#title #ubuntu-meeting Meeting Meeting started by cjwatson at 15:02:15 UTC. The full logs are available at http://ubottu.com/meetingology/logs/ubuntu-meeting/2011/ubuntu-meeting.2011-09-07-15.02.log.html . == Meeting summary == *lightning round *Bugs *AOB ''LINK:'' https://launchpad.net/ubuntu/+milestone/ubuntu-11.10-beta-2 isn't looking too bad, but obviously needs continuing attention (cjwatson, 15:34:48) Meeting ended at 15:39:17 UTC. == Votes == == Action items == * (none) == People present (lines said) == * cjwatson (59) * bdmurray (30) * stgraber (23) * ubottu (12) * jhunt (6) * doko (5) * barry (4) * meetingology (3) == Full Log == 15:02:15 #startmeeting 15:02:15 Meeting started Wed Sep 7 15:02:15 2011 UTC. The chair is cjwatson. Information about MeetBot at http://wiki.ubuntu.com/AlanBell/mootbot. 15:02:15 15:02:15 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 15:02:32 so I believe that mvo and slangasek are on holiday and ev is at Plumbers 15:03:07 [TOPIC] lightning round 15:03:15 $ echo $(shuf -e cjwatson barry doko stgraber jhunt bdmurray) 15:03:15 jhunt stgraber bdmurray doko barry cjwatson 15:03:18 not rigged, honest 15:03:19 Fix for bug 834813 now in upstream libnih. Finishing tests for Upstart 15:03:21 job logging (written 82 new Upstart tests so far!). 15:03:21 Launchpad bug 834813 in libnih "nih_str_split() can return garbage array elements." [Undecided,Fix committed] https://launchpad.net/bugs/834813 15:03:25 ☕ 15:03:40 argh, missed mumble 15:03:49 - This past week 15:03:49 - Monday was a public holiday (labour day) 15:03:50 - Spent most of last Wednesday and Thursday doing ISO testing and bug fixes for beta1 15:03:52 - Did quite a bit of ubiquity debugging and bug fixing 15:03:55 - Started poking at friendly recovery. 15:03:58 - Got it running from /lib (though /usr will be required for whiptail). 15:04:01 - Still need to add some fallback code for when /usr isn't there. 15:04:03 - Asked James for some ideas on how to start friendly recovery really early in the boot sequence. 15:04:06 - Between his e-mail and our mumble discussion, I have enough ideas to start poking at possible implementations. 15:04:10 - Continued working on IPv6 testing. Got the code to generate all the VMs done, just need to work on preseeding now. 15:04:12 - Gave a session on the ARB at the Ubuntu App Developer week 15:04:15 - Worked on a few bugfixes for Arkose and tried to get it translated 15:04:17 - Started setting up an iSCSI test environment for bug 838809 15:04:19 Launchpad bug 838809 in Ubuntu Oneiric "authenticated and unauthenicated iscsi clients fails to complete boot" [High,Confirmed] https://launchpad.net/bugs/838809 15:04:20 - TODO 15:04:22 - Debug that iSCSI bug 15:04:25 - Get a new friendly-recovery released fixing bug 234409 and bug 575469 (and probably some others) 15:04:28 Launchpad bug 234409 in friendly-recovery (Ubuntu) "friendly-recovery violates the Linux Filesystem Hierarchy Standard" [Medium,Triaged] https://launchpad.net/bugs/234409 15:04:28 - Finish implementing IPv6 testing 15:04:29 Launchpad bug 575469 in friendly-recovery (Ubuntu Oneiric) "recovery mode mounts filesystems read-write rather than read-only" [Medium,Confirmed] https://launchpad.net/bugs/575469 15:04:31 - TPM stuff 15:04:33 (done) 15:04:38 stgraber: in some ways I rather liked Lennart's idea that if /usr is separate then the initramfs should deal with mounting it too; although it potentially has boot time implications 15:04:56 bug triage of foundation's bug reports 15:04:56 iso-testing bug review and triage 15:04:56 usb-creator bug review and testing 15:05:04 UDD querying for standard output apport package install failures 15:05:04 blog post regarding using UDD when Launchpad text searches are inadequate 15:05:07 investigation into ubiquity version tagging in the source package hook 15:05:09 wrote a bug bot function to deal with failures running update-initramfs on LiveMedia which results in package install failures 15:05:13 created a duplicate siganture for dkms driver failures 15:05:14 review of dkms bug report duplicate signatures 15:05:14 email to bug control regarding filtering bugpattern-needed mailing list bug reports 15:05:17 review of and email to dholbach regarding Bugs/Testing progress tracking 15:05:20 replaced power supply in desktop system 15:05:21 hp touchsmart tm2 troubleshooting and debugging (updated BIOS to resolve boot failure) 15:05:24 holidays (half of Thursday, all of Friday and Monday) 15:05:26 done 15:05:52 bdmurray: is your blog post on voices? 15:06:11 barry: no just on planet.ubuntu.com 15:06:24 cjwatson: yeah, I guess /usr is important enough that the initramfs should take care of mounting it, though for the recovery mode, the least file systems we depend on the better (as your /usr partition can be corrupted) 15:06:33 stgraber: yes, true 15:07:23 doko: 15:07:56 - GCC update (should be final for anything but ARM) 15:08:19 - NBS whole week, in need of anti-depressives now 15:08:22 .. 15:08:53 i guess that's my cue... 15:08:57 short week due to usa holiday. bug 832864 (pyside ftbfs); ffe attained and uploaded new versions of apiextractor, generatorrunner, shiboken, and pyside. sbuild hackery fun: . patch pilot, bugs: 770752, 835763, 835763, 835765, 831411, 829078, bug 831411 (zope packages to dhpy2). xorg crash diagnostics. bug 407862 (rsyslog - still investigating, can't reproduce). working on lts ppa for lp. done. 15:09:00 Launchpad bug 832864 in pyside (Ubuntu Oneiric) "pyside version 1.0.4-1 failed to build in oneiric" [High,Fix released] https://launchpad.net/bugs/832864 15:09:01 Launchpad bug 831411 in zope.testbrowser (Ubuntu) "[FFe] dh_python2 transition of zope packages" [Undecided,Fix released] https://launchpad.net/bugs/831411 15:09:18 Quite a bit more work on NBS/FTBFS. 15:09:18 Prepared transitional openssl098 package based on Debian's (but we had a non-trivial set of patches against openssl which needed to be ported, so it took a while). 15:09:23 Uploaded a few grub2 fixes and a Debian merge (bdmurray, one of the things fixed here was bug 537123 - a Debian developer noticed a problem in the previous fix which I think accounts for something you were asking me about at the rally). 15:09:26 Launchpad bug 537123 in grub2 (Ubuntu) "package linux-image-2.6.31-20-generic 2.6.31-20.57 failed to install/upgrade: /etc/grub.d/README executable bit set" [Undecided,Fix released] https://launchpad.net/bugs/537123 15:09:28 Found a strchr bug on ARM, rather to my surprise! (Bug 842258) 15:09:29 Launchpad bug 842258 in eglibc (Ubuntu) "ARM strchr fails to convert c to char" [Undecided,New] https://launchpad.net/bugs/842258 15:09:30 Patch piloted today. 15:09:33 Worked on bug 420080. I think I have this almost done; still needs a little work to ensure that /etc/crypttab is written correctly. 15:09:34 Launchpad bug 420080 in partman-crypto (Ubuntu) "Configure encrypted volumes destroys existing data" [Critical,In progress] https://launchpad.net/bugs/420080 15:09:36 Some work on packaging OpenSSH 5.9. 15:09:38 Got my Debian kilt :-) 15:09:41 done 15:10:19 have to forward the ARM report to linaro ... 15:11:01 [TOPIC] Bugs 15:11:10 bdmurray: anything we should be paying attention to this week? 15:11:46 I was looking at usb-creator bugs a bit as it made the 5 packages 15:12:00 and I ran across bug 826716 which seems rather important 15:12:02 Launchpad bug 826716 in usb-creator (Ubuntu) "error "uncaught exception", "no such file or directory" creating persistent usb key" [High,Triaged] https://launchpad.net/bugs/826716 15:13:07 wibni the oserror included the file name? 15:13:30 I "think" I got that bug at the global jam on Sunday. Ended up just dding the ISO image to the usb stick ;) 15:14:03 bleh, my line numbers don't match 15:15:36 I don't get it, why would mkfs.ext3 not be there? 15:15:43 missing /sbin in PATH maybe? 15:16:05 I guess usb-creator could force /sbin:/usr/sbin onto the PATH 15:16:20 oh, I've personally ran into some /sbin not in path issues 15:16:56 that suggests a problem elsewhere; /sbin is supposed to have been on everyone's default PATH for years now 15:16:59 at least in Ubuntu 15:17:07 usb-creator probably ought not to assume that anyway 15:17:21 I'm pretty tempted to blame something desktopy 15:17:23 right this happened to me somewhat recently on an oneiric system 15:17:44 ps auxewww plus a lot of patience can help to track down the root of such problems 15:18:30 cjwatson: did you mean blame something desktopy for messing up PATH? 15:18:50 yeah, because the thing that opens the PAM session is supposed to deal with this, and that's probably lightdm 15:19:09 I can't see a problem here, but my .bashrc has been adding /sbin and /usr/sbin to PATH for a decade or so which probably masks it 15:19:23 if I ssh to localhost PATH is properly set 15:19:41 ssh does its own PAM handling which gets this right 15:20:03 look for the gnome-session process and see what its PATH is 15:20:09 (in ps auxewww) 15:20:15 The switch to lightdm broke my path at some point. Though I think that got fixed 15:20:29 I'm using gdm on both the systems I just looked at 15:20:31 anyway, I'll volunteer to fix usb-creator to tolerate this 15:20:39 (but at the same time my .bashrc still contains ". /etc/environment" so I wouldn't notice ;)) 15:20:45 assigned/milestoned 15:21:20 There is also bug 818177 which seems to be missing some info from apw 15:21:21 Launchpad bug 818177 in udev (Ubuntu) "HP DL380G5 root disk mounted read-only on boot and boot fails" [High,Confirmed] https://launchpad.net/bugs/818177 15:21:35 not volunteering for that 15:21:41 heh 15:22:01 though I agree with Steve, simply moving the race isn't a solution 15:23:18 Is having it assigned to the team sufficient? 15:23:48 I think somebody needs to take it and discuss with upstream 15:24:43 I started to look into this, but couldn't recreate the prob. 15:25:00 I'll ping apw (at Plumbers I think) re the further info and see what we can find. 15:25:48 mm, being a race it may be difficult to reproduce directly; a reproduction case would involve getting a udev event into the queue, making sure it isn't processed before udevadm control --exit, and then checking that the event is processed somehow 15:26:05 I suppose flooding udev with events just before the control --exit might do it 15:26:23 err - checking that the event is lost, I mean 15:26:47 the udev_exit case in udevd.c does: 15:26:49 /* discard queued events and kill workers */ 15:26:49 event_queue_cleanup(udev, EVENT_QUEUED); 15:26:50 worker_kill(udev, 0); 15:27:15 which is probably fine if udevd has been SIGTERMed, but shouldn't it be a bit more careful in the control --exit case (i.e. graceful exit)? 15:28:38 also I'm surprised that udev in the real root doesn't catch up with the lost events 15:29:33 perhaps we assume that the device for the real root exists by the end of the initramfs 15:30:03 jhunt: anyway, I think you're closest to this; can I assign it to you for the time being, and you can see what you and apw can work out? feel free to ask me for ideas for debugging methods if need be 15:30:36 cjwatson: count on my calling on you at some point :-) 15:31:18 finally there is bug 835625 which steve seems to be watching 15:31:18 yep; it's a messy one, not expecting anything immediately 15:31:20 Launchpad bug 835625 in apt (Ubuntu Oneiric) "apt may try to unpack a foreign-arch multiarch library before the native package is at a multiarch version, prohibited by dpkg" [High,Triaged] https://launchpad.net/bugs/835625 15:31:33 however it is milestoned and unassigned 15:32:12 yes, I think that's best left to Steve; I'll assign him for the time being 15:32:40 okay, that's all I've got 15:33:03 OK, thank you 15:33:18 [TOPIC] AOB 15:33:27 anything else? 15:34:02 I'm guessing pitti is at plumbers too? I could use a second pair of eyes regarding the ubiquity apport package hook 15:34:38 I'm not sure; he has been on IRC today 15:34:48 https://launchpad.net/ubuntu/+milestone/ubuntu-11.10-beta-2 isn't looking too bad, but obviously needs continuing attention 15:39:16 sounds like that's a wrap; thanks all 15:39:17 #endmeeting Generated by MeetBot 0.1.5 (http://wiki.ubuntu.com/AlanBell/mootbot)