[00:11] hi RAOF - you didn't spend any time looking at the gjs build failure did you? [00:12] chrisccoulson: I did, but then punted it to michag [00:12] heh, it looks like it's been punted to me now then ;) [00:12] i just wanted to make sure i wasn't duplicating effort there [00:13] Right. I'll describe what I know for you then. [00:13] It seems there's probably two bugs; 1) gjs just plain segfaults in mozjs on armel whenever you ask it to do anything. [00:14] the bug i was looking at was the test-suite failure due to a leaked object [00:14] 2) the behavioural ABI for mozjs on i386 seems to have changed when the JIT is enabled - it seems that at least one finaliser isn't called when you destroy the context. [00:14] Right. That's the second one. [00:14] That can be worked-around by just disabling the JIT on i386; the armel one needs something more. [00:14] ok, i'll spend some time trying to figure that out [00:15] We could also just disable the test-suite; I'm not sure how critical the leak is for i386 [00:15] i'd rather not disable JIT on i386 if i can avoid it, as it could be a real problem [00:15] yeah, i'm not surehow big a problem it is really :-/ [00:16] Armel isn't a simple matter of disabling jit, sadly :/ [00:16] (Otherwise I would have uploaded packages that just disable the jit on armel & i386) [00:19] looks like it's built alright [00:20] gah... compiz + cursors [00:21] * dutchie pushes to LP [00:25] kenvandine, back around yet? [00:33] compiz will ONLY let me use compiz mega-large black cursor === almaisan-zzz is now known as almaisan === almaisan is now known as almaisan-zzz === BalSak1 is now known as multi-x [00:43] hi guys. in the upcoming 10.04 lynx, any idea if the gdm will upport multiseat-X, and if not, if it will be possible under KDE? [00:43] * RAOF is unsure what you mean by “multiseat X” [00:44] RAOF: using a single machine with multiple seats via multiple video cards, USB keyboards & mice: https://help.ubuntu.com/community/MultiseatX [00:45] it's (apparrently) been a standard feature of X for a very long time, up to gdm 2.20, but got bloken under the recent 2.28 rewrite [00:46] Ah. That version of multiseat. [00:46] I don't know; try it with a livecd? [00:47] been doing some work on it recently for implementation in schools, offices & home (with kids). especially now that desktop virtualization is becoming a reality, I can remote-X/NZ/VNC/RDP to a destop in a VM [00:48] RAOF: ???? don't understand what you mean by liveCD. it's an extremely complex setup, initially, so I'll loose all settings between reboots [00:48] well, let me then rather ask: what version of gdm is rolling out with 10.04? [00:49] 2.30 [00:49] where are pointer themes stored? [00:49] RAOF: thanks [00:49] You could do it with a liveusb, where settings can be stored. [00:50] I think you'll also be fighting ConsoleKit, unless you've already worked that out. [00:55] RAOF: yip. it may actually be an asset in this case, though [00:59] we don't have multiseat support in consolekit in ubuntu [01:00] there is a lot of work happening in separate branches of GDM and consolekit to support true multi-seat [01:01] chrisccoulson: thanks for that. could you please elaborate? [01:02] well, the multi-seat branch of consolekit is here: http://cgit.freedesktop.org/ConsoleKit/commit/?h=multi-seat [01:02] and i think the gdm work is here too: http://git.gnome.org/browse/gdm/tree/daemon?h=multi-stack [01:02] oops, sorry, http://git.gnome.org/browse/gdm/commit/?h=multi-stack [01:04] chrisccoulson: what does that exactly mean? my (poor) understanding of konsolekit is a way of abstracting the user from the system, similar to what was done with PulseAudio, so that the user can use services on a server without actually being present. (client/server). this may be an oversimplfication, but seems compatible with milti-seat [01:05] consolekit is for primarily for tracking sessions [01:05] but currently, it only supports one local seat (which all local users are added to) [01:06] but with multi-seat, you will be able to define multiple seats and assign hardware to those [01:06] so, future/beta implementations of consolekit would be very well-suited to multi-seat, and ease the setup & config thereof? [01:06] that is my understanding of how multi-seat will work anyway ;) [01:06] yes [01:07] sweeeeeet! [01:07] it was pretty difficult to get it working in the first place. any help is appreciated.... [01:15] can someone review https://code.edge.launchpad.net/~jshholland/ubuntu/lucid/poppler/backport-anti-alias/+merge/22511 please? [01:17] dutchie, wasn't seb128 reviewing that? [01:18] he seems to have gone to bed [01:18] yeah, you would be better off asking him to have a look at it tomorrow [01:19] shall I subscribe him to the merge proposal? [01:20] you might be better off just pinging him tomorrow. i'm not sure we use the lp:ubuntu branch for the poppler packaging [01:26] thanks guys. will keep an eye out for gdm+consolekit+multiseat [02:29] kenvandine: when you get back give me a ping [04:09] kwwii: any idea about the status of the OOo splash change? [04:18] rickspencer3: ping, you happen to be around? [04:21] ccheney, 'sup? [04:24] RAOF, thanks for fixing the * bug in f-spot [04:24] :) [04:25] I love that turn around! [04:25] rickspencer3: i haven't heard anything back from the art team about the new splash, but we already got the logo a while back and i had sent it to ken, any idea what i should do? [04:26] ccheney, what specifically is blocking you? [04:26] and does the logo contain the new wording? [04:26] the logo's are just the oracle logo itself it has to be combined with the artwork from ubuntu art team to produce the final product [04:27] so right now all i have is the oracle logo and haven't heard anything back from the art team yet [04:27] ccheney, was there not some legalize that need to be refined as well? [04:27] * ccheney double checks [04:28] ccheney, in any case, can you please send an email to kwii reminding him that you need this? [04:28] and cc ivanka and I? [04:28] there are a few minor changes to the copyright file and update in the wiki to do, but the main blocker is the artwork [04:28] ok will do [04:28] ccheney, I'll ask about the wording right now [04:29] once we get the artwork from the art team we have to show it to oracle to get approval but that shouldn't take too long, might miss apr 1 though (2 days from now) [04:34] rickspencer3: ok sent [04:34] thanks ccheney [04:34] I also asked Amanda about the wording [04:35] I'll keep you posted [04:35] rickspencer3: ok thanks [04:35] wow, I installed UNE fresh from the daily today, and now have 109M of updates! [04:36] rickspencer3: if we get the images and oracle/amanda approval by sunday it should be ok assuming its processed during the freeze [04:36] ccheney, it depends where the hold up is [04:36] it takes about 2 days to build on arm, but the arm buildds have been apparently having issues and couldn't build the last upload :-\ [04:36] yea [04:37] If they got us the stuff by April 1, I feel duty bound to do as promised [04:37] the last upload arm couldn't build wasn't materially different from the previous one that did work afaict [04:38] ccheney, is a rebuild actually required to replace artwork? [04:38] that seems rather, uh, old fashioned [04:40] yea, well its part of the build process for the filename to use which currently has sun in it and also the fact that the artwork itself is in the openoffice.org source instead of in a separate theme package, etc [04:41] also they are planning on changing up the color and location of the progress bar which requires a code patch [04:41] its a very minor patch but is more than just replacing the graphic for that part [04:51] thank goodness I'm getting the new aislerot [05:00] RAOF, I see you also liked bug #536925 yesterday ! [05:00] Launchpad bug 536925 in gnome-rdp "gnome-keyring-sharp uses deprecated socket interface; apps cannot use keyring" [Undecided,Fix committed] https://launchpad.net/bugs/536925 [05:18] bryceh, are you around? [05:23] rickspencer3: guessing you have a problem after the updates? tjaalton just uploaded the evdev and synaptics packages to fix it if so [05:24] Sarvatt, a slight problem, yes [05:24] the touchpad is sensitive in the button area on my dell mini 10v again [05:24] it is quite enraging [05:25] Sarvatt, is this the regression you are mentioning? I just did a fresh install of UNE daily today, so wasn't an update [05:25] but "working", so it's not the same issue :) [05:25] tjaalton, yes [05:25] if by "working" you mean make me want to throttle the person who created such a diabolic device [05:25] ;) [05:25] but in fact, it does point the mouse and such [05:26] I'm afraid that people upgrading to the new xserver will be disappointed until -evdev/synaptics are released [05:26] but I need someone to shove those past the build queue [05:26] anyway, I assigned a bug to tseltiot [05:26] tjaalton, oh? will there be breakages on touchpads tomorrow? [05:26] rickspencer3: http://sarvatt.com/downloads/xorg.conf.d/ has the steps to fix it if you need the fix ASAP [05:27] rickspencer3: well that's what Sarvatt suggests, though we took measures to get around that.. [05:27] I'm confused (which is normal) [05:27] the new xserver with xorg.conf.d/inputclass backports [05:28] I'm thought tseliot patched the synaptics driver for this sensitiviy in the button area usability nightmare [05:28] like a 6 months ago [05:28] Sarvatt, but thanks for the xorg.conf files [05:28] I'll use those if it comes down to it [05:28] should just work with the input drivers exporting 'x11_driver' from the udev rules (included a patch for that in the xserver), but apparently doesn't [05:29] JumpyCursorThreshold - lolz [05:29] tjaalton: ah actually I believe it may be working for him but the quirks for his model aren't [05:30] Sarvatt: but I don't think he's got the new xserver yet [05:31] thanks gents [05:31] gots to run [05:31] ttyt [05:33] rickspencer3: can you pastebin your /var/log/Xorg.0.log? [05:34] rickspencer3: ah nevermind, you need both the udev rule and the xorg.conf.d snippet for the quirk to work which you dont have yet, hopefully it'll be built soon [05:38] rickspencer3: I can reproduce the “Save as” freeze, too, but it looks like the problem's in GTK# rather than f-spot. Freezing is a pretty obnoxious behaviour for a save dialog! [06:08] RAOF, can you seed the dialog with a name to at least mitigate the bug a tiny bit? [06:09] rickspencer3: I *do* seed the dialog with a name - are you additionally not seeing a default filename? [06:09] RAOF, correct [06:11] RAOF, http://imgur.com/LFvxv [06:13] Sarvatt, fwiw, here's bug #552317 [06:13] Launchpad bug 552317 in xserver-xorg-input-synaptics "touchpad on dell mini10v pretty much unusable" [High,New] https://launchpad.net/bugs/552317 [06:14] you can reassign, comment, etc... as you like [06:23] rickspencer3: thanks, it's definitely just the quirk not getting applied and the xserver-xorg-input-synaptics with the fix just finished building - https://edge.launchpad.net/ubuntu/+source/xserver-xorg-input-synaptics/1.2.2-1ubuntu1/+build/1593577 [06:24] thanks Sarvatt, looking forward to getting that tomorrow [06:26] you can just run this in a terminal to fix it until the upgrade is there by the way - synclient JumpyCursorThreshold=90 AreaBottomEdge=4100 [06:30] thanks Sarvatt [06:30] * rickspencer3 copies and pastes [06:56] Oh, dear. That's actually a GTK+ bug. [06:56] That *will* hit people. [06:56] Over the head, with swords. [07:07] Good morning [07:08] pitti: Good morning [07:28] hey RAOF, how are you? [07:29] Good. Yourself? [07:30] I've been having fun finding bugs that go all the way down the stack :) [07:43] RAOF: yeah, just saw the save as dialog bug.. [07:43] weird that this hasn't been noticed much earlier [07:45] Yeah. I would have thought that *someone* would complain loudly that they accidentally hit “save” without a filename and it froze the program without saving. [07:50] RAOF: reproduces perfectly in gedit [07:52] And in the trivial test program. [07:52] (I trust :)) [08:13] good morning [08:26] good morning didrocks [08:26] * kenvandine goes to get some sleep, good night :) [08:26] have a good night kenvandine! [08:28] night kenvandine [08:35] hey aquarius [08:36] * kenvandine really goes to crash now [08:37] hey kenvandine [08:40] hello there [08:40] bonjour seb128 [08:40] hello pitti [08:41] how are you? [08:42] I'm pretty good, thanks; you? [08:42] trying to finish off the db4.7 deprecation [08:44] seb128: btw, gnome-panel amd64 binaries still aren't published, which caused some FTBFS; I asked #soyuz, and they sorted it out (they think) [08:44] pitti, I'm great thanks [08:44] so we need to wait for the publisher to finish [08:44] ok [08:44] then we can give back stuff [08:44] didrocks: ^ FYI (if you saw amd64 FTBFS) [08:44] I tried to give some builds back yesterday [08:44] armel is wracked [08:44] hey seb128, pitti [08:44] pitti: thanks for the notice :) [08:44] salut didrocks [08:45] seb128: hm, ok, they are still investigating [08:45] didrocks, did you start on any update today? [08:45] seb128: not yet, still finishing triaging email, I'll then (almost finished) [08:45] k [08:45] I'm doing gnome-desktop and libgnome-keyring now [08:47] ok [08:48] didrocks, bug #552132 [08:48] Launchpad bug 552132 in gnome-control-center "gnome-appearance-properties crashed with SIGSEGV in g_cclosure_marshal_VOID__VOID()" [Undecided,Confirmed] https://launchpad.net/bugs/552132 [08:49] could you investigate? [08:49] it started recently [08:49] not sure if that's the update or your changes [08:49] Ola ! [08:49] but since you are not one who touch it ;-) [08:49] salut baptistemm [08:49] seb128: yeah, that's on what I am in fact :) [08:50] thanks [08:52] are the retracers down ? === almaisan-zzz is now known as al-maisan [08:53] seb128: gnome-panel should be good now [08:53] * pitti gives back gdm [08:54] baptistemm, they keep going up and down why? [08:56] pitti, I'm looking to gnome-menus caching issues [08:56] pitti, not sure why you filtered non displayed items out [08:56] seb128: well, that was the point of it? [08:56] pitti, it's what create bug #542300 [08:56] Launchpad bug 542300 in gnome-menus "Missing Menu Entries - Lucid Beta 1" [Low,Confirmed] https://launchpad.net/bugs/542300 [08:57] gnome-menus filters them out anyway? [08:57] pitti, those are usually listed as hidden in editor [08:57] and you can check those [08:57] ie enabling by clicking [08:57] and without the pre-filtering, the cache would get twice as big, and become useless [08:57] but with your cache not seeing those at all you can't do that now [08:58] well NoDisplay means not displayed by default [08:58] the editors should still list those are not displayed [08:59] pitti, quite some users enable the gnome-control-center item in the system menu this way for example [08:59] seb128: the bug description of that is very weird, though; the caching certainly doesn't cause _all_ of those to go away; most people have the capplets/system tools, etc? [08:59] well as you commented [08:59] it filter out all the NoDisplay=true [08:59] they are not in the cache [08:59] so they don't exist for the editors [09:00] but they should [09:00] lot of users tweak their menus and active things we don't display by default [09:00] gnome-control-center [09:00] ok, so some of them should appear (like "control center"), and some others shouldn't (like apport, gdebi, etc.) [09:00] eog [09:00] gconf-editor [09:01] I don't think you should change the spec semantic with your cache [09:01] can we make a difference between "do not ever show this" (gdebi/apport which are mime links) and "we just hide it by default"? [09:01] things with no categories would not be listed [09:01] well, I don't understand why you want to change the semantic [09:02] the cache should do what we had before but in a faster way [09:02] ah, ok; so NoDisplay=True should be in the cache, but not empty categories and non-matching OnlyShowIn, right? [09:02] I'm not sure about that and the sideeffect that can have [09:02] I think we should just list all entries installed [09:02] the way it works without cache [09:02] oh, we shouldn't change the semantic _in_ the caching patch [09:02] to avoid breaking softwares relying on that behaviour [09:02] I was just pointing out that there are desktop files which should not EVER be shown [09:03] right [09:03] but gnome-menus take care of that [09:03] I get tons of bugs "I enabled the apport.desktop and clicking on it doesn't work" [09:03] I regularly close them with "don't do that" [09:03] I don't think we should duplicate that work by tweaking what goes in the cache or not [09:03] yes, as I said, it's not the cache patch which should enforce that [09:03] well then fix gnome-menus or the desktop entry you use [09:03] it does right now though [09:03] it's all the [09:04] but if we can mark those differently, then the cache doesn't need to even contain the ones which aren't ever displayed [09:04] and my question was if there's a flag to mark them differently [09:04] if de.getNoDisplay() or de.getHidden(): [09:04] return [09:04] e. g. /usr/share/applications/apport-gtk-mime.desktop currently uses NoDisplay=true [09:05] which apparently isn't strong enough [09:05] right [09:05] but it should be the cache enforcing that policy [09:05] (should "not")? [09:05] it means gnome-menus with and without cache doesn't behave the same [09:05] yes sorry [09:05] yes, we agree [09:05] should *not* [09:05] ok [09:05] can I drop those filters in the cache code then for lucid? [09:06] http://lists.freedesktop.org/archives/xdg/2009-May/010389.html [09:06] seb128: yes, I guess we have to [09:06] seb128: well, I don't know about "Hidden" [09:06] but certainly getNoDisplay() [09:06] I would drop everything [09:07] reading the cache should be the same than reading the directory content [09:07] then we can fix issues like the apport one as they should [09:07] ie by updating the desktop or fixing gnome-menus code [09:07] ok, seems that the spec itself doesn't really support that :( [09:07] no it doesn't [09:07] we might be able to hack around by having no categories I think [09:08] see gnome-theme-installer.desktop [09:08] seb128: we don't use "Hidden" in /usr/share/applications/ anyway, so you can as well drop the check [09:08] right [09:08] Hidden means hidden from menus and mimetypes too [09:09] ie that's equivalent of not installed [09:09] there is no point to do that in system entries [09:09] that's good to delete some from the user config though [09:09] pitti, ok, doing that, thanks [09:09] pitti, but see gnome-theme-installer.desktop for your apport bug [09:11] ah, changing it to "Categories=GNOME;GTK;" works indeed [09:11] i. e. it doesn't appear any more in alacarte [09:11] seb128: thanks! [09:11] * pitti commits [09:11] pitti, you're welcome [09:11] pitti, thanks for taking the time to clarify that with me there ;-) [09:17] didrocks, yet another bug eating the dust thanks to you ;-) [09:17] * seb128 hugs didrocks [09:17] thanks for fixing that crasher [09:18] * didrocks hugs seb128, was my fault :) [09:19] * didrocks gives back glade, gnome-menus, gnome-utils on amd64 and back on update [09:27] nice, libdb4.7 is 653 kB; which will now drop from the CDs [09:28] 20% of a langpack :) [09:28] how much do we need to get an extra one back? [09:28] ie are those the first 20%? [09:28] haven't checked yet [09:28] or do we have like 80% space now [09:28] CD builds were broken in the last couple of days [09:29] I guess the GNOME 2.30 uploads don't help there [09:29] nope [09:29] pitti: If you are working seeds atm, could you cut a langpack or two from powerpc? The desktop CD is oversized. [09:29] seb128: taking seahorse and seahorse-plugins [09:30] Otherwise I can take care of it myself. [09:30] didrocks, thanks [09:30] TheMuso: I'm not, it's all your's [09:30] TheMuso: I just did some libdb4.7 -> 4.8 migration [09:30] oh ok [09:30] thanks [09:32] seb128: investigating apport retracer crashes now [09:32] pitti, thanks [09:32] seb128: the amd64 one stumbled over a broken bug report again (https://bugs.edge.launchpad.net/ubuntu/+source/nspluginwrapper/+bug/551044) [09:32] Error: Could not parse data returned by Ubuntu: list index out of range (https://launchpad.net/bugs/551044) [09:32] -> oopses [09:32] pitti, do you have an opinion on when to turn apport off? [09:32] same for the lpi "report a bug" [09:33] seb128: apport> right after the release candidate, traditionally [09:33] lpi> can we do a test drive (package in ubuntu-desktop with that disabled)? [09:33] and then disable it aronud that time as well? [09:34] pitti, well it's easy, it's one line to comment in lpi [09:34] ok [09:34] seb128: what do you think? you want to do it later? [09:34] erm, eralier? [09:35] no, that seems fine [09:35] seb128: right, but we did the apport disabling a lot of times now [09:35] seb128: I'm slightly worried that disabling the menu entry has some side effects [09:35] you want me to upload that before beta2? [09:35] like apps trying to change the menus after calling lpi, etc. [09:36] seb128: to the PPA? not that time critical, but I guess it would be good to do it soon [09:36] I've no concern about not having bugs for annoying issues [09:36] we have enough bugs sent our way menu item or not [09:36] ok [09:36] I will upload to the ubuntu-desktop ppa today [09:36] * pitti hugs seb128, merci! [09:37] * seb128 hugs pitti [09:37] de rien! === tkamppeter_ is now known as tkamppeter [09:49] hrm only en is on powerpc desktop CD. Have to cut something else, and I think part of openoffice is what it may have to be. [09:51] TheMuso: any idea why powerpc is so big? [09:51] does it have that much extra packages? [09:51] s/much/many/ [09:51] pitti: No it doesn't, but it doesn't help that one of the kernels is huge, the kernel in questino being powerpc64-smp. [09:51] question [10:03] seb128, this is just someone opened a bug yesterday, and the retrace is not yet done [10:03] baptistemm, ok, it might take a while [10:03] and it seems to be a reproductible crasher I'm interested to have [10:04] do you have the bug number? [10:07] yep [10:08] bug 551955 [10:08] Launchpad bug 551955 in gnome-bluetooth "bluetooth-applet assert failure: *** glibc detected *** bluetooth-applet: double free or corruption (fasttop): 0x08870470 ***" [Undecided,New] https://launchpad.net/bugs/551955 [10:10] retracing will not be useful [10:10] double free requires a valgrind log [10:11] hrm livefs is not the latest anyway, so will wait for tomorrow to see where things stand. [10:11] seb128, okay [10:13] re, I'm striking with any no more keyboard on gdm and gnome session after upgrade + reboot. On my right TTY now [10:13] hum [10:13] ot nice [10:13] not [10:14] no, it's not :( I'm looking at the bunch of update I had this morning to see what's in fault [10:14] gdm is already out of this :) [10:14] I would tend to blame tjaalton [10:15] what now? :) [10:15] there was quite some xorg updates [10:15] tjaalton, some update from today broke xorg keyboard for didrocks [10:15] tjaalton, I would think it's due to some of your uploads ;-) [10:15] didrocks: pastebin Xorg.0.log [10:15] but I'm random blaming there [10:16] seb128: Sarvatt saw the same, but I couldn't reproduce it [10:16] so feel free to ignore me :p [10:16] things should be fine with the new evdev [10:16] * seb128 doesn't update [10:16] tjaalton: I'm trying to downgrade xserver-xorg-input-evdev, reboot in progress [10:16] didrocks: which arch? [10:16] tjaalton: i386 [10:17] so you got a new evdev? then things should definitely work [10:17] ok, it's back [10:17] kbd/mouse? [10:18] seb128: I note you updated tomboy to 1.2.0. Did you take into account the new files that are deployed in tomboy 1.2.0 that aren't in previous releases? [10:18] one sec, finishing booting :) [10:18] popey, which? [10:18] seb128: http://paste.ubuntu.com/406951/ [10:19] MacSlow: ping [10:19] poningru, do you have any issue? [10:19] ups [10:19] popey, ^ [10:19] popey, going backward is usually not a good thing [10:19] popey, try rather to tell us what doesn't work than trying to guess what is broken [10:19] tjaalton: so, the guilty version for me was 1:2.3.2-5ubuntu1, installing back 1:2.3.2-3ubuntu2 works [10:19] popey, and yes the new files are installed [10:20] didrocks: you need to install xserver-xorg-core [10:20] seb128: ok. i'm not saying something doesn't work, i just couldn't see the difference in the debian/rules, sorry to have bothered you [10:20] tjaalton: ok, trying upgrading ok and installing it [10:20] tjaalton, you should use breaks in those cases [10:20] didrocks: no guarantees about partial upgrades :) [10:20] tjaalton, or versioned depends [10:20] too ugly [10:20] tjaalton, wrong reply [10:20] seb128: did you see https://bugs.edge.launchpad.net/ubuntu/+source/poppler/+bug/248355/comments/46? [10:20] tjaalton, ? it has been made for those cases [10:20] Ubuntu bug 248355 in poppler "Evince doesn't anti-alias graphics" [Low,Fix committed] [10:21] dutchie, yes, did you see the comment from james_w? [10:21] that's the comment [10:21] seb128: well, too late anyway [10:21] I don't see the what Makefile.in problem is [10:21] tseliot: xserver-xorg-core is already installed here [10:21] the xserver was uploaded first just for this reason [10:21] didrocks: which version [10:22] ? [10:22] tjaalton: 2:1.7.99.2~git20091215.77221c91-0ubuntu0tormod [10:22] uh oh [10:22] tseliot: sorry, wrong hl (dummy weechat ;)) [10:22] aah, ok [10:22] yeah if you use edgers you're busted [10:22]