[00:00] jdstrand: I think he meant cu2d [00:00] jdstrand: I mean, are there integration tests for this before it gets uploaded to the archive [00:00] infinity: I'll let you and tyhicks discuss it [00:01] slangasek: there is a testsuite in the build. there haven't been integration tests due to lack of running click packages in the testing. that is changing and lool found the problem with webapps because of those tests [00:01] slangasek: If it went through cu2d, it would have been a sync, not a sourceful upload. [00:01] infinity: the dbus changes fix a few subtle bugs in the apparmor mediation code that I recently discovered [00:01] infinity: you're assuming I looked that closely ;) [00:01] slangasek: the AP test runs on the stack will pull it, but from the archive, so too late [00:01] lool: right [00:01] slangasek: clue is that if it's a sync then queuediff doesn't work [00:02] I'm assuming you at least looked at the diff briefly ... :) [00:02] tyhicks: *nod*... If Jamie's happy with it, upload away. I'll give it a quick twice-over in the queue. [00:02] cjwatson: idem [00:02] ;) [00:02] cjwatson: I really didn't, I pushed it through on the grounds that phone is not supposed to be frozen [00:02] ah [00:03] Oh, bugger. I was going to update gdb to 7.6.1 for the tasty, tasty bugfixes, but zumbi rolls the Debian tarballs by hand, and I don't want to desync. [00:03] (Among other things, it fixes the Most Annoying GDB Bug Ever: https://sourceware.org/bugzilla/show_bug.cgi?id=15415) [00:03] sourceware.org bug 15415 in gdb "gdb resolves symbolic links when passing argv[0]" [Normal,Resolved: fixed] [00:04] cherry-pick that one? [00:05] jdstrand: do you need anything else from me prior to sponsoring the dbus upload? [00:06] cjwatson: Yeah, but 7.6.1 is all bugfixes, probably makes sense to pull it anyway. I'll see if zumbi's cut a tarball for it (or, indeed, why he does his own at all). [00:08] tyhicks: no-- I uploaded it. just answer infinity's questions [00:08] infinity: when doing your twice-over for the dbus upload, I listed the tests that I wrote and ran for this upload here: https://bugs.launchpad.net/ubuntu/+source/dbus/+bug/1229280/comments/1 [00:08] Launchpad bug 1229280 in dbus (Ubuntu) "Eavesdroppers confined with AppArmor can see all method_return and error messages" [High,In progress] [00:08] * jdstrand steps away for some food [00:08] thanks jdstrand [00:08] tyhicks: thank you! :) [00:08] slangasek and infinity: and thank you for the reviews [00:08] * jdstrand -> food [00:12] tyhicks: Eww, 1233895 is an icky bug. [00:14] infinity: it is, but jdstrand handled the policy updates in apparmor-easyprof-ubuntu and that was the only policy affected [00:15] infinity: we were lucky that there isn't really any complex dbus policies outside of apparmor-easyprof-ubuntu [00:16] cjwatson: so I *think* lp:~vorlon/ubuntu/saucy/grub2/lp.1236625 is a right fix... this seems like it's important enough to get in for release IMHO [00:16] tyhicks: So, dropping the compat bit. What will that mean for people who upgrade from older releases? [00:16] infinity: nothing - saucy is the first release to have apparmor mediation extended into userspace [00:17] tyhicks: Sure, but when the userspace is upgraded and the kernel isn't, will things explode somehow? [00:17] infinity: only if you're still on a dev release kernel from about a month or two ago [00:18] hey. i just did a upload of cloud-init :-(. wish I did not have to. changes to last accepted version are very minimal. if someone could accept that would be wonderful to get it into a build tonight to fully test it. [00:19] tyhicks: Okay, so if you're on raring's 3.8, it'll all just no-op cleanly, and if you're on a current saucy kernel, it's good, but if you're in between, the world goes sideways? [00:20] infinity: sorry, my connection dropped for a few minutes after I said "only if you're still on a dev release kernel from about a month or two ago" [00:21] 18:19 < infinity> tyhicks: Okay, so if you're on raring's 3.8, it'll all just no-op cleanly, and if you're on a current saucy kernel, it's good, but if you're in between, the world goes sideways? [00:22] infinity: yes, but the "in between" is only a couple saucy dev release kernels [00:22] tyhicks: Or, another way of phrasing that: "if I'm upgrading online from raring to saucy, and dbus gets restarted in this new version on my raring kernel, did you just hose my system until I reboot?" :P [00:23] tyhicks: Kay. If it's just a few dev kernels that explode, I don't care. [00:23] infinity: dbus-daemon looks for /sys/kernel/security/apparmor/features/dbus, which doesn't exist when running a raring kernel [00:24] Kay, cool. [00:24] Acceptiferated, then. [00:24] if that directory isn't there, then the apparmor mediation hooks are disabled [00:24] infinity: thanks! (and good questions ;) [00:27] slangasek: Thanks, staged locally and will think about it during my daytime (since it isn't trivial) [00:27] slangasek: Would you mind renaming the patch to ubuntu_something.patch? [00:27] I prefer to keep a stack of ubuntu_* at the top [00:51] is this in the right state https://bugs.launchpad.net/ubuntu/+source/software-properties/+bug/1233486 . i'm hoping for (i realize last minute) FFe. [00:51] Launchpad bug 1233486 in software-properties (Ubuntu) "[FFe] add support for 'cloud-archive:' like 'ppa:' but for cloud archive" [Medium,New] [01:14] would be nice to get a britney run for dbus to migrate; it seems all autopkgtests ran now [01:15] It'll get one, there's stuff publishing now [01:16] thanks [01:20] dbus valid candidate -- cool [01:21] and accepted, more to the point :) [01:22] right :-) [01:23] doko_: Something went sideways with your contol.m4 magic. All the cross packages have a broken dep on "dnlcpp-4.8-$triplet" [01:44] cjwatson: ok, renamed and pushed [02:08] * slangasek looks blankly at launchpad. mountall 2.51 accepted? but er, that was uploaded weeks ago [02:09] oh, arm64 build \o/ [02:09] * slangasek scratches his head, noticing only now that it's called arm64, not aarch64 [02:10] You're too late if you want to object. :) [02:10] slangasek: It's arm64 in the kernel too. Linus had the same objection as I did to the silly name. [02:10] slangasek: Did you get a mail about it or something? [02:10] cjwatson: I did [02:10] (Though, I appreciate the GNU name being oddball, because it doesn't match arm*) cases) [02:10] Uh [02:10] infinity: yeah, at least if it's consistent with the kernel [02:10] cjwatson: Yeah, people get fresh accepted mails when britney does the copy. [02:11] Oh, this was a build in PROPOSED [02:11] Right, fine (ish) [02:11] :) [02:11] Won't apply to the bulk of builds in the RELEASE pocket [02:11] A bit confusing, since it's a full copy of the source again, and the mail has no indication that it's because a new arch caught up. [02:11] But oh well. [02:11] And the mail is kind of half-arsed IIRC. [02:11] It's missing the changelog bit. [02:11] Which is such an indication - but only if you know WTF it means [02:12] slangasek: Well, it's consistent with the kernel source tree, not with uname. Confused yet? :) [02:12] infinity: uname == aardvarch? [02:12] Anyway, arm64 is building real stuff now, scored to try to supersede the bootstrap archive first [02:12] slangasek: The Debian arch arm64 builds the GNU arch aarch64-linux-gnu on the kernel that reports aarch64 to uname -m, where the source is in arch/arm64. [02:13] slangasek: NOT CONFUSING AT ALL. ;) [02:13] It's not remotely in the right order so loads of stuff will depwait [02:13] Tore through lapack fairly well. [02:14] ~Twice the speed of akateko/kishi07, ~half the speed of allspice/sagari. [02:15] But akateko probably doesn't count; I bet it was spending ages on docs or something. [02:15] Yeah. [02:15] The build log is certainly huger. [02:15] Lotta docs. [02:20] cjwatson: Hah, the one-line response to your well-thought-out gnulib/as-needed analysis is great. [02:21] infinity: I think that's Paul's way of saying "please send a patch for your preferred option" [02:22] (Paul is the star of my favourite Uli story, so I'm inclined to give him the benefit of the doubt) [02:22] cjwatson: Or "can't type; cat in face". [02:24] That story being essentially http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=59829, especially http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=59829#55 [02:24] Debian bug 59829 in libc6 "libc6: [PATCH] fnmatch() behaves oddly with *s and FNM_LEADING_DIR" [Fixed,Open] [02:24] Uli demands bug report from somebody he's heard of: Paul says OK and copies and pastes my bug report verbatim [06:15] Laney, sure, it's ok for UbuntuKylin. === rbasak_ is now known as rbasak [09:08] cjwatson: Trade you reviews. [09:09] infinity: Yep [09:10] infinity: blt is basically for arm64 too although it's less urgent [09:10] cjwatson: You could have skipped changing the build-dep if you used --with autotools-dev instead. [09:10] I could've. [09:10] (I intensely fear autoreconf unless knowingly patching configure.in and testing the result is sane, but to each their own) [09:10] I do use that sometimes when it's significantly easier (not dhed yet, or autoreconf blows up) [09:11] I test-built in this case [09:11] * infinity nods. [09:12] (indicator-power was uploaded to fix a settings:/// link for touch but is seeded on desktop; it was tested by sil2100 successfully) [09:15] lool: I'll look at it. [09:15] lool: mirv, that is :) [09:16] it's a one char change on the url_dispatcher's url that's used on touch. [09:16] Mirv_: Developers are interchangeable parts. Sufficient time in management teaches this. [09:17] Mirv_: oh sorry [09:17] Mirv_: braing bug [09:17] so, the change is one char change in url string that is used on touch only (on_phone_settings_activated) [09:17] infinity: except you of course, you're unique! [09:18] lool: I'm as unique as a snowflake tattoo on a hispter's neck. [09:18] thank you === davmor2_ is now known as davmor2 === peterm-ubuntu is now known as Peter-meetign === Peter-meetign is now known as Peter-meeting [12:16] cjwatson: ^ [12:19] I think it's okay to upload a minor fix for a wrong dependency for packages in universe? [12:20] happyaron: You think correctly. [12:20] thanks === Mirv_ is now known as Mirv [12:25] USE_LATEX_LOL_DOLPHIN_HASHTAG_YOLO [12:25] blink [12:27] cjwatson: don't blink that's when the weeping angels get you :D [12:28] I went to Edinburgh shortly after watching Blink the first time [12:28] CITY WITH GARGOYLES EVERYWHERE [12:29] It was suboptimal [12:36] cjwatson, ohhhhh that would be bad === yofel_ is now known as yofel [12:41] cjwatson, hi, could you help to take a little time reviewing the merge request of ubiquity for ubuntukylin? We just modify one line to change the background pic name. [12:42] maclin: The weird bit for me was the claim that this was a "name rule" for Ubuntu [12:42] It's not [12:42] The reason Ubuntu is warty-final-ubuntu.png is basically that we made a mistake back in the 4.10 days and need to stick with it to preserve upgrades [12:43] I don't understand why UbuntuKylin feels it needs to mimic this mistake [12:43] ubuntukylin-default-settings.png is actually a better name for it, rather than having a bogus series name in there [12:44] If you think this is a "name rule", then somebody has got the wrong end of the stick at some point [12:45] oh, we just change it according to the name of ubuntu, we really don't know the mistake :P [12:45] cjwatson: http://www.youtube.com/watch?v=FwK4eAhLGYg&feature=player_detailpage#t=18 [12:46] cjwatson: (Reference for the USE_LATEX frustration) [12:46] maclin: I think you're making a mistake in trying to match Ubuntu here, and should revert whatever else you've done rather than changing ubiquity. [12:46] Unless that's impossible at this point. [12:48] cjwatson, yes, if it's true, we will change it back in default-settings. thanks:) [12:50] maclin: Thanks. Sorry, I think I may have reviewed another one of your changes and not noticed this; should have raised this earlier. [13:10] cjwatson: maclin: next cycle I think i'll purge hard-coded names, and simply use gsettings (xfce-settings, etc) to figure out the default wallpaper and purge that duplication. [13:13] xnox, +1, it's a great idea to identify the name out of ubiquity codes. [14:21] can I upload gnupg and gnupg2 security updates? [14:28] mdeslaur: Yes [14:29] thanks cjwatson === doko_ is now known as doko === charles_ is now known as charles [15:17] I'm reviewing the rest of the queue but would be nice if someone could take that LXC upload [15:21] did someone review it that quickly, or was that an inadvertent autoaccept? [15:34] slangasek: can't have been auto-accept, it's in a package set and seeded on a bunch of images [15:41] stgraber: well, that's why I said "inadvertent" :-) [15:42] we can't get seem to work out a fix for qtwebkit-source on arm, since the current version also doesn't work on arm is it ok to force it? [15:43] No, don't force it [15:44] If the current version doesn't work and can't be made to work, perhaps it should be removed. Does it have rdepends? [15:44] ... a ton [15:44] We can't release this way ... [15:46] I wonder how the previous version got in [15:47] Oh. You forced it. Don't do that [15:47] sorry [15:47] Yeah, that was the thing I warned at the time would cause problems later. [15:48] Uh oh [15:49] Question: Should ~ubuntu-sru review requests go here or #ubuntu-devel? [15:49] The indication is not clear... [15:50] Riddell: So I really think we need to fix this by release rather than continuing to ignore it; we can't ship out-of-date binaries. What progress has been made? [15:53] cjwatson: shadeslayer had a fix disabling 3d stuff which I'm trying to look up, but he ran into another problem [15:58] building now with DEFINES+=WTF_USE_3D_GRAPHICS=0 on my pandaboard, let's see what result I get [15:58] result will probably be tomorrow [16:18] stgraber: why libchewing sync is rejected? [16:18] it was me [16:18] because there were two [16:18] ah [16:18] oh no, someone rejected the new one [16:18] ignore that [16:20] Laney: do I need to sync it again? [16:20] you need to find out why it got rejected [16:20] did you get an email? [16:20] no, no accept/reject [16:22] Wouldn't be surprised if they are busted for syncs, sadly [16:25] what to do then? [16:26] Wait for someone to own up [16:30] happyaron: I gave a reject reason when I rejected one of them. The diff is 1MB large, there are significant code change, a mile long changelog and at lease one new symbol added [16:30] happyaron: that's a bit much when we're past FeatureFreeze and that close to release [16:31] I think the reject emails don't work for syncs [16:33] Laney: ah, that'd explain it [16:34] https://bugs.launchpad.net/launchpad/+bug/830614 [16:34] Launchpad bug 830614 in Launchpad itself "Email not immediately sent for copied packages which end up in NEW" [Low,Triaged] [16:34] Not certain that's the same thing [16:35] (It might be, but would have to analyse) [16:36] Can someone help to decide on https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/1234887 please? [16:36] Launchpad bug 1234887 in network-manager (Ubuntu) "[FFE] Update NetworkManager to 0.9.8.4" [Wishlist,Confirmed] [16:36] Lot of changes this late on but a lot of fixes === rtg is now known as rtg-afk === rtg-afk is now known as rtg === psivaa is now known as psivaa-afk === rtg is now known as rtg-afk === psivaa-afk is now known as psivaa [17:22] cjwatson, Laney: please could you approve gdb and elfutils (to fix arm64) === rtg-afk is now known as rtg [18:21] hi! [18:22] can someone review apparmor-easyprof-ubuntu? it is seeded in supported but only on touch [18:22] two small fixes-- one for mir access and one for the upcoming kernel update === Ursinha is now known as Ursinha-afk [18:25] jdstrand: I'll take a look [18:26] question: how come the lastest kernel isn't making it to the CD Images? [18:26] stgraber: thanks! [18:26] if you look at http://cdimage.ubuntu.com/ubuntu-server/daily/pending/saucy-server-amd64.list you'll see that the kernel is linux-generic_3.11.0.11.12_amd64.deb but the latest is linux-tools-3.11.0-11-generic [18:29] probably because either d-i or the seeds haven't been updated [18:29] stgraber: who usually handles the seed updates for that? [18:30] infinity [18:30] ah, thanks [18:32] utlemming: actually, I don't see anything wrong with that kernel version [18:34] stgraber: yeah, you're right, I was looking at the wrong package...I was waiting for .17, but it looks like .12 is getting served up [18:35] utlemming: current linux-image-generic is 3.11.0.11.12 and current linux-image-3.11.0-11-generic is 3.11.0-11.17 [18:36] utlemming: the linux-image-generic one comes from the linux-meta source, the linux-image-x.y.z-* come from the linux source [18:36] so the first usually only changes when the ABI is bumped [18:37] stgraber: ack, okay that makes sense [18:37] stgraber: thanks for humoring my question [18:44] stgraber: mind also taking a look at apparmor upload? [18:44] stgraber: tyhicks has the details [18:44] * jdstrand uploaded just a moment ago [18:45] stgraber: my two things in the upload are simple policy updates. tyhicks has the details on the patches (and a couple other policy updates) [18:46] jdstrand: I may also need you to chime in on 0072-lp1229393.patch, which is the patch to fix the cache location for the .features file [18:46] that's fine [18:47] tyhicks: but I didn't actually change the initscript or anything else, so it is really just your patch [18:47] well, jj's patch, but yeah. I can comment [18:47] ok [18:48] I want to redo our policy load, but not for 13.10 [18:54] stgraber: not sure if you saw my ping re apparmor? [18:54] stgraber: s/ping/request/ [18:54] stgraber: just in case you didn't: http://paste.ubuntu.com/6210598/ [18:55] meh, I chopped one off [18:55] stgraber: http://paste.ubuntu.com/6210600/ [18:56] jdstrand: I accepted it a couple of minutes ago [18:56] * stgraber kicks queuebot [18:56] stgraber: you accepted apparmor too? (not just apparmor-easyprof-ubuntu?) [18:57] ah yes [18:57] stgraber: thanks! :) [18:57] * jdstrand hadn't gotten the email yet [18:58] thanks! === thomi_ is now known as thomi [19:42] infinity, slangasek, bdmurray, sru team: could anyone review the libreoffice upload to precise that's waiting since august? it's a small diff with backported patches, not an updated version... [20:05] re-ping about the NM bugfix upload ^^^ [20:18] I don't see that u* in the queue have any actual changes [20:22] 18:53 < stgraber> dobey: so is that waste of buildd time really necessary? [20:22] 18:54 < dobey> yes [20:22] I didn't get any better reason out of him, so I decided not to care and keep them in the queue for someone who cares more to deal with [20:23] (since my other alternative was a simple reject and I didn't feel like arguing for hours) [20:24] sigh [20:36] infinity: around to look at that linux upload or shall I? === Ursinha-afk is now known as Ursinha [21:00] maybe SRUers would be happy putting NM in that way to get it some extra testing [21:02] stgraber: I am now, but it looks like you got to it. :P [21:04] infinity: yep, pretty small diff this time around [21:16] Hello again - if someone could review evince, I'd appreciate it [21:16] It is a small patch that only grants more permissions to the evince apparmor profiles [21:17] most importantly, if grants access to the accessibility bus [21:17] it already got accepted [21:17] there's no need to ping right away :-) [21:17] ah, thanks! :) === _bjf is now known as bjf[alt] [22:22] ^- at the head of a build-dep chain on arm64 - can somebody have a look? [22:26] Drat. The aarch64 WIP port for elfutils not only doesn't quite work yet on arm64, it also breaks the x86 build. [22:27] Now to decide how deeply I care about helping to fix that this week, or if I should just disable the failing tests on arm64 and scream "la la la, elfutils doesn't actually work with aarch64 binaries but I don't care, la la la". [22:29] advantage of doing this kind of thing: I get to take touched-it-last on merges off people who were doing drive-by changes and clearly haven't cared for multiple releases [22:29] disadvantage: I'll have to do those merges next cycle ;-) [22:29] cjwatson: You already had TIL on opemipmi. :P [22:30] I meant my next one [22:30] ... twice. [22:31] How's the graphviz cycle-breaking going? [22:32] Sleep got in the way. It's first on my list today. [22:33] Well, it was second on my list after elfutils, but that's a black hole of doom I'll revisit after I've had a drink and found my Ballmer Peak. [23:13] can someone please reject ubuntu-touch-session-0.77 when it comes up? [23:13] I uploaded with a .orig + .diff but it's native [23:13] lool: If it's native, where did you find an orig? :P [23:14] infinity: bzr bd was too smart [23:14] infinity: ^ [23:14] quick before the * cron! [23:14] Rejected. [23:14] thanks! [23:14] Gah. [23:14] No. [23:14] The bot beat me. [23:14] I'm disappointed [23:14] Upload a 0.78. [23:14] yeah [23:18] reuploaded after fixing bzr bd config in trunk to native = True [23:46] infinity: libquvi's for arm64