[00:50] jcastro: how do I submit a juju charm to the massive repo of them? [00:50] http://juju.ubuntu.com/Charms [00:51] has the steps [00:51] kees, push to your branch, file bug, attach branch, tag it "new-charm" when you want a review [00:51] kees, is this for the contest? [00:54] jcastro: well, I thought I did all that before, but I figured I'd try it again as long as there was a contest. ;) [00:54] link me up with the bug # when you have it and I'll pile it on the queue! [00:55] kees, nice to you have you around rocking it still. [00:57] jcastro: thanks! [00:57] https://bugs.launchpad.net/charms/+bug/961819 tada [00:57] Launchpad bug 961819 in Juju Charms Collection "New "sbuild" charm for build environments" [Undecided,New] [00:57] okay, looks like I had just missed that last step then. heh [01:29] any tips for dealing with test suites in packages that call libraries (python library, in this case) that try accessing things that are off limits on the build system (/dev/log, in this case) [01:46] adam_g: fixing the test suite not to access things that are off limits? [01:47] jamesh: its not the test suite, its a library that the test suite requires. suppose there is no fix other than skip those tests [01:48] adam_g: or provide a fake implementation of the library function [01:48] or perhaps one of the mock libraries === Ursinha is now known as Ursinha-away === jalcine is now known as jalcine_ === jalcine_ is now known as jalcine [02:52] can someone sponsor some attention to this bug. given that it's openssl, it will likely affect quite a few installs and packages. this applies to 64bit installs that do not have processors with AES-NI [02:52] https://bugs.launchpad.net/ubuntu/+source/openssl/+bug/940230 [02:52] Launchpad bug 940230 in openssl (Ubuntu) "openssl on 64bit is much slower than before" [Undecided,New] [02:53] it should be easily fixable with compile flags. i.e. do not compile ASM in for the x86_64 build... === jalcine is now known as jalcine_ === jalcine_ is now known as jalcine === jalcine is now known as jalcine_ === jmp is now known as Guest72352 === Guest72352 is now known as _jmp_ === jalcine_ is now known as jalcine [05:52] Good morning [05:53] unity-2d in precise use to work and now it doesn't [05:53] at 800x480 on fbdev @ 16 bpp [05:57] scientes: Can you be more specific than “doesn't work”? [05:58] it crashes at launch [05:58] it takes al long time to time out before apport comes up [05:58] but then its "cant acess memory at XXXXX" [05:58] i tried with guest login too [05:58] and i purged xorg-edgers [05:59] the desktop also gets posted [05:59] *desktop image [06:03] how do i launch unity from the command line? [06:06] scientes: unity [06:07] unity-2d [06:08] pooh wow, that worked [06:08] maybe its a multi-seat problem [06:08] oh no, its using llvmpipe, and slow as balls [06:08] and horribly corrupt [06:08] how do i run unity-2d from command line? [06:09] #ubuntu might know [06:10] unity-2d-shell: [FATAL] Settings schema 'com.canonical.Unity2d.Launcher' is not installed [06:12] balls move quickly when kicked. [06:12] just saying, hehe [06:15] old version, fixed [06:28] Hi, I'm using python-distutils-extra, and when I build under Lucid, data_files are missing from POTFILES.in, while under Precise everything is OK. I'm currently using a potfiles-workaround/ dir with symlinks to the files that need translations, but that's too ugly, any better solutions? === BenC__ is now known as BenC [07:08] cjwatson: ops, screwed check then [07:14] micahg: what's the state of the lightdm gtk greeter? [07:15] Hi, I'm using python-distutils-extra, and when I build under Lucid, data_files are missing from POTFILES.in, while under Precise everything is OK. I'm currently using a potfiles-workaround/ dir with symlinks to the files that need translations, but that's too ugly, any better solutions? [07:46] good morning === tkamppeter_ is now known as tkamppeter [08:18] hrw: well, I fixed pcmciautils, anyway [08:21] cjwatson: thx [08:27] cjwatson: what would you suggest to do with rest of packages? [08:27] cjwatson: ignore, convert to debhelper? [08:28] hrw: not sure :) [08:28] it doesn't seem desperately vital that ed doesn't have debug symbols, say [08:28] sure, but would be nice to have one for bzip2 for example [08:29] maybe an Ubuntu-specific change to call pkg_create_dbgsym in the appropriate place [08:29] I don't think converting packages to debhelper as an Ubuntu delta is a good idea [08:29] like binutils do ;) [08:29] no, sending delta back should be done [08:29] it would have to check /CurrentlyBuilding [08:30] there are cases where maintaining a delta is the right thing to do [08:30] btw - can package has arch specific dependency? build-dependencies can do that but I am not sure about dependencies [08:30] you probably aren't going to make many friends by sending patches that completely change the packaging [08:30] I know [08:30] yes, provided that the package is not Architecture: all [08:30] thats the problem [08:30] then no [08:31] laptop-detect is shellscript which is arch:any cause on x86 it has one more dependency [08:31] that's a legitimate reason to be arch: any [08:31] it doesn't just mean "compiled code only" [08:31] I know, just wondered [08:32] mvo, can you have a look at https://bugs.launchpad.net/apt/+bug/960914 ? [08:32] Launchpad bug 960914 in update-manager (Ubuntu) "Can not update packages - nothing happens after package download" [Undecided,New] [08:40] toabctl: sure, let me check that one [08:42] toabctl: let me follow up in the bugreport I asked a bunch of questions there [08:42] (now) [08:45] hey mvo, good morning [08:45] hey pitti good morning to you as well [09:01] mvo: in bug 950676, do you think it would help to remove libseed0 early with a breaks: or even conflicts:? it seems to be pretty close to the root of the evil [09:01] Launchpad bug 950676 in apt (Ubuntu Precise) "lucid->precise upgrade failure due to gir1.0->gir1.2 conflicts" [High,New] https://launchpad.net/bugs/950676 === jalcine_ is now known as Guest61506 === Guest61506 is now known as JackyAlcine === JackyAlcine is now known as jalcine === jalcine is now known as Guest81870 [09:43] StevenK, hey - any change you could accept the zentyal-* binary packages into precise please? [09:44] * StevenK reaches for cocobanana. [09:51] jamespage: I'm a little concerned some of the source packages are 2.3.4 and some are 2.3.3, but they all look good, so I'm accepting them. [09:51] StevenK, great - upstream release each component separately hence the different minor version numbers... [09:52] thanks v much === fenris is now known as Guest95871 [10:13] pitti: sorry, I was deep in a code review, let me check [10:15] pitti: getting rid of libseed0 early would certainly help [10:15] pitti: from what I see in the logs [10:16] mvo: still not sure whether this is a real apt bug, or just bad dependencies [10:17] mvo: ok, I'll try to upload a seed with that conflicts, we'll see how far we'll get [10:17] s/try to// [10:17] I wonder why its installed in the first place [10:17] mvo: some universe package in lucid apparently pulled it in [10:18] pitti: give me some minute to read the log some more [10:18] at least: [10:18] Broken libseed0:i386 Depends on gir1.0-glib-2.0 [ i386 ] < 0.6.8-1 > ( libs ) (>= 0.6.3) [10:18] Considering gir1.0-glib-2.0:i386 17 as a solution to libseed0:i386 -1 [10:18] Removing libseed0:i386 rather than change gir1.0-glib-2.0:i386 [10:18] looks like its doing the right thing [10:44] openssl-1.0.1 is released. is it too late to ship for precise? [10:44] I was actually pondering that myself [10:45] planning to look into it today and do some testing, and get a feature freeze exception request in if it looks good; there are some problems I've not been able to solve otherwise [10:45] there are numerous improvements. which would be beneficial for a LTS [10:45] yeah, I know [10:45] if you look at it, it would be much appreciated. [10:45] I'm not averse to it, it would be better than giant hairy backports [10:46] but there is certainly a risk so I need to do due diligence [10:46] also if you could disable asm for ubuntu x86 64bit build [10:46] I'd rather investigate first [10:46] there's a big regression of using asm vs. C starting with 1.0.0 [10:46] I'm aware of that bug but I do not propose to agree to any particular solution until I've looked into it myself [10:47] no worries. i have a bug opened for it. with a bunch of links that should explain. [10:47] as well as timings. [10:48] heh, there's already an FFe open. I do so love it when non-developers open FFes ;-) [10:48] i was just hoping to get these all fixed given that precise is LTS and openssl affects so many other packages that perf would be fixed. =) [10:48] what's the url to that FFe? [10:49] pitti: so what is still depending on the old gir? is it just a incomplete transition? Considering gir1.0-gtk-2.0:i386 3 as a solution to gir1.2-gtk-2.0:i386 3 shows that there are probably 3 rdepends (but they might be still on the installed system) [10:50] jhojho: it's got no useful information in it yet; I'd rather people didn't pile on it please [10:50] pitti: the log is a bit confusing, nevermind I keep diging [10:50] i wasnt planning on adding anything to it. [10:51] just to watch/track. [10:51] jhojho: also are you sure that 1.0.1 still has the performance regression on 64-bit? the thread at http://thread.gmane.org/gmane.comp.encryption.openssl.devel/19836 suggests this shouldn't be so [10:51] aside from some slowdown to resist cache timing attacks [10:51] pitti: aha, I think that contributes to the problem that there are no direct gir1.2-gtk-2.0 dependencies, this means apt has a harder time figuring out if the package is important or not [10:52] the FFe is bug 958430 [10:52] Launchpad bug 958430 in openssl (Ubuntu) "[FFe] Please sync openssl 1.0.1 from Debian unstable" [Undecided,Confirmed] https://launchpad.net/bugs/958430 [10:52] pitti: I think that moving gir1.0-gtk-2.0 from option to extra (and the other gir1.0 stuff) fixes the bug too [10:53] sladen, do you think you could have a look into the ubuntu-font MP again? [10:53] mvo: we can't, the gir1.0-* packages have gone long ago [10:53] cjwatson: it's not clear. I'm happy to do some timing tests if a package is built. [10:54] cjwatson: looking at the changelog, the constant time implementation that I see, is the stuff contributed by google [10:54] for elliptic curve [10:54] so not the same thing. [10:54] mvo: "Installing gir1.2-gtk-2.0 as Depends of gir1.2-dbusmenu-gtk-0.4 [10:54] "Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command [10:54] line to include this in your build of OpenSSL" [10:55] for 64bit only [10:56] cjwatson: in that gmane link you posted, the last sentence says "If non-AES-NI [10:56] performance in FIPS context is important to you, contact [10:56] opensslfoundation.com." [10:57] so I'm thinking it's probably not fixed for asm (or considered a priority item). and since using the asm codepath drops off about 50% for aes, it's kind of a big regression. [10:59] dholbach: if it's going up, I'd like it with the ~medium (as we have different combinations in different PPAs and this means the combinations are clear) [10:59] sladen, can't you upload it with that change? [11:00] sladen, we are going to have Beta 2 Freeze today [11:00] cjwatson: here's what I've found so far.. https://bugs.launchpad.net/ubuntu/+source/openssl/+bug/940230 [11:00] Launchpad bug 940230 in openssl (Ubuntu) "openssl on 64bit is much slower than before" [Undecided,New] [11:00] so better get it in quick before [11:00] dholbach: however it is a massive (but minor) kludge that could be done at any point, and so I think this it makes sense to do it at T-5/6 weeks when it's probably possible to get a real fix in that time [11:00] hi there! [11:00] T-5/6 weeks? you mean after the release? [11:00] dholbach: and additionally; if it's happening, I don't want to do it before the wallpapers go up, so I don't lose the CD space [11:00] jhojho: I think I'm more worried about non-FIPS context for Ubuntu [11:01] dholbach: release is 5 weeks away [11:01] so I disregarded that last sentence [11:01] mvo: do you think a conflicts: libseed0 or breaks: will be better? [11:01] pitti: ohhh [11:01] cjwatson: i dont need FIPS so i'm fine with that =) [11:01] pitti: I think a conflict will be (slightly) better [11:01] mvo: ok, that's my gut feeling, too [11:01] mvo: "oooh" sounds promising [11:02] sladen, ok, I don't understand what you're saying, but I'll leave the decision to you - the merge proposal looked like you wanted Florian just to update the version number, which looked like a minor thing to block on [11:02] mvo: will a replaces: help to ease apt's mind? [11:02] pitti: no, sorry, it was about that its gone from the archive :/ [11:02] pitti: as the optionl->extra would be a nice fix IMO [11:02] mvo: it was renamed for parallel installability, but now it _is_ the libseed library [11:02] dholbach: that one merge results in 3-4 other uploads needing to go into PPAs so they keep ahead [11:02] dholbach: I'm not keen on doing it [11:03] dholbach: however, if it's done, I'd prefer it being done in the prettiest way [11:03] ok, I'll leave it to you - shall I mark it as "WIP"? [11:03] pitti: the conflict should be good enough (I think), but a C/R/P will not hurt if its really the same [11:03] it doesn't seem to make sense to have it in the sponsoring queue [11:04] mvo: "provides:" would be exaggerating, and technically wrong [11:04] mvo: I have C/R now [11:05] ok [11:05] fingers crossed :) [11:05] mvo: uploaded; looking forward to tomorrow's run then [11:05] mvo: danke! [11:07] cool [11:08] cjwatson: anyways thanks for looking at it as not having to wait another 2 years for TLSv1.1 and TLSv1.2 would be nice. [11:14] infinity: eek, that d-i FTBFS looks unhealthy.. [11:16] pitti: Nah, it's fine. [11:16] pitti: Images just grew a bit when nic-firmware.udeb got more stuff added. [11:16] oh, "disk full" [11:16] pitti: Either Colin will fix it while I sleep, or I'll fix it when I wake up. [11:16] ok, thanks [11:17] pitti: Yeah, "disk" being the FAT image, not the buildd's disk. ;) [11:17] right, I just missed that at my first look [11:19] it's running through sbuild locally now [11:41] can someone try to rebuild mawk? it fails for me on testing [11:46] hrw: works fine here in a precise/i386 sbuild instance [11:49] thx [11:50] hrw: do you want the log? [11:52] no, built fine in pbuilder here as well [11:53] ok === _salem is now known as salem_ === MacSlow is now known as MacSlow|lunch === Ursinha-away is now known as Ursinha === toto is now known as Guest76177 === Guest76177 is now known as doctorpepper === smb` is now known as smb [13:04] gcc 4.7.0 is out… http://article.gmane.org/gmane.comp.gcc.devel/125441 [13:11] hello. i've got a question on ubuntu audio dev ppa. i'm missing alsa-driver-1.0.25 package for precise x64. does anybody know if this one will be part of the release? [13:12] right now there seem to be only x86 packages === MacSlow|lunch is now known as MacSlow === bladernr_afk is now known as bladernr_ === dendro-afk is now known as dendrobates [13:23] have you the package name at hand? [13:25] can I express a dependency on a version like this: postgresql (8.4 | 9.1)? [13:27] no [13:27] the policy manual has a complete syntax for dependencies [13:28] cjwatson: right, I was looking at this page but it wasn't mentionned. I wasn't sure whether it was because it wasn't supported or just wasn't mentionned: http://www.debian.org/doc/debian-policy/ch-relationships.html [13:30] anything not there is not supported [13:30] package should be like this: http://packages.ubuntu.com/en/source/precise/alsa-driver [13:30] but there seem to be no builds for x64 [13:33] i'm confused [13:33] jenni82: all the binary packages in the alsa-driver source package are Architecture: all, so it only needs to be built on one architecture [13:33] jenni82: this source package only builds arch:all packages (can be installed on any architecture), so it only needs to get build once (which is done on the i386 buildd) [13:36] hmm ... but why does "$ cat /proc/asound/version" deliver Advanced Linux Sound Architecture Driver Version 1.0.24. [13:37] That's up to the kernel packages. In Ubuntu the alsa-driver source package really just delivers configuration [13:38] i was hoping to get rid of some sound-issues. so this means i have to install 1.0.25 manually? [13:39] I don't know whether the kernel team plan to bump to 1.0.25 [13:41] ok. thank you. === yofel_ is now known as yofel [14:53] jdstrand, apw: sorry, in bug 947254 I accidentally also copied to -security :/ [14:53] Launchpad bug 947254 in Kernel SRU Workflow "linux: 2.6.38-13.57 -proposed tracker" [Undecided,In progress] https://launchpad.net/bugs/947254 [14:54] but I can't just remove the current package from -security again, as we actually want to keep 2.6.38-13.56 there [14:55] so we'd either need to reupload, and cause yet another update, or do an USN explaining this? [14:55] or find a Launchpad DB hacker to unpublish it? [14:55] cjwatson: ^ if you happen to have a brilliant idea [14:56] I suspect you're better off leaving it there, unless it's really bad [14:57] it was regression tested and everything, it's just not a security update [14:57] ... for the first time in months, I guess I got way too used to always copying everything to -security [14:58] saw too late that the "promote-to-security" task was "invalid" [15:00] jjohansen: ^ [15:02] pitti: okay, we can write a USN to explain this [15:03] jjohansen: thanks; and sorry [15:07] ok, I think I got the other umpteen kernels correctly === jespada_ is now known as jespada === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates === vibhav is now known as Guest43747 [15:39] pitti, i think i'd vote for just leaving it [15:59] pitti: re lightdm-gtk-greeter> I've got one more thing to finish before I finish that, should be ready by 19:00 UTC [15:59] thanks [16:14] is there an archive admin around who can bump the nova upload thru https://launchpad.net/ubuntu/precise/+queue ? the new binary package is just a launch script that was previously packaged with nova-console, plus a manpage and upstart job [16:14] pitti: do you recall what checks are done by apport before filing a bug in Launchpad? I'm looking at bug 960851 from a Zorin OS install [16:14] Launchpad bug 960851 in ubuntu-meta (Ubuntu) "package ubuntu-desktop 1.245 failed to install/upgrade: dependency problems - leaving unconfigured" [Undecided,New] https://launchpad.net/bugs/960851 === dendrobates is now known as dendro-afk [16:18] seb128: is my desktop on precise supposed to be completely broken? [16:18] seb128: i log in an i see nothing on unity or gnome3 [16:18] didrocks: or it's your fault?? ^^ [16:19] nxvl: before asking "who's fault it is" you should look at what you did upgrade === dendro-afk is now known as dendrobates [16:19] as there has been no unity upload at all in the official repo, I would doubt of it [16:20] didrocks: the "fault" think was just a joke, not actually looking for whos fault is, sorry about confussion [16:20] didrocks: i did an upgrade earlier this week and my desktop is completely gone [16:20] no worry :) [16:21] hey nxvl [16:21] didrocks: is that known or just randomness? [16:21] dholbach: hi daniel! [16:21] nxvl: not known from here at least [16:22] sorry, handling unity release, don't really have time for debugging it [16:22] look at what you upgrade [16:22] upgraded* [16:22] didrocks: funny part, when i get to lightdm i get no icons, just blank squares with red 'x' [16:22] there has been some new lightdm as well [16:39] Does i18n/keyboard.txt actually have any effect in ubuntu-defaults-builder / ubuntu-defaults-image ? [16:39] The context is that I 'm trying to build a localized image for Greek [16:40] and I noticed that in the resulting image there was no english keyboard :( [16:41] grepping the various scripts I see no reference to keyboard.txt besides verifying it [16:42] fwiw the source of the package is here: https://github.com/ubuntu-gr/ubuntu-defaults-el-gr [16:47] slangasek: are we planning to switch to a tmpfs /tmp/ in 12.10? [16:52] depends, you volunteering for the installer work? :-) [16:52] cjwatson: well, there are certainly programs to be fixed [16:52] I can instantly think of Firefox (uses it for downloads) [16:52] and presuambly also some CD burners [16:53] these should use /var/tmp/ for large things [16:53] cjwatson: OOI, what needs fixing in ubiquity? that's not somehing that sprang to my mind [16:53] (just got a question about it) [16:53] I think there's a reasonable argument that tmpfs /tmp ought to be selectable in the partitioner rather than mandated [16:54] and it may well require adjustment of autopartitioning limits and such [16:54] ah, ok [16:54] https://blueprints.launchpad.net/ubuntu/+spec/server-karmic-tmp-as-tmpfs refers to bug 386554 as a blocker, I haven't checked yet whether that's still current [16:54] Launchpad bug 386554 in linux (Ubuntu) "System behaved as if OOM when it had plenty to spare" [Medium,Triaged] https://launchpad.net/bugs/386554 [16:54] so either way, it's not a trivial change for multiple reasons [16:55] I was just curious whether I was missing a spec, or a bug (there's none against mountall) [16:55] that was the canonical spec for it, I think, yes [16:55] but I'm fairly sure I read about it on some ML [16:55] or some bug, or other place [16:55] damn memory [16:55] parts of my brain are on tmpfs, too :) [16:55] pitti: it was mentioned recently on the TB ML IIRC [16:56] ah, perhaps that one, thanks === zyga is now known as zyga-food === fenris is now known as Guest70667 === Guest70667 is now known as ejat === jalcine_ is now known as Guest10125 === Guest10125 is now known as JackyAlcine [17:21] mvo: do you have a good suggestion on how to fix bug 944876? Just shifting the mapping is more of a hack than a fix, although s-p-kde would probably need a rewrite anyway at some point [17:21] Launchpad bug 944876 in software-properties (Ubuntu) "changed mapping of release_upgrades_policy causes software-properties-kde to set the wrong policy" [High,New] https://launchpad.net/bugs/944876 === JackyAlcine is now known as Guest91681 [17:28] @pilot in === udevbot changed the topic of #ubuntu-devel to: Precise Beta-1 Released. Archive: open | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bryceh [17:28] bryceh: enjoy the flight :) [17:31] ivoks: patch in bug 961166 - was that tested? there's a .. vs. ../.. discrepancy, which seems to be due to having backported only half of 8316bd2d9813cbc7b2b8288b6618eec2c2004028 [17:31] Launchpad bug 961166 in live-build (Ubuntu) "lb_binary_disk doesn't check compression of the initramfs" [Medium,Triaged] https://launchpad.net/bugs/961166 [17:32] cjwatson: let me revisit that [17:32] cjwatson: ups... my bad, sorry [17:33] it might be worth taking the whole of that patch [17:33] pitti, thanks, actually since we're nigh-frozen, probably just going to tend srus [17:33] bryceh: there might be a couple of universe fixes, too [17:33] but yeah, ther's a fair number of SRUs, too [17:37] pitti, is universe not included in this freeze? [17:37] bryceh: it is, but it's feature/UI freeze just like main [17:37] but UI is "less" frozen for universe [17:38] the definition even says it doesn't apply to universe [17:38] and of course they are less dangerous usually === zyga-food is now known as zyga [17:42] pitti, ok, thanks. I still plan to make srus my focus; like you say there's a lot of 'em. [17:44] pitti: I thought the discussion we had a (few) weeks ago concluded in a general consensus that U/I freeze applied to everything (like FF), but that for unseeded packages that aren't in any docs package they would be trivially approved? [17:44] right, that's why I said "less" frozen [17:45] I haven't checked whether there are any UI breaks in teh sponsoring queue, and whether they have UIFE bugs [17:47] agateau: are you sure 'libqt4-gui' is the package you meant to ask me for in bug #915801? Because libqt4-gui:i386 isn't installed (it's not a dep of skype:i386) [17:47] Launchpad bug 915801 in sni-qt (Ubuntu Precise) "sni-qt seems to no longer let skype show up as an indicator in precise" [High,Confirmed] https://launchpad.net/bugs/915801 [17:48] slangasek: ahah, maybe that is the problem [17:48] really [17:48] ? [17:48] is your skype package statically linked against qt? [17:49] well, it's the stock package [17:49] let's see [17:49] $ ldd /usr/bin/skype |grep QtGui [17:49] libQtGui.so.4 => /usr/lib/i386-linux-gnu/libQtGui.so.4 (0xf6b07000) [17:49] ah no, I was wrong [17:49] it is libqtgui4 [17:49] ok [17:49] libqt4-gui is a transitional pacakge [17:49] *package [17:51] ok, versions sent to the bug [17:54] this is really a curious one; sni-qt has worked fine here all the time [17:54] I wonder what the heck is different [17:54] (with skype) [17:54] pitti: Right, but I think we need to update the definition on the wiki to match that. [17:56] pitti: yeah, I'd also like to know ;) [17:57] mvo: hey, do you have any time to talk about bug #876298? [17:57] Launchpad bug 876298 in update-notifier (Ubuntu) "[MASTER] We need to better handle external payloads (Flash, msttcorefonts) not being available." [Critical,Triaged] https://launchpad.net/bugs/876298 [18:03] slangasek: #962378 is making me cry [18:03] bug #962378 [18:03] Launchpad bug 962378 in ca-certificates-java (Ubuntu) "ca-certificates-java tries to call "java" binary before it's installed" [Undecided,New] https://launchpad.net/bugs/962378 [18:05] mdeslaur: that's not a new issue, the circ dep has been there for quite some time IIRC - what's making you cry now? [18:06] slangasek: I'm trying to fix bug #920758 [18:06] Launchpad bug 920758 in ca-certificates-java (Ubuntu Precise) "DigiNotar Root CA still present in ca-certificates-java" [Undecided,Confirmed] https://launchpad.net/bugs/920758 [18:07] and the circ dep is getting in the way of fixing, or in the way of testing the fix? [18:08] hrm, actually...just testing, so I'll work around it for now [18:08] bdmurray: can we block new reports of bug #929219 please? launchpad is starting to creak [18:08] Launchpad bug 929219 in eglibc (Ubuntu) "chromium-browser crashed with SIGSEGV in __nscd_get_mapping()" [High,Triaged] https://launchpad.net/bugs/929219 [18:10] bdmurray: the duplicate signature should check for that __nscd_get_mapping() call in any package, not just chromium-browser or eglibc; I just duped bug #929437 to it, which was reported via gvfs+software-center [18:10] Launchpad bug 929219 in eglibc (Ubuntu) "duplicate for #929437 chromium-browser crashed with SIGSEGV in __nscd_get_mapping()" [High,Triaged] https://launchpad.net/bugs/929219 [18:11] mdeslaur: right - we may be able to improve handling of the circular dep, but I'd like doko to weigh in and he's out today [18:12] slangasek: ok, thanks [18:14] slangasek: nscd_get_mapping only appears in the retraced stacktrace which isn't patternable. do you have another suggestion? [18:14] bdmurray: heh, darn [18:14] bdmurray: gethostbyname2_r() + the eglibc version may be sufficient? [18:15] bdmurray: yeah, that should be enough to identify it [18:16] libc6 version >= 2.15~, <= 2.16-0ubuntu6 [18:16] sorry, 2.15-0ubuntu6 [18:38] cjwatson: can you confirm if the perf regression is fixed in 1.0.1? nehalem, sandy bridge and ivy bridge all have aesni support [18:38] so it's not a perf issue on those processors. [18:39] the regression is on 64bit processors without aesni [18:49] jibel: would it be possible to re-trigger precise-upgrade-lucid-universe ? [18:49] jibel: as long as it fails, it "only" takes 1:45 hours [18:49] jibel: but we got that seed fix today which hopefully helps [18:49] jibel: (seed as in the package, for the libseed0 -> libseed-gtk3-0 transition and the gir1.0 mess) [18:51] pitti, it's in the queue and will start after lucid server. ETA 3min 9sec according to jenkins :) [18:51] jibel: thanks! [18:52] pitti: what was the fix for seed, OOI? [18:52] ah, pushing libseed0 out manually, ok [18:53] slangasek: we made libseed-gtk3-0 conflicts/replaces libseed0, to force the removal of libseed0 instead of trying to fulfill its dependencies in precise [18:53] it seems apt did not clean it up by itself [18:53] yeah... it's a shame to need to do that [18:53] but certainly quicker than figuring out why apt isn't happy to remove the lib :/ [18:53] well, I looked at that apt log for over an hour, and this seems to be the root cause [18:54] I guess if I do that ten times more, I'll write a script to graphvizise apt.log :) [18:54] yeah, it's the root package cause certainly [18:55] but the *root* cause is apt keeping an obsolete lib when it should throw it away :) [18:56] slangasek: sure, the bits in comment #22 ? [18:56] mvo: well, and the merge proposal which I marked you a reviewer on :) === dendrobates is now known as dendro-afk [18:57] slangasek: rigth, I was off yesterday and crazy day today, but let me scan over it [18:59] jhojho: my results are in the bug === dendro-afk is now known as dendrobates [18:59] jhojho: based on the assertions I've seen that there was some deliberate slowing down of AES to defeat cache timing attacks, I'm reluctant to assume that slower is necessarily a showstopper here [19:00] i see the bug report but no actual timings. [19:00] slangasek: yeah, workarounds'r'us [19:00] cjwatson: that makes no sense as processors with aesni exhibit no slowdown [19:00] I posted timings from a Core 2 Duo and a Nehalem to the FFe bug [19:01] slangasek: I will reply in the bug and in the MP [19:01] Anyway, I'm not going to change it further before beta 2 at this point unless there's some kind of serious regression from 1.0.0g, so there should be plenty of time to collect data [19:01] ah I see the timings now. let me go check them out. [19:02] I'm afraid I didn't post comparative timings from lucid; I can do that later if necessary [19:02] on the "one step at a time" principle I wanted to get 1.0.1 sorted out first [19:02] anyway, it's thoroughly pub time now [19:02] please if you could. [19:03] there is one thing I would ask for on the amd64 build [19:03] its the inclusion of the config flag that turns things on for the google contributed elliptic curve code [19:04] well, let's talk with the Debian maintainer about that; 1.0.1-2ubuntu1 very significantly reduces the delta against Debian and I'm pretty happy about that [19:04] (makes it easier to maintain) [19:04] for whatever reason, they made things so that you have to explicitly pass that flag. [19:04] ok [19:04] that's "enable-ec_nistp_64_gcc_128"? [19:05] yes === dendrobates is now known as dendro-afk [19:13] actually i guess i should have asked that over here instead… [19:13] hrmm. is apport retracer blocked/disalbed for amd64 bugs? [19:18] slangasek: so the bugpatterns are regular expression would 2.15~pre work? [19:20] ah no also need one for 2.15-0ubuntu6 [19:21] bdmurray: as a regexp, I'd do '2\.15\(~pre\|-0ubuntu[1-6]\)' [19:22] cjwatson: it does look like things are a tad faster between 1.0.0 vs 1.0.1 [19:34] @pilot in === udevbot changed the topic of #ubuntu-devel to: Precise Beta-1 Released. Archive: open | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bdmurray, bryceh [19:35] cjwatson: just installed to 1.0.1 and it looks like Lucid is still faster. here's the pastebin for lucid and precise (2 results, 1.0.0 and 1.0.1) http://pastebin.com/FG6mfabw [19:40] jhojho: wouldn't that be dependent on the kernel instead? [19:40] why? [19:40] this is a userspace test. it has nothing to do with the kernel [19:41] the fact the aes numbers dont change here either make me think its using the kernel aes stuff but i really have no clue [19:42] sha1 rc4 md5 all change drastically [19:44] Hi all! [19:48] slangasek: I followed up in the MP [19:48] mvo: seen, thanks [20:22] How does the app-install-data-ubuntu menu-data get generated? === fishor_ is now known as fishor [20:47] slangasek: sni-qt> *chuckle* [20:47] slangasek: I'm glad that the mystery is resolved [20:47] pitti: :/ [20:47] pitti: I feel bad now for wasting people's time [20:47] [ubuntu/precise] lightdm-gtk-greeter 1.1.4-0ubuntu1 (Accepted) [20:47] \o/ [20:47] and am running debsums -s now :P [20:47] thanks micahg and mr_pouit ! [20:54] pitti: What, no love for the guy who put all the effort into pressing the "accept" button? [21:00] infinity: not unless you can prove it was you who pressed it ;) [21:01] slangasek: There's a bug for that. :P [21:01] slangasek: And I can prove I filed it! === salem_ is now known as _salem [21:10] I'm sure someone will pick it up eventually, but if a sponsor is lurking, I could use one for https://bugs.launchpad.net/ubuntu/+source/landscape-client/+bug/962448 [21:10] Launchpad bug 962448 in landscape-client (Ubuntu) "update package to 12.04.2" [Undecided,New] [21:35] msg chanserv topic #ubuntu-devel Precise Beta 2 Freeze in Effect. Archive: pre-release freeze | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bdmurr [21:35] ay, bryceh [21:35] skaet, yes? [21:36] * Daviey paases skaet a / :) [21:36] * skaet accepts the / from Daviey [21:36] :) [21:36] * maco wonders if the linebreak indicates it wont all fit in /topic either [21:36] heh, max topic length? [21:36] suck it and see. [21:36] * maco eyebrow [21:37] it looks ok to me... [21:37] yeah ok here too [21:37] i dont see the topic as having changed... [21:37] topic seem to be longer than the line limit. interesting. [21:37] Missing two letters off of the patch pilot's nick here. [21:39] * skaet edited out the redundant info about the Archive, Beta 2 Freeze in Effect pretty much summarizes it. === Daviey changed the topic of #ubuntu-devel to: Precise Beta 2 Freeze in Effect. Archive: pre-release freeze | Development of Ubuntu (not support, not app development) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bdmurray, [21:39] ScottK, better? [21:39] skaet: wasnt bryceh also supposed to be a patch pilot? [21:39] I only see bdmurray, [21:40] Getting there. === Daviey changed the topic of #ubuntu-devel to: Precise Beta 2 Freeze in Effect. Archive: pre-release freeze | Dev' of Ubuntu (not support or app devel) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bdmurray, bryceh [21:40] \o/ [21:40] There we go. [21:40] Thanks Daviey [21:41] sigh... [21:41] bryceh was on the last version i thought... oh well. sorted now. Thanks ScottK, maco. :) [21:42] You're welcome. [21:46] having it go over length is probably a sign the topic needs some cleanup ;-) === mwhudson_ is now known as mwhudson [23:11] so um, why is libvte suddenly complaining about a missing /etc/termcap? [23:12] (and as a result, failing to correctly handle PgUp/PgDn and others) [23:13] gnome-terminal not restarted for 18 days; not sure if it needs restarted, or if trying to restart it is the worst idea ever [23:19] slangasek: I see no such whining from my xfce4-terminals (which also use vte) [23:19] slangasek: I say restart gnome-terminal, and if it all goes south, we'll see you from your console? ;) [23:38] @pilot out === udevbot changed the topic of #ubuntu-devel to: Precise Beta 2 Freeze in Effect. Archive: pre-release freeze | Dev' of Ubuntu (not support or app devel) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bryceh === PaulW2U_ is now known as G4MBY [23:39] @pilot out === udevbot changed the topic of #ubuntu-devel to: Precise Beta 2 Freeze in Effect. Archive: pre-release freeze | Dev' of Ubuntu (not support or app devel) | build failures -> http://bit.ly/xmGdCW | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: