== Meeting information == * #ubuntu-meeting Meeting, 04 Sep at 15:04 — 15:30 UTC * Full logs at [[http://ubottu.com/meetingology/logs/ubuntu-meeting/2013/ubuntu-meeting.2013-09-04-15.04.log.html]] == Meeting summary == === lightning round === The discussion about "lightning round" started at 15:05. * ''LINK:'' https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/833994/comments/8 === AOB === The discussion about "AOB" started at 15:27. == Vote results == == Action items == * (none) == People present (lines said) == * stgraber (35) * slangasek (29) * ubottu (28) * ev (26) * cjwatson (24) * stokachu (21) * xnox (19) * bdmurray (14) * doko (13) * barry (6) * meetingology (3) == Full Log == 15:04 #startmeeting 15:04 Meeting started Wed Sep 4 15:04:39 2013 UTC. The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 15:04 15:04 Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired 15:05 $ echo $(shuf -e barry doko stgraber jodh ev bdmurray slangasek cjwatson xnox stokachu) 15:05 doko ev jodh stgraber slangasek cjwatson stokachu barry bdmurray xnox 15:05 [TOPIC] lightning round 15:05 doko: you're first :) 15:06 - Some unexpected GCC work, looking at regressions in the testsuite: 15:06 - Testsuite wasn't enabled, so it was difficult to track down :-/ 15:06 - Fix libjava test failures, seen with binutils 2.23 15:06 - Fix boehm-gc test failure 15:06 - Look at gcc and g++ test failures only seen with hardening defaults, and fix 15:06 - Re-discovered that -Wformat-security is disabled in GCC 4.8, because the boo 15:06 tstrap failed with this default. Now enable that in the specs instead of the op 15:06 tions processing, and it builds ... 15:06 - Ignore the additional mudflap test failures. Will work on removing mudflap u 15:06 pstream alltogether now that the sanitizer is available in 4.8. 15:06 - Upload is pending 15:06 - Restart on cmake cross builds. 15:06 (done) 15:07 - Tail end of vUDS. Had a very productive discussion around the next steps for 15:07 continuous integration. 15:07 - Mostly spent the week getting ready for the Lexington QA sprint next week. 15:07 - Calls with jibel, Martin, Francis, and Adam. 15:07 - Meeting with IS to discuss Cassandra cluster next steps. They've asked to 15:07 delay the final steps of migrating the data from our existing clusters into a 15:07 new one. This is because they're working on building Prodstack 4.0 and doing 15:07 these steps now would mean having to move a lot of data around once the new 15:07 Prodstack is ready. I've in turn asked that more attention be given to moving 15:07 daisy, errors, and the retracers onto prodstack and opening up access to the 15:07 Cassandra nodes again, so we can at least continue with feature development 15:07 while we wait. 15:07 - Patch piloting. Woo. 15:07 - Discussion around reporting runaway applications. 15:07 Other: 15:07 - Note: I am in Boston from this Sunday for the QA sprint and then on holiday 15:07 for two weeks after that in somewhat remote territory (Nevada and Arizona 15:07 desert). If you need my help with something, we have until Friday :) 15:07 - Attended a talk by Adrian Cockcroft at Cassandra London on how they have 15:07 built Netflix. A lot of it was a repeat of the Cassandra SF talk that I 15:07 summarised on canonical-tech, but there were definitely some useful bits in 15:07 there, like their Zuul API proxy tier. Didn't get a chance to ask him how the 15:07 move of all their systems to Ubuntu was going :-/ 15:07 (done) 15:08 Blueprint-related work: 15:08 - Image based updates (BLUEPRINT: foundations-1305-image-based-updates) 15:08 - Landed the daily-proposed channel 15:08 - Re-numbered all the current channels (base version is 1 instead of YYYYMM00) 15:08 - Bumped system partition size from 1.2GB to 2GB 15:08 - Following-up with QA to get this landed by default this week 15:08 (jodh is on holiday) 15:08 Other work: 15:08 - LXC 15:08 - Linux Plumbers mini-conference preparation 15:08 - Licensing review and fixes 15:08 - Python attach API review and fixes 15:09 - Some code review 15:09 - Android port fixes (building again with current NDK) 15:09 - Prepared and pushed out the LXC 1.0~alpha1 pull request 15:09 (FFe already accepted for once it's released) 15:09 - Network 15:09 - Merged ifupdown from Debian. Thanks to Steve for fixing the conffile handling of /etc/init.d/networking! 15:09 - Other 15:09 - vUDS 15:09 15:09 TODO (hopefully this week): 15:09 - Land LXC 1.0~alpha1 in the archive ASAP (need to nag Daniel to process my pull request) 15:09 - Implement the boot-time hooks for Ubuntu touch 15:09 - Help QA run tests against touch_ro so we can switch to it by default 15:09 - Re-design channels.json to be flexible enough to support channel aliases and hidden channels 15:09 - Blog/g+/email about the version number changes as soon as we get a new phablet-flash landed in the archive 15:09 (DONE) 15:10 stgraber: what do you think about backporting ifupdown to precise? 15:11 stokachu: a straight backport wouldn't work because of other changes that happened around it, so we'd have to compare the effort to cherry-pick the fixes you care about vs the work required to backport saucy to precise. 15:11 stgraber: (android src package rebuild has landed in .deb package as well now, if still needed) 15:11 stgraber: ok ive got a bug on ifupdown in precise where using simple service networking restart fails 15:12 stgraber: you have time after the meeting to discuss it a bit? 15:12 xnox: thanks! I didn't actually need it since we aren't using the binaries from the android package yet, but that means that once we do, we won't regress (which was the goal here). 15:13 * short week, Labor Day was on Monday 15:13 * revisiting the status of UEFI netboot support; currently arguing with a machine that only wants to netboot over IPv6 15:13 stokachu: I don't see how backporting a more recent version of ifupdown would solve that 15:13 * finally getting the partition resizing work done for phablet-flash, though too late to make it for the initial switch to system images 15:13 * poking around the edges of beta1, want to revisit the freeze rules for optional milestones based on some of the experiences here 15:13 * patch piloting today 15:13 (done) 15:13 stokachu: considering we don't actually support restarting networking to begin with 15:14 stgraber: dont support restarting networking? 15:14 stokachu: right 15:14 even for server 15:14 stokachu: we support bringing individual interfaces up and down, but as the networking in Ubuntu is event based, we can't support restarting the whole thing sequentially 15:14 * ev steps out to chat with IS 15:15 have you tried `sudo service networking restart`? don't! ;) 15:15 cjwatson: 15:15 Launchpad: 15:15 - Landed build cancellation (bug 54730) and retrieval of build logs for cancelled builds (bug 1050514). 15:15 stokachu: I mean, it may work and for simple systems it'll, just don't expect it to really work when using bridges, vlans or bond interfaces 15:15 stgraber: ok good to know, thanks 15:15 bug 54730 in launchpad-buildd "launchpad buildd abort does not work as expected" [High,Fix released] https://launchpad.net/bugs/54730 15:15 - Implemented automatic ppa-reset of failing virtual builders (bug 1039935), which has made the PPA build farm significantly more stable. 15:15 bug 1050514 in Launchpad itself "Cancelling build loses log file" [High,Fix released] https://launchpad.net/bugs/1050514 15:15 bug 1039935 in Launchpad itself "buildd-manager failure counting is too trigger-happy when killing virtual builders" [High,Fix released] https://launchpad.net/bugs/1039935 15:15 - Wrote blog entry on recent work: http://blog.launchpad.net/general/launchpad-build-farm-improvements 15:15 - Removed limits on the percentage of available builders that a given public PPA can take up, which now do more harm than good (bug 666308). Will be in the next deployment. 15:15 bug 666308 in Launchpad itself "public non-virtual ppas are unnecessarily limited on builders per architecture" [High,Fix committed] https://launchpad.net/bugs/666308 15:15 Click: 15:15 - Added "icon" (optional) and "installed-size" (automatically generated) to manifest for the benefit of UI code. 15:15 - Cleaned up various bugs in the preloaded wrapper. 15:15 - Added support for multiple installation root directories, which we'll use for preinstalled apps. Much internal heavy lifting to make this work but no incompatible interface changes. 15:15 TODO: 15:15 - Finish deploying preinstalled app handling in click. 15:15 - Removal support in click, since UI work on this is due to start soon. 15:15 - Big pile of OEM bugs, probably involving all the coffee :-/ 15:15 .. 15:15 barry: haha, yeah, then you notice that dbus is "stop on deconfiguring-interfaces" taking down the whole system :) 15:16 bug 833994 ϟ needs discussion on latest comments 15:16 bug 1157943 ϟ Needs more prodding of David (apt upstream) 15:16 bug 1204507 ϟ Verification completed, needs promotion out of -proposed queue. 15:16 bug 833994 in debian-installer-utils (Ubuntu) "debian-installer does not support https when using with preseed files" [Medium,Triaged] https://launchpad.net/bugs/833994 15:16 bug 1179781 ϟ Needs apt uploaded in order to test. 15:16 bug 1157943 in apt (Ubuntu Precise) "apt-get update fails hash checks on https repositories when file size changes" [Undecided,New] https://launchpad.net/bugs/1157943 15:16 (done) 15:16 bug 1204507 in maas (Ubuntu Quantal) "MAAS rejects empty files" [Undecided,Confirmed] https://launchpad.net/bugs/1204507 15:16 bug 1160490 ϟ uploaded to saucy, needs sponsorship in precise, quantal, and raring. 15:16 bug 1179781 in apt (Ubuntu Raring) "If-Modfied-Since undhandled case causes apt lists corruption with https repositories" [Medium,Triaged] https://launchpad.net/bugs/1179781 15:16 bug 1160490 in ifupdown (Ubuntu Raring) "race condition updating statefile" [Medium,In progress] https://launchpad.net/bugs/1160490 15:16 stgraber: i think it's shorthand for `ssh deadmachine sudo reboot` 15:17 Short week due to USA holiday. vUDS. 15:17 system-image: LP: #1208909; LP: #1218612. fixed traceback when -c gives a bad channel name. system-image 1.3 and 1.4. various meetings. update blueprints. 15:17 Launchpad bug 1208909 in Ubuntu system image "Add support for forcing full image update" [High,Fix released] https://launchpad.net/bugs/1208909 15:17 Launchpad bug 1218612 in Ubuntu system image "Support new version number scheme" [Critical,Fix released] https://launchpad.net/bugs/1218612 15:17 This week: LP: #1196991 15:17 Launchpad bug 1196991 in Ubuntu system image "Support the new download dbus service" [High,Triaged] https://launchpad.net/bugs/1196991 15:17 done 15:17 uploaded apport SRU (R) fix for bug 1168849 15:17 uploaded ubuntu-release-upgrader SRU (Q and R) fix for bug 1215526, bug 1210649 15:17 bug 1168849 in apport (Ubuntu Raring) "crashes while reporting a Synaptics bug - fills up /tmp" [High,Fix committed] https://launchpad.net/bugs/1168849 15:17 bug 1215526 in ubuntu-release-upgrader (Ubuntu Raring) "/usr/lib/ubuntu-release-upgrader/check-new-release-gtk:ImportError:new_dist_available:build_ui" [Undecided,Fix committed] https://launchpad.net/bugs/1215526 15:17 uploaded fixed SRU (Q and R) for bug 1035136 15:17 bug 1210649 in ubuntu-release-upgrader (Ubuntu) "/usr/lib/ubuntu-release-upgrader/check-new-release-gtk:TypeError:on_button_upgrade_now_clicked:run:error:error:__init__:__init__:function" [High,Triaged] https://launchpad.net/bugs/1210649 15:17 bug 1035136 in lsb (Ubuntu Raring) "[SRU] Fix lsb-core for 12.10 and 13.04" [Medium,Fix committed] https://launchpad.net/bugs/1035136 15:17 uploaded oneconf fix for bug 1051935 15:17 bug 1051935 in ubiquity (Ubuntu Saucy) "Fails with SystemError when too many files are open" [Undecided,Confirmed] https://launchpad.net/bugs/1051935 15:17 sponsored patches in bug 1191704 15:17 review of unsubscribed packages 15:17 fix of tz issue with comparing dates in sru-report 15:17 investigation into bug 1209442 (not phased update issue) 15:17 bug 1191704 in heimdal (Ubuntu Raring) "KDCs complain about not having enough file handles for /var/lib/heimdal-kdc/heimdal" [High,Fix committed] https://launchpad.net/bugs/1191704 15:17 bug 1209442 in Unity "compiz crashed with SIGSEGV in getCompPluginVTable20090315_unityshell()" [High,Triaged] https://launchpad.net/bugs/1209442 15:17 * slangasek buys some Bosnian coffee from Brazil at the European deli and drop-ships it to cjwatson 15:17 stokachu: 1204507> blocked on a slew of other verifications, as shown on http://people.canonical.com/~ubuntu-archive/pending-sru.html 15:17 fwiw 15:17 investigation into, fixing bug 1211511 15:17 reported errors bug 1217521 15:17 reported update-notifier bug 1220435 regarding crash reporting from Live CDs 15:17 bug 1211511 in update-manager (Ubuntu) "update-manager hides but wants to install ignored phased updates" [High,Triaged] https://launchpad.net/bugs/1211511 15:17 bug 1217521 in Errors "Include repository information in package versions table on a problem page" [Wishlist,New] https://launchpad.net/bugs/1217521 15:17 cjwatson: thanks checking now 15:17 research into apport bug 1098844 15:17 bug 1220435 in update-notifier (Ubuntu) "not notified of a crash when running a saucy live cd" [High,Triaged] https://launchpad.net/bugs/1220435 15:17 bug 1098844 in apport (Ubuntu Raring) "apport-gtk crashed with PermissionError in thread_collect_info(): [Errno 1] Operation not permitted: '/var/crash/openafs-modules-dkms.0.crash'" [Medium,Triaged] https://launchpad.net/bugs/1098844 15:18 tested multiple kernels for bug 1218004 15:18 bug 1218004 in linux (Ubuntu) "Apple Wireless Trackpad causes kernel oops" [High,Confirmed] https://launchpad.net/bugs/1218004 15:18 ✔ done 15:20 * xnox notices it's me 15:20 cjwatson: do you have any thoughts on the latest comments on bug #833994 (from stokachu)? 15:20 bug 833994 in debian-installer-utils (Ubuntu) "debian-installer does not support https when using with preseed files" [Medium,Triaged] https://launchpad.net/bugs/833994 15:21 * ev back 15:21 imho it's unwise to introduce insecure https, and at the moment using gnu wget fails all installations (due to exact text parsing) so that linked bug should really be fixed first. 15:21 .. 15:21 - sent off summary to promote 64bit images by default on the website 15:21 - finished proof-of-concept cross-compiling qml extensions, send of 15:21 email to ubuntu-phone mailing list. Further changes are required in 15:21 qt* & packaging 15:21 - help out landing (and testing) upstart 1.10 in the archive for FFe, ubiquity 15:21 I remain concerned that providing HTTPS when there's no way to make it secure is inviting people to do it wrong 15:21 uploads for FFe/Beta1, unbreak boost-python linking, unbreak 15:21 dh-python, sponsor patches for a few bugs. 15:21 - continuing work on emulator, emulator config is now using kernel 15:21 from the archive. 15:21 .. 15:21 And yeah, as xnox said 15:21 I really don't have time to take on a fix for this set of things at the moment though :-/ 15:22 if we could possibly get a definitive stance on the subject that would help i think 15:22 * slangasek nods 15:22 posted to the bug publicly 15:22 stokachu: I think cjwatson's / xnox's stance has been pretty consistent / definitive 15:22 "we're not going to give users a false sense of security about https" 15:22 stokachu: bug 1172101 must be fixed. 15:22 bug 1172101 in wget (Ubuntu) "wget-udeb should install to /usr/bin/wget instead of /usr/bin/wget.gnu" [Low,Triaged] https://launchpad.net/bugs/1172101 15:23 This is it, my stance has been definitive, just not popular 15:23 stokachu: then we can further proceed with remaining https cert issues. 15:23 And I did propose a solution in comment #8 15:23 ok 15:23 stokachu: and then yeah, implement cjwatson's comment #8 15:23 (Which certainly needs implementation from our end) 15:23 https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/833994/comments/8 15:23 Launchpad bug 833994 in debian-installer-utils (Ubuntu) "debian-installer does not support https when using with preseed files" [Medium,Triaged] 15:24 tb is technically right about links that are sniffable but not pwnable, but that basically requires a webserver that does PFS, and very few of them actually do and most admins have no idea if theirs does or not 15:24 so the vast majority of people who would see this functionality and try to use it would probably still be doing it wrong 15:25 ok -- i can pass this along and do we want to leave this case as is for a potentional resolution from comment 8 down the road? 15:25 it's certainly still open 15:27 ok 15:27 yeah, and supporting gnu wget is also useful (with & without https) 15:27 any other topics for today? 15:27 [TOPIC] AOB 15:27 ? 15:27 just confirmation on the rest of those bugs 15:27 theres a few low hanging fruit ones i believe 15:28 stokachu: bug #1179781> I'll sponsor those today 15:28 bug 1179781 in apt (Ubuntu Raring) "If-Modfied-Since undhandled case causes apt lists corruption with https repositories" [Medium,Triaged] https://launchpad.net/bugs/1179781 15:28 slangasek: thanks! 15:29 (chiluk probably should have resubbed ubuntu-sponsors for the reupload, but that's ok) 15:29 slangasek: ah i missed that sorry! 15:29 nothing else? 15:30 going... 15:30 going... 15:30 #endmeeting Generated by MeetBot 0.1.5 (http://wiki.ubuntu.com/meetingology)