[00:00] hggdh: it boggles the mind that runit has been upstartified; looking [00:00] upstart runs runit runs daemontools runs home and cries to djb ;) [00:01] hggdh: this package is calling stop/start directly - that's a bug [00:01] hggdh: the only interface policy says packages are supposed to use in their maintainer scripts is invoke-rc.d [00:01] slangasek, ah, I feel better now :-) [00:02] slangasek: I'm going to send the idea to upstart-devel ... I think this may be the right thing to ship and recommend people use until we have 'while' [00:03] SpamapS: ok, cool [00:04] hggdh: so yeah, it's a bug in runit, which is definitely not a package I'm spending any effort fixing right now :) [00:08] slangasek, NP. I understand all that is needed is change the postinst to call invoke-rc.d instead of /sbin/start. If this is it, I can do it mesself. Correct? [00:09] hggdh: well, the *sensible* thing to do is to make the package use debhelper, which automatically spits out the correct maintainer script snippets... but yeah, I don't think you want to maintain that :) so just replacing with invoke-rc.d is the way to go [00:09] hggdh: the real question is why isn't it using dh_installinit [00:09] err.. what he said [00:09] SpamapS: I don't think that's a question, really [00:09] I mean, it's runit [00:09] did you expect the packaging to not also be idiosyncratic? ;) [00:11] SpamapS, slangasek: adding debhelper will be a bigger change ;-) [00:12] * hggdh would rather start with small changes [00:14] hggdh: yeah, I generally wouldn't be in favor of debhelperifying a package in Ubuntu only unless it's something we plan to maintain [00:15] deal === tjaalton_ is now known as tjaalton === nijaba_afk is now known as nijaba [01:45] geser, tumbleweed, etc. Feel free to fix. I need to run. === asac_ is now known as asac [04:21] Does anyone know if Python3.1 will remain in Natty, or will it be dropped, leaving just Python3.2? [04:45] jderose, I can't speak authoritatively, but I doubt anyone is going to try really hard to drop python3.1. [04:45] I suspect the default python3 interpreter will be python3.2 [04:45] Python2.6 will likely also be available in parallel. [04:46] persia: yeah, 3.2 already seems to be default. i'm working on a python3 package and was just curious if i should fully QA for running under 3.1 also, or just 3.2 [04:47] persia: thanks for the insight! [04:48] I'd recommend focusing on 3.2: it will be a rare case that someone would be using 3.1 unless they asked for it especially. [04:48] And in the longer term, 3.1 will go away: I'm just not sure that the actual final removal is on someone's natty agenda. [04:48] persia: okay, thanks :) [04:49] Any archive admins available to push through the -vmmouse sync in bug #709134? It's the last of the pieces required to transition to Xserver 1.10. [04:49] Launchpad bug 709134 in xserver-xorg-input-vmmouse (Ubuntu) "Sync xserver-xorg-input-vmmouse 1:12.6.99.901-1 (main) from Debian experimental (main)" [Wishlist,New] https://launchpad.net/bugs/709134 === jtechidna is now known as JontheEchidna [06:40] Good morning [06:41] pitti: ahoi! :) [06:45] pitti: Good morning! [06:48] pitti: Are you archive admininging today? The sync in bug #709134 will finish the Xserver 1.10 transition. [06:48] Launchpad bug 709134 in xserver-xorg-input-vmmouse (Ubuntu) "Sync xserver-xorg-input-vmmouse 1:12.6.99.901-1 (main) from Debian experimental (main)" [Wishlist,New] https://launchpad.net/bugs/709134 [06:49] RAOF: I'm not, but I can do the sync [06:49] RAOF: is that the reason why dist-upgrade currently croaks? [06:49] The following packages have unmet dependencies: [06:49] xserver-xorg-core : Breaks: xserver-xorg-video-8 [06:49] hm, hardly, that's -input [06:50] oops, seems the new pygobject breaks a lot [06:51] It will be breaking dist-upgrade, once the rest of the uploads are built and published. [06:51] will fix [06:51] synced [07:04] Hm. That's ususually long. -ati finished building 3 hours ago, doesn't seem to have made it to archive.ubuntu.com yet. [07:09] O [07:10] I'm wondering, if a few packages build-dep on a package only on archs that we don't have, can the build-dep be dropped from Ubuntu (gcc-4.3) [07:11] micahg: yes; we should drop it if it reduces delta from Debian, and keep it if that b-dep is in Debian (IMHO) [07:12] pitti: the build-dep is gcc-4.3 and it's only on archs we don't have [07:12] micahg: right, I understand === tkamppeter_ is now known as tkamppeter [07:13] micahg: as it's basically a no-op for us, we should keep it if the Debian package has it as well (to avoid an unnecessary delta), and drop it if not [07:13] pitti: I'm wondering about dropping gcc-4.3 :) [07:13] ah [07:14] sorry, misunderstood [07:14] * micahg probably isn't too clear at this hour :) [07:14] there's still a couple of rdepends, though [07:15] I'm looking now for ones that aren't no-ops [07:15] e. g. chromium-browser [07:15] Build-Depends: g++-4.3 | g++-4.2 [07:16] pitti: is there a germinate output for it? [07:18] * micahg is wondering if there's an easy way to check for remaining deps [07:18] micahg: http://paste.ubuntu.com/560844/ [07:19] pitti: is that a special tool? [07:19] micahg: a lot of that is self-referential and there might also be alternative dependencies for those [07:19] micahg: it's checkrdepends from lp:ubuntu-archive-tools [07:19] ah [07:19] checkrdepends gcc-4.3 natty [07:19] * micahg needs to remember that [07:21] pitti: thanks, will look into this [07:57] good morning [07:58] hey dholbach [07:59] hey mvo [08:03] <\sh> moins [08:11] good morning [08:50] Who wants to give a session at UDW? Please sign up at https://wiki.ubuntu.com/UbuntuDeveloperWeek/Timetable - thanks :) === Claudinux_ is now known as Claudinux [08:56] didrocks, anyone reporting 'focus' issues with yesterdays unity... i am seeing the wrong cursor and lack of sensitivity on some elements ... actiling like one of the semi-transparent places boxes is on the screen; though there is nothing visible [08:57] apw: hey, there are some bug report in compiz, all new info like xwininfo + click on the "invisible" window can be great [08:58] apw: bug #709461 [08:58] Launchpad bug 709461 in compiz (Ubuntu Natty) "semi-random invisible window with x geometry on top layer possible, all viewport only (one ws though)" [High,Triaged] https://launchpad.net/bugs/709461 [09:00] didrocks, will see what i can gleen [09:12] pitti: 'checkrdepends gcc-4.3 natty' isn't the right invocation any more :-) probably works by accident ... [09:21] bryceh,RAOF,Sarvatt: is there going to be an upload of -video-mach64 for the new X? [09:22] bryceh,RAOF,Sarvatt: actually, there seem to be several video drivers missing, per http://people.canonical.com/~ubuntu-archive/testing/natty_probs.html [09:24] cjwatson: Meep, that's a big list of armel. [09:25] I'm guessing that's just due to no (input|video) drivers being installable yet, so xserver-xorg's uninstallable, so xserver-xorg-core's uninstallable [09:26] Yeah, looks like it. [09:26] *All* the drivers built against the wrong server?! Bah. === cjwatson changed the topic of #ubuntu-devel to: Archive: Soft freeze for Alpha 2 | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: [09:30] Ok. I'll bump the X server dependencies and get those rebulit. [09:32] OK, thanks [09:32] RAOF: does that go for the ones that are still uninstallable on x86 too? [09:33] Yeah. [09:33] I see the rebuild-the-world script is going to need to be a bit smarter next time :/ [09:33] OK, I'll leave that with you then, thanks. Do you need upload assistance? [09:33] I will, yes please. [09:34] Point me at something moderately digestible then ... [09:34] I'll get everything batched up, then ping you once? [09:34] yep [09:34] Is *_source.changes suitably digestible, or would you like debdiffs as well? [09:36] just source packages should be fine === hunger_ is now known as hunger [10:13] cjwatson: oh? [10:31] pitti: 'checkrdepends -s natty gcc-4.3' or just 'checkrdepends gcc-4.3' (plus -b if you want binary) [10:32] pitti: I sent a mail about this at the time, I think - I had to change the syntax to make the interface sane for looking up multiple packages at once [10:32] cjwatson: ah, so I suppose it additionally checked for rdepends of the "natty" pacakge [10:32] yeah [10:32] cjwatson: right, forgot about that apparently; thanks! [10:46] The following packages have unmet dependencies: [10:46] unity : Depends: unity-common (= 3.2.16-0ubuntu1) but it is not going to be installed [10:46] E: Broken packages [10:46] GRRR ! [10:47] ogra: on armel? [10:47] didrocks, yeah, i suspect only arch any vs arch all out of syncness [10:47] ogra: right [10:47] ogra: unity FTBFS yesterday on armel IIRC [10:47] but it gets annoying, we dont have images for weeks due to that skew [10:47] let me recheck [10:48] the FTBFS was due to something not being able to install [10:48] I retried it earlier this morning [10:48] cjwatson, you had some ideas how to solve that properly, i want to have a spec/BOF at next UDS about that [10:48] I asked for a rebuild a little bit later without any success [10:48] yeah, it's built now, thanks cjwatson [10:48] it has gotten really bad this release [10:49] ogra: the theory isn't difficult, but it requires soyuz implementation time [10:49] should i mention here that canonical's gcc madness causes mono to fail more than three times as many test cases on natty as on sid, on the same arm hardware? [10:49] it doesn't need a bof, it needs to get done [10:49] cjwatson, right, i'll take care that we have soyuz people around for that [10:49] directhex: given that our gcc maintainer is on holiday, I'm not sure what you would be achieving [10:49] i think with a new non x86 arch it will get bad for others too [10:49] ogra: around> it still doesn't need a bof [10:50] cjwatson, well, then a spec with the right people assigned [10:50] the solution is to keep multiple versions of packages in the Packages and Sources files, with appropriate refcounting [10:51] ok [10:52] so you have as many versions of architecture: all packages in the db as are required to match published builds of each architecture [10:53] well, wouldnt it be easier to just block the new ones from showing up until the set is complete ? [10:54] i know there will be a slight out of sync-ness across the arches due to that, but thats still better than completely uninstallable sets and seems easier to implement [10:54] at least in my imagination (not knowing the code at all) [10:55] that would cause different problems [10:55] ah, k [10:56] my solution is not theoretical [10:56] http://ftp-master.debian.org/wiki/projects/arch-all-tracking/ [10:56] (nor is it "my solution", come to that, but anyway) [10:57] geez 2009 ! === diwic is now known as diwic_otp [10:57] i'll talk to davidm that he organizes resources for implementing that in O [10:58] also http://twerner.blogspot.com/2009/11/dak-dominate-will-dodadoda-debian.html [10:58] we have literally lost weeks this cycle due to it [10:59] so this is implemented in practice and the various client-side things that had problems have generally already been fixed now; there is certainly no need to design something afresh, and it would probably be actively undesirable to do so [11:00] well, i think linaro designed something different from scratch [11:00] linaro probably didn't know about this [11:00] they us an archive snapshot afaik [11:00] this should be purely a catchup project [11:00] *use [11:00] but that adds the difficulty to know when the achive is in sync at all [11:02] right, i'll do some writeup for david and hand it to him to make sure we have resources assigned for an implementation === diwic_otp is now known as diwic === wgrant_ is now known as wgrant === MacSlow is now known as MacSlow|lunch [11:33] @pilot in === udevbot changed the topic of #ubuntu-devel to: Archive: Soft freeze for Alpha 2 | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: pitti [11:36] ogra: Does your crazy AC100 still need -input-kbd? I'll include that in the rebuild frenzy. [11:38] RAOF, yep, it does, thanks ! [11:44] * dholbach hugs pitti [11:44] * pitti hugs dholback === oubiwann is now known as oubiwann_ [12:17] I just installed Wireshark in lucid, but for certain functionality it needs to be started as root. In previous versions it installed 2 menu option (1 to start with root privs.). Is this a bug? Anyone know if I should report this issue? [12:21] The issues is I manually had to make a second menu item for starting with root privs. [12:23] RAOF: how's it going? [12:23] pitti: Thanks for merging that evolution-mapi patch :) [12:23] cjwatson: I'm just running a test-build of everything. [12:24] jelmer: at your service :) [12:24] cjwatson: If you'd prefer, tjaalton has offered to sponsor the packages, too. :) [12:25] don't mind [12:25] I just want to get going on alpha-2 builds ASAP [12:26] * didrocks sneaks two small fixes for unity places meanwhile :) [12:27] cjwatson: Yeah, sorry. [12:32] pitti: free? [12:32] ari-tczew: what's up? [12:33] pitti: could you look on gedit merge? === oubiwann_ is now known as oubiwann [12:33] ari-tczew: you know main is (soft-)frozen? [12:34] ari-tczew: will look, if it's safe for a2 [12:34] cjwatson: oh! nope. but Robert Ancell has sponsored bluez for me 11 hours ago. [12:34] ari-tczew: that was before the soft freeze started. [12:34] pitti: translations update [12:34] ari-tczew: looks safe enough, yes [12:34] pitti: please don't [12:35] gedit has an arch any/all split; build skew will render it uninstallable [12:35] ok [12:35] :( [12:35] ari-tczew, pong btw :) [12:35] well, maybe not since the dependencies aren't actually all that strict [12:35] it doesn't have a strict dependency, though [12:35] but surely it can wait for after alpha-2? we have a lot of stuff still to build, like most of X [12:36] ogra: hi, do you will have time for my 2 bugs? (we talked PM last time) [12:36] so I'll just review then [12:36] so I would rather not have main uploads that will have to be scored down [12:36] cjwatson: ok but how long it needs to wait? [12:36] alpha-2 is due on Thursday, although it will depend on QA cycles [12:36] you can see this from the release schedule [12:37] ari-tczew, yes, working on defoma atm but wont upload until friday (due to softfreeze) [12:37] it doesn't change anythign visible anyway, so it's not urgent in any way [12:37] (defomas build-deps are hilarious) [12:37] for now I just reviewed/approved the MP === MacSlow|lunch is now known as MacSlow [12:39] ogra: yep [12:40] cjwatson: I think everything in http://cooperteam.net/Packages/ is good to go. [12:41] RAOF: OK, great, thanks - I'll start pushing [12:42] modulo wet-string internet [12:42] want me to take half of them? [12:43] cjwatson: ^ [12:43] nah, it's ok [12:46] hi cjwatson , there is the bugreport for os-prober: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611670 === diwic is now known as diwic_afk [12:49] Ok. Midnight says hello, time to go to bed. [12:49] dholbach, i had the pleasure of meeting Kate LCA2011 she informed me that packages are still being accepted for natty. I was wondering if acire will be packaged for natty? [12:51] jfer, I have no idea [12:51] ogra: Argh, urgh. Despite my first test-build somehow suggesting otherwise -kbd doesn't build against 1.10; I'll look at that tomorrow morning. Sorry. [12:52] RAOF, no hurry [12:52] RAOF, i cant use natty on the ac100 yet until the new QT is uploaded [12:52] (since i need to use unity-2d and QT without NEON) [12:53] is it possible to be included in natty? [12:54] jfer, can you talk to Jono about that? I really don't want to be assumed to be the "maintainer" for it [12:54] ok thanks. [12:54] dholbach, sissy [12:54] jfer, I did one upload to the PPA to backport it and could see if it's possible to upload a natty version to the PPA [12:54] ogra, you do it! [12:54] nah [12:54] :) [12:54] there you go [12:55] i am keen to get involved in packaging. is the mentoring program still happening? [12:55] jfer, unfortunately not, but I suggest you still check out https://wiki.ubuntu.com/MOTU/GettingStarted and just ask your questions in #ubuntu-devel or #ubuntu-motu [12:56] there's no dedicated 1-on-1 mentoring that I know of [12:56] but there's still a lot of very helpful people [12:56] #ubuntu-packaging is a good spot for packaging questions. Its very quiet, though... [12:56] ok. i have watched a few of your videocasts and they are most helpful but i am keen to get working on some real examples [12:57] excellent [13:06] dholbach: hello. when we can except first conclusions from surveymonkey? [13:07] ari-tczew, in a few weeks [13:08] aha [13:08] lunch time :) [13:11] RAOF: all done now, thanks! [13:12] oh, damn, except it would have helped to use dupload -t ubuntu :-/ [13:13] * cjwatson expects a giant pile of reject mails [13:41] RAOF,bryceh,Sarvatt: the new xserver-xorg-input-mouse fails to build - do you know what's going on there? [13:41] cjwatson: FYI, I uploaded the pywebkitgtk fix, as it's quite important, but lowered its build score [13:42] so it doesn't get into the way [13:42] pitti: is that the one that fixes a crash in ubiquity? [13:42] oh, no [13:42] it previously shipped a broken .pyc file [13:42] ok, that's fine, thanks [13:43] cjwatson: can I see the build log of xserver-xorg-input-mouse, please? [13:44] http://launchpadlibrarian.net/63296363/buildlog_ubuntu-natty-i386.xserver-xorg-input-mouse_1%3A1.6.0-1ubuntu1_FAILEDTOBUILD.txt.gz [13:44] ah, thanks [13:44] it probably needs an update to build against the new headers [13:45] right, there hasn't been a release for 1.10 [13:45] yes, I agree [13:45] raof just bumped the build-dep [13:45] hmm [13:45] yeah that's not quite enough :) [13:49] the warnings aren't critical, but looks like at least two commits from master should be needed for input abi 12 [13:50] I guess with nvidia/fglrx being broken right now, I better disable those in Jockey for now [13:51] there could probably be a check for the video abi, and not offer those if the current package doensn't provide the current abi? [13:51] right, that'd be a more permanent solution [13:51] pitti: yes, that would be a good idea, until I fix the dependecies as discussed in the ubuntu-x mailing list [13:51] there's been some talk about how to handle the abi's on ubuntu-x@ [13:53] tjaalton, tseliot: so if I get the -video-abi from xserver-xorg-core, and check it against the driver package's Provides: line, would that be correct? [13:53] (why is it provides: and not depends:?) [13:54] pitti: yes, I think so [13:54] tseliot: fglrx doesn't seem to have a video-abi provides at all? [13:54] I haven't read the thread with thought, yet, so can't comment in detail :) [13:55] tjaalton: that's how it currently seems to work, anyway [13:55] pitti: well, it doesn't really work currently [13:55] pitti: RAOF pointed me to ${xviddriver:Depends} which is what I should use instead of provides [13:55] hm, so perhaps I just generally disable them for now, and postpone the ABI matching when this got settled [13:55] yeah [13:56] pitti: I thought I added that in fglrx but I'll definitely fix it in the next upload [13:56] +1 [13:57] tseliot: I'll check -mouse and upload whatever builds against the new abi [13:57] tjaalton: ok, good. Let me know if you need a hand [14:00] tseliot: when we can except the fix for nvidia? [14:01] * cjwatson cleans up the xserver-xorg-video-s3virge reject [14:02] ari-tczew: if you mean "expect", I'm still thinking about it [14:03] I noticed the thread in ubuntu-x only last night [14:11] tseliot: grah, those abi patches don't apply on 1.6.0, probably best to pull master (1.6.99) [14:11] not that it's hugely different in functionality.. [14:12] tjaalton: they are going to release something stable together with X, aren't they? So it shouldn't be a problem [14:13] tseliot: right, it was recently bumped to 1.6.99 [14:14] yes, I see that in git [14:14] actually, that happened two months ago, but there's only one commit after that :) [14:16] are one of you two able to do the upload? [14:16] cjwatson: yeah, I'll pull git master and wrap it up [14:16] cool, thank you === dholbach_ is now known as dholbach === evilvish is now known as vish [14:31] can anyone tell me why mcollective got rejected? === Quintasan_ is now known as Quintasan === Dr_Who is now known as tgall_foo === oubiwann is now known as oubiwann_ [14:37] zul: you didn't get an e-mail? [14:37] zul: Riddell was doing some source rejections earlier ... [14:37] a bit of historical curiosity: perhaps it's a false memory, but wasn't the plan a while back to use the desktop wallpaper for the plymouth theme so that the visual transition was minimal? [14:37] cjwatson: no i might have missed it or maybe dustin got it [14:38] zul: https://lists.ubuntu.com/archives/ubuntu-archive/2011-February/039057.html [14:38] Riddell: ok thanks [14:38] which also went to uploader kirkland [14:39] ev, quite a *while* back ;) [14:40] any idea why it didn't come to fruition? [14:40] * ogra hanst heard anything of that plan after dublin [14:40] *hasn't [14:40] hm [14:40] i guess because it was still usplash back then [14:41] and likely nobody thought about it later when we got plymouth [14:58] cjwatson: well, the new -mouse doesn't build on pbuilder because the headers are not available (yet). but it should be good for the archive builder, I guess [14:59] tjaalton: go ahead if you think it's good, I don't think it can get significantly worse [14:59] heh, right.. let's see how it goes [14:59] there === sconklin-gone is now known as sconklin [15:11] cjwatson, have you or know of a 2.6.38 kernel with grub 0.97 ? asking due to bug 710754. [15:11] Launchpad bug 710754 in linux (Ubuntu) "natty kernel does not boot on t1.micro in arch i386" [Medium,In progress] https://launchpad.net/bugs/710754 [15:11] smb has a workaround for -virtual kernel on ec2, but I was curious if it affected all grub 0.97. [15:15] For completeness, the bug seems to be in pv-grub which is based on the old grub. It may or may not have the same potential in there. [15:15] smoser: I don't know, sorry [15:15] pvgrub makes my head hurt [15:16] Heh, it surely did that to me even without looking at its code [15:16] let me check grub's changelog briefly [15:16] but I seriously doubt there've been any recent fixes in grub; last upstream commit in our package was 2005 [15:17] don't see any relevant changes post 0.97, on a quick glance [15:18] cjwatson, Probably just something to keep in the back of our memories should there be somone approaching us with strange boot failures with grub0.97 and Natty or later i386 kernels. [15:18] Not this should be expected to happen... [15:20] @pilot out === udevbot changed the topic of #ubuntu-devel to: Archive: Soft freeze for Alpha 2 | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: [15:22] cjwatson: -mouse built, whee [15:47] kees, jdstrand: postgresql just announced new security/bug fix release, I'm preparing them [15:47] kees, jdstrand: are we actually still concerned about this buffer/int overflow in a security sense? http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=7ccb6dc2d3e266a551827bb99179708580f72431 [15:47] kees, jdstrand: or do we merely handle that as a DOS now? [15:48] i. e. do you want -security or SRU? [15:48] well, I guess at least teh dapper backport should go to -security [15:48] pitti: a DoS in a server we consider medium priority [15:48] jdstrand: ok, so still -security? [15:48] pitti: so -security would be great [15:48] jdstrand: ok; I'll open a tracker bug for it, and put the updated source packages somewhere [15:49] jdstrand: 8.1 went out of support, so for that I'll backport the fix [15:49] * pitti hopes there won't be too many security vulns in 8.1 any more until June [15:49] pitti: thank you. not sure who will process it, it could be sbeattie or mdeslaur (or me or kees) [15:49] * jdstrand nods === bjf[afk] is now known as bjf === deryck is now known as deryck[lunch] [15:50] pitti: but if you subscribe us to the bug (and ping as you helpfully do :), we'll sort that out. thanks for preparing them :) [15:50] jdstrand: yup, as usual; thanks! === beuno is now known as beuno-lunch [16:20] has marking a function (a constructor) as "inline" (or not marking anymore) any influence on the API or ABI of a library? [16:20] in some cases at least [16:20] I painfully remember that with apt [16:20] inline constructors in C++ really really suck, and should be avoided [16:21] I'm looking at http://launchpadlibrarian.net/58061583/buildlog_ubuntu-natty-i386.gecode_3.4.0-1_FAILEDTOBUILD.txt.gz [16:21] as you can't change them without breaking ABI [16:21] (in a behavioural way, not in a compile sense) === juliank0 is now known as juliank [16:23] as the .so version got bumped would that be a good chance to drop the "forceinline" from that contructor (or even all constructors)? [16:23] Inline functions should be avoid in general, unless you can prove that there is a large advantage in a performance-critical path [16:23] s/avoid/avoided/ [16:27] geser: sounds good; but that should probably happen upstream? [16:29] pitti: upstream "fixed" it by not providing a constructor with a va_arg list anymore. I'll wait on the DD to package it (upstream released it today) [16:30] geser: but still inline? [16:30] yes, even "forceinline" :( [16:31] geser: that's still broken, as with that you can't change the implementation of the ctor at all without breaking abi [16:49] /var/lib/dpkg/tmp.ci/config: 127: cannot create /dev/null: Permission denied [16:49] * ogra scratches head [16:49] something's removed your /dev/null [16:49] i'm in a chroot [16:50] make a new one [16:50] and am root [16:50] unfortunately it's extremely hard to retrospectively find out what [16:50] and /dev isnt bind mounted or anything [16:50] oh [16:50] so i should be able to just write to it as a file [16:50] ogra: cat /proc/self/status | grep Cap [16:50] proc isnt mounted either [16:50] biab [16:51] nodev mount option? [16:51] oh ! [16:51] indeed !!! [16:51] * ogra higs geser [16:51] err [16:51] *hugs === deryck[lunch] is now known as deryck === beuno-lunch is now known as beuno [17:44] is there a quick tool that I can use to check two revisions against the debian version sort algorithm? [17:45] dpkg --compare-versions [17:45] I knew it had to be simple. :) === sforshee is now known as sforshee-lunch [18:08] RAOF, bryceh: You broke my unity :/ [18:11] kenvandine: you have the same issue on intel as well, isn't it? ^^ [18:11] Amaranth is suggesting an FBO issue [18:11] compiz itself works but if you enable the unityshell plugin and start compiz it segfaults, getting a backtrace now [18:11] tumbleweed: Thanks for fixing up requestsync. Sorry I had to run off (due to $WORK stuff). [18:13] #5 0x00007fce31d361c5 in nux::IOpenGLFrameBufferObject::Deactivate() () from /usr/lib/libnux-graphics-0.9.so.0 [18:13] So yeah, FBO :/ [18:13] I think it fails to setup the FBO and crashes trying to clean up [18:15] Amaranth, your welcome [18:16] Amaranth, actually, I booted unity successfully yesterday with the new stack and it seemed to be working fine for me [18:21] didrocks, what's the issue? [18:21] didrocks, yes mine is on intel [18:22] kenvandine: look ^^ [18:23] what are you debugging? [18:23] didrocks, i didn't try disabling unity [18:23] let me do that [18:24] I just restart after some upgrades and my session was empty [18:24] no compiz, no nautilus, no gnome-panel running [18:24] it's weird [18:24] seb128: maybe you have the same issue than Amaranth is describing [18:25] Amaranth, isn't that a grain used by the aztecs? [18:25] bug 711396 [18:25] Launchpad bug 711396 in xserver-xorg-video-intel (Ubuntu) "segfault in nux::IOpenGLFrameBufferObject::Deactivate" [Undecided,New] https://launchpad.net/bugs/711396 [18:26] I'm thinking we have a gnome-session issue here as well [18:26] Apparently this case has never been tested [18:26] Amaranth: what issue in gnome-session? [18:26] didrocks: It doesn't do any kind of fallback [18:26] or even start nautilus... [18:27] Amaranth: not, nautius and gnome-panel is starting [18:27] Amaranth: look bug #711378 [18:27] Launchpad bug 711378 in compiz (Ubuntu) "after compiz crashed, gnome-panel isn't mapped again" [High,Triaged] https://launchpad.net/bugs/711378 [18:28] didrocks, yes compiz works fine with the unity plugin disabled [18:28] after compiz crashed compiz has no say in gnome-panel being mapped ;) [18:28] Amaranth, right... that is the bug i filed earlier [18:28] Amaranth: right, but gnome-panel is running if you look at a tty [18:29] Amaranth: so as I have no input there, we filed the bug report [18:29] didrocks: In this case I'm pretty sure nothing is running [18:29] Amaranth: did you try on a tty? [18:29] I know nautilus isn't, at least [18:29] geser: kenvandine and I tried [18:29] Yeah, I looked with pgrep, didn't look for gnome-panel though, just nautilus [18:29] and confirm it's running in that case [18:29] But when I got a terminal running and ran unity --reset it was complaining about dbus and such not running so it seems I got no session [18:30] s/terminal/gnome-terminal/ [18:30] Amaranth: try with that, you should have gnome-session and others… it has been confirmed [18:30] just that's it's not mapped [18:30] Amaranth: my gnome-panel is running as I send a SIGHUP to the gnome-panel to make it "reappear" [18:32]