38] yeah, it's not too good [20:38] chrisccoulson: as such, maybe focus on that after all the firefox bits? [20:39] i'm a bit confused though, it's not actually using the totem plugin to play media on lucid [20:39] but, whatever it uses isn't working on hardy ;) [20:40] maybe it is smart with mime types? [20:50] is there an equivalent of perl's "<=>" in python? [20:51] ("$a <=> $b" return -1, 0 or 1 if a is lower, equal, greater than b) [20:56] fta: cmp [20:56] fta: but not exatcly [20:56] fta: see http://docs.python.org/library/functions.html [21:01] >>> cmp("2", "11") [21:01] 1 [21:01] >>> cmp(2, 11) [21:01] -1 [21:01] :( [21:05] eek [21:06] chrisccoulson: ctxextensions is pretty hosed. if I install all the extensions from the ppa, then upgrade, most of (all?) the extensions don't show up until I disable ctxextensions [21:07] jdstrand - yeah, i noticed that too [21:07] did you recall if it worked before the upgrade? [21:07] chrisccoulson: I phrased that wrong. if I install all the extensions corresponding to those in the ppa... [21:07] chrisccoulson: it seemed to, yes. I had an Extensions menu to the right of Help [21:08] hmmm, that's not good. i'll look at that later then, but i was thinking if it didn't work before, that we could just leave it at the old version ;) [21:09] heh, yeah [21:20] chrisccoulson: firebug seems broken too. with ctxextensions disabled, firebug still doesn't show up even though firebug and xul-ext-firebug are installed [21:22] Any word on when FF4.0 alpha is coming out, anyone know? [21:31] chrisccoulson: and mozilla-noscript [21:32] there's lots for me to fix then ;) [21:32] chrisccoulson: well, to be fair, there is a lot that works too :) [21:32] heh, that's good :) [21:33] right, time for me to move downstairs [21:33] chrisccoulson: I am updating the qa tracker as I go, so you can look there [21:33] thanks [21:34] I I'll rely on others to test some stuff though (eg bindwood) [21:37] Question: Is 3.7 going to be 3.6.4, or FF4? What exactly is being done with 3.7? [21:39] 3.7 is going to be 4.0 [21:39] Okay, thanks, I wasn't really sure. [21:40] 3.6.4 is a collection of specific backports to allow plugins to run in a different process [21:40] Yeah [21:40] Mozilla needs to make better roadmaps, or just not release any roadmaps at all :P [21:41] Okay, another question: It is June, and I am using the 3.7a6pre... How many more alpha/beta releases will there be, especially if they are aiming for November? [21:44] beta 1 is soon [21:44] Oh, okay, thanks [21:44] and then we'll ship betas on ~2 week cycles [21:44] until we're ready to ship [21:47] Oh, I see. I'm surprised that will take to November or further, 3.0 betas didn't take very long, I think there were only 4 or so [22:06] (ftr: Fx3 b1 was nov 2007, shipped final in june 2008, so seven months) === almaisan-away is now known as al-maisan [22:22] b'ah, my wifi keeps dropping out tonight [22:23] chrisccoulson: openjdk doesn't work on karmic [22:23] I've not tested sun-java6-plugin yet [22:23] is it just not detected or just not working properly? [22:23] sorry [22:23] not detected [22:23] i haven't had a chance to test everything yet [22:23] ah, not detected should be easy to fix :) [22:23] I'm still going through the testing and will report in the qatracker [22:24] thanks [22:24] i'm going to try and get jaunty ready for the weekend, so i've not had much time to test everything yet [22:25] chrisccoulson: sure. hopefully my testing will help save you some time (Firefox3.6.4Upgrade/SecurityPublication for what I am doing) === al-maisan is now known as almaisan-away [22:28] chrisccoulson: also note that my testing of the extensions (the ones I reported on) was fairly cursory. some better than others, but none 'fully' [22:29] chrisccoulson: basically I spot checked several things within each before I gave it a pass [22:43] http://blog.chromium.org/2010/06/bringing-improved-pdf-support-to-google.html [22:50] jdstrand: thunderbird 3.0.5 has been pushed [22:51] chrisccoulson: ^^ [22:51] micahg: you mean by upstream? [22:51] jdstrand: yes, but no CVEs yet, since FF is on hold [22:51] k [22:51] they are trying their hardest to make this exciting for us, aren't they? [22:52] heh [22:52] jdstrand: heh, idk what the procedure is since we reference a USN in the changelog w/out actually publishing the USN [22:53] chrisccoulson: I might be in a pickle w/lighnting [22:53] i suppose there is no urgency to push 3.0.5 if there are no CVE's disclosed [22:53] micahg: can we at least confirm it is security relevant? [22:53] jdstrand: yes, some of the same CVEs as FF3.5.10 [22:54] jdstrand: not all though as JS is disabled by default [22:54] micahg: ok, then I can give you a USN [22:54] jdstrand: no, there's one in the changelog already :) [22:54] micahg: it can sit in the ppa until they disclose them [22:54] * micahg did this last month [22:54] micahg: oh right. ok, then fine. we just won't push until they disclose [22:54] jdstrand: k [22:55] chrisccoulson: so, lightning 1.0b2 won't work w/TB3.1 and 1.0b1 won't work with TB3.0 [22:55] OOPs [22:55] chrisccoulson: switch those [22:55] chrisccoulson: so it seems like I have to package lightning before I can push TB3.1 to maverick or I can't backport [22:57] ah, ok. that's a bit of a pain [22:57] chrisccoulson: so, would it be ok to push off TB 3.1 to beta? [22:59] we could push it off a bit, but i wouldn't want to delay it for too long [22:59] chrisccoulson: I can shoot for the beginning of the beta cycle, daily builds are still scheduled for alpha3 [23:05] jdstrand: would it be a problem to push Thunderbird 3.0.5 before we push to Lucid/Maverick? [23:05] jdstrand: oops, I meant to PPA [23:05] micahg: can you rephrase? [23:06] jdstrand: would it be ok for me to push to the thunderbird-stable PPA? [23:06] jdstrand: before we push to archive? [23:06] micahg: not at all. as long is it complies with upstream policy [23:07] wait [23:07] hehe [23:07] micahg: it is totally ok to push to ppa [23:07] jdstrand: k, it's the team PPA, so I think we're good on upstream policy as well [23:07] micahg: you rephrased but I answered with your original 'would it be a problem' phrasing :) [23:08] chrisccoulson: I'll push TB3.0.5 to the tb-stable PPA, archive can wait :) [23:08] * micahg needs to learn how to run fta's bot :) [23:08] chrisccoulson: fyi, only problem with firefox on karmic is the openjdk one (excepting the extensions we talked about) [23:09] micahg - thanks [23:09] jdstrand, that's good :) [23:09] well, that I found anyway :) [23:09] jdstrand: chrisccoulson: should we try a full backport of openjdk? [23:09] I figured it could be fixed like chrisccoulson fixed it before, but I'll let you guys decide [23:10] i'm not sure if we'll need that yet, i need to have a look at what the problem is [23:10] jdstrand: well, before it was a separate package, I have the commit though for the main xul192 changes [23:10] I see. I defer to chrisccoulson [23:11] a full backport seems pretty heavy-handed if all we need to do is tweak a path or something. but I still defer to chrisccoulson :) [23:12] i'll have a look in a bit. i'm just about to do a hardy -> lucid upgrade, so i can't do anything with karmic just yet [23:12] (my laptop won't handle 2 kvm's ;) ) [23:12] heh [23:12] I typically run 4-5 concurrently for a given security update [23:13] though for this, only 2 (i386 and amd64) for a given release [23:13] chrisccoulson: let me know, the openjdk PPA should have a backported karmic build, I can do that once I finish my hardy stuff [23:13] jdstrand, i need a faster machine ;) [23:14] i didn't realise how much my laptop would struggle when i purchased it [23:14] * micahg wants a laptop w/8GB of RAM :) [23:16] if you have sata, the intel ssd would probably help the most [23:16] * micahg has sata [23:16] (ie, I can test a desktop with 384M vm fine-- so even a couple of gig of ram is ok) [23:16] it is expensive, but man is it nice [23:17] * micahg is hitting .5GB of swap during normal usage w/4GB RAM [23:17] yeah, i was thinking of getting a ssd. the harddrive in my laptop is really the bottleneck [23:17] i'm using a lot of swap in normal usage too, on both of my machines [23:17] wow [23:17] i'm sure the lucid kernel is leaking memory :/ [23:18] I swap very little and I have 4G [23:18] :( [23:18] if i leave my desktop on all weekend, then it's usually unusable by monday (the oom killer has done it's work) [23:18] TB apparently has a leak, there's a bug report I haven't gotten to yet, BTW, I use Xfce on top of it :) [23:19] i can't find anything in userspace that leaks memory on my desktop, but i can leave it on all weekend and it's used up all 2GB of RAM and 3GB of swap with just a single session and no applications open [23:19] and there's nothing that appears to be using all that :/ [23:20] but that happens on my laptop too, i just reboot it more often [23:21] chrisccoulson: could it be X + your video driver? [23:21] i thought so, but both machines have different video drivers [23:21] and restarting Xorg doesn't free up the memory [23:22] chrisccoulson: well, what's using the memory when you come back to it? [23:22] nothing, that's the problem. if i look at the applications that are running, there's nothing that's using an abnormal amount of memory [23:22] but if i look at the output of free, then it shows there is nothing left [23:23] (and only a couple of MB of the used memory is cache) [23:23] and the swap is full [23:23] and eventually the oom killer just kicks in [23:40] micahg: ping [23:41] bdrung__: pong [23:41] micahg: can you show me some more complex work you did? [23:43] bdrung__: https://edge.launchpad.net/ubuntu/+source/thunderbird/3.0+nobinonly-0ubuntu1 https://edge.launchpad.net/ubuntu/+source/seamonkey/2.0.4+nobinonly-0ubuntu1 [23:45] bdrung__: here's a smaller, more complex one: https://edge.launchpad.net/ubuntu/+source/thunderbird-locales/1:3.0-0ubuntu1 [23:46] chrisccoulson: so, the apparmor profile is disabled on the karmic upgrade [23:46] chrisccoulson: ie, if someone enabled it in 3.5, then upgrades to 3.6, then it is disabled [23:46] ah, that's not meant to happen is it? [23:46] chrisccoulson: I'll need to look at this more closely tomorrow [23:46] thanks [23:47] chrisccoulson: it tries to be smart, but it never accounted for the 3.6 transition [23:47] jdstrand: chrisccoulson: there's a version check because of a snafu in the early 3.6 biulds [23:47] * micahg thinks [23:47] yeah. I'll work on it tomorrow [23:47] I'm fairly frazzled atm [23:47] :) [23:48] chrisccoulson: the basic idea is the profile should be disabled on new installs. if the user later enables it, it should stay enabled on upgrades [23:48] yeah, that makes sense [23:49] that very simple statement is surprisingly hard to get right, esp. with the shared debian/ directory... [23:49] micahg: looking at http://launchpadlibrarian.net/39334337/thunderbird-locales_1%3A2.0.0.14%2B1-0ubuntu2_1%3A3.0-0ubuntu1.diff.gz, using PATH_TO_RDF=usr/share/thunderbird-locale-$${CURLOCALESMALL}/install.rdf is not save [23:50] you should call install-xpi with -i to specify the path! [23:53] bdrung__: k, well, I want to install in the appropriate path [23:53] micahg: i changed the default path from /usr/share/$package to /usr/share/xul-ext/$package-without-xul-ext [23:54] bdrung__: ah, was that in 0.23? [23:54] in 0.21 [23:54] "Install extensions into xul-ext subdirectory" [23:55] * micahg thinks that was fixed in the next upload [23:55] bdrung__: yes, I fixed that after you made the change :) [23:56] bdrung__: https://edge.launchpad.net/ubuntu/+source/thunderbird-locales/1:3.0.4-0ubuntu1 [23:56] :) [23:57] bdrung__: yeah, I remember the discussion and we decided to keep the previous path in Lucid since we were so close to release, I might revert that in Maverick [23:58] micahg: one suggestion: have the xpi files extracted in the source tarball. then you can patch the install.rdf files with quilt before building. [23:58] ho! [23:58] bdrung__: you mean, build a *real* source tarball? :) [23:58] micahg: yes ;) [23:58] hi asac, can you comment on my developer app? [23:59] bdrung__: if I have time to rework the package this cycle, I'll try it :) [23:59] micahg: we have a bunch of xpi-* scripts for that [23:59] micahg: when is the meeting? [23:59] bdrung__: ah, ok [23:59] hi asac [23:59] asac: Tuesday [23:59] hey bdrung__ [23:59] micahg: thats enough time to remind me again ;) ... sorry. i will do it next time i am awake! [23:59] asac: k, np