[09:49] <CIA-3> debian-installer: ogra * r1278 ubuntu/ (3 files in 3 dirs): fix omap udeb selection for netboot flavour, drop gzip compression in omap mkimage call for netboot flavour, make sure we build INITRD_FS = initramfs for omap netboot
[09:51] <CIA-3> debian-installer: ogra * r1279 ubuntu/debian/changelog: releasing version 20081029ubuntu97
[10:26] <CIA-3> ubiquity: evand * r4066 ubiquity/ (debian/changelog ubiquity/components/ubi-timezone.py): Catch invalid iterators in on_region_combo_changed (LP: #521851).
[10:49] <CIA-3> ubiquity: evand * r3925 testing/tests/run-frontend: Add test for moving backwards past the manual partitioning page.
[11:12] <saispo> cjwatson: hi, have you have a roadmap about the latest kernel integration in the debian-installer ?
[11:13] <cjwatson> what's to integrate?
[11:16] <saispo> just move kernel 2.6.24-27 proposed to stable for creatingcd
[11:16] <saispo> just move kernel 2.6.24-27 proposed to stable for creating cd
[11:17] <CIA-3> choose-mirror: cjwatson * r623 ubuntu/choose-mirror.c: consistent indentation
[11:17] <cjwatson> why didn't you say that instead of "latest kernel integration"? :-)
[11:17] <cjwatson> not very specific
[11:17] <saispo> yes :) excuse me
[11:18] <cjwatson> anyway, I already uploaded it, looks like it's just awaiting the end of the standard 7-day aging period
[11:18] <saispo> ok
[11:18] <cjwatson> not my call, I don't push my own uploads through to -updates since that's generally bad form
[11:18] <saispo> i see your upload and will wait, it's just for having a date
[11:19] <saispo> i'm not hirry
[11:19] <saispo> hurry
[11:19] <cjwatson> should be this week
[11:19] <saispo> ok, thanks
[11:30] <CIA-3> ubiquity: evand * r4067 ubiquity/ (debian/changelog ubiquity/components/ubi-language.py):
[11:30] <CIA-3> ubiquity: Don't let not being able to talk to the system bus crash the entire
[11:30] <CIA-3> ubiquity: language page.
[11:33] <CIA-3> ubiquity: evand * r3926 testing/ (4 files in 3 dirs): Merge with trunk.
[11:40] <CIA-3> ubiquity: evand * r3927 testing/tests/run-frontend: monkey patch switch_progress_windows as a no-op.
[11:42] <CIA-3> partman-base: cjwatson * r203 ubuntu/ (4 files in 3 dirs):
[11:42] <CIA-3> partman-base: Allow preseeding partman/alignment to "cylinder", "minimal", or
[11:42] <CIA-3> partman-base: "optimal"; "cylinder" restores old alignment behaviour for the benefit
[11:42] <CIA-3> partman-base: of those with crotchety BIOSes, while "optimal" is the default.
[11:44] <CIA-3> partman-base: cjwatson * r204 ubuntu/debian/changelog: releasing version 139ubuntu4
[11:49] <CIA-3> ubiquity: evand * r3928 testing/tests/run-frontend: Add test case for r4066, testNoActiveItemRegionCombo.
[11:52] <amichair> is anyone able to reproduce Bug #556376 ?
[11:53] <amichair> (I have a guess what the cause is by looking at the code, but can't recreate to test)
[11:56] <ev> I was able to reproduce it fairly reliably in KVM.
[12:00] <amichair> ev: how exactly? jr said only in non-english?
[12:00] <amichair> (I'm on a clean virtualbox, amd64)
[12:01] <ev> I booted into "Install Kubuntu", then pressed alt-f until I was on the timezone page
[12:01] <ev> I pressed alt-f again to continue
[12:01] <ev> and then left it for a few minutes
[12:01] <ev> it didn't move to the next page until I moved the mouse
[12:03] <ev> I may have selected a language other than English, I don't recall exactly
[12:06] <amichair> ev: thanks, I'll give it another shot
[12:06] <ev> amichair: cool, let me know if you need me to test any potential fixes
[12:11] <amichair> ev: nope :-(
[12:11] <amichair> alt-f goes all the way through
[12:21] <ogra> what is partman doing that could trigger the following (/me doesnt know what to look for to reproduce it in kernel)
[12:21] <ogra> Apr 12 11:18:47 main-menu[208]: INFO: Menu item 'partman-base' selected
[12:21] <ogra> Apr 12 11:18:48 kernel: [  556.388214] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40200000
[12:21] <ogra> Apr 12 11:18:48 main-menu[208]: (process:7234): Bus error
[12:22] <cjwatson> could be misaligned memory access
[12:23] <cjwatson> /var/log/partman might at least pinpoint which parted_server command was being run when it fell over
[12:23] <ogra> likely, i would like to give something to amitk so he can easily reproduce though ...
[12:23] <ogra> ah
[12:23] <ogra> /bin/partman: *******************************************************
[12:23] <ogra> /lib/partman/init.d/25md-devices: *******************************************************
[12:23] <ogra> /lib/partman/init.d/30parted: *******************************************************
[12:23] <ogra> thats all i have
[12:23] <ogra> rotating due to several tries
[12:24] <ogra> hmm, init.d is a script
[12:26] <cjwatson> stick set -x near the top of /lib/partman/init.d/30parted
[12:26] <ogra> tried that
[12:26] <ogra> but i cant seem to capture the output anywhere
[12:26] <cjwatson> you should have got stuff in /var/log/syslog then
[12:26] <ogra> nope
[12:26] <cjwatson> err
[12:26] <ogra> (note i'm running through serial atm)
[12:26] <cjwatson> always goes there for me
[12:26] <cjwatson> but if that doesn't work, then:
[12:27] <cjwatson> exec 2>/tmp/trace; set -x
[12:27] <ogra> ok
[12:28] <ogra> ~ # ls /tmp/trace
[12:28] <ogra> ls: /tmp/trace: No such file or directory
[12:28] <ogra> *sniff*
[12:29] <cjwatson> err
[12:29] <cjwatson> no idea what went wrong there :)
[12:29] <ogra> i guess it dies to early (in /bin/sh already)
[12:29] <cjwatson> that seems pretty unlikely.  huge swathes of the installer before that are in shell.
[12:29] <cjwatson> where did you put the set -x?
[12:29] <ogra> first i turned set-e into -ex
[12:30] <ogra> then i put the line above below that
[12:30] <ogra> *set -e
[12:30] <cjwatson> make sure you do the exec; set before . /lib/partman/lib/base.sh
[12:30] <ogra> yeah, it is
[12:30] <cjwatson> I don't know what to suggest then ...
[12:30] <ogra> well, amit seems to have a suspicion
[12:31] <ogra> so i'll see
[12:31] <cjwatson> I wouldn't rule out it being a userspace bug yet, we just don't know where
[12:31] <cjwatson> unaligned accesses are a common enough bug on architectures that have strict alignment requirements
[12:31] <cjwatson> it's a bit surprising that we wouldn't have run across it already elsewhere, that's all
 ogra: I am almost sure it has to do with clocks being turned off during module access
[12:32] <cjwatson> ok
[12:32] <ogra> so lets see where he gets to :)
[12:57] <ogra> wow, thats weird
[12:58] <ogra> so i decided to go for a non serial install test and i cant switch consoles
[13:03]  * ogra doesnt get it ... it completely stops working ... like no enter key and the like, but the caps-lock and num-lock keys still work
[13:05] <ogra> oh !
[13:06] <ogra> it switches consoles, it simply doesnt display them ... i have tty1 on the screen no matter where i switch to
[13:07] <ogra> so indeed pressing enter doesnt move the picture on screen :P
[13:07] <ogra> oh my, that kernel is so far from ready :(
[13:31] <CIA-3> debian-installer: ogra * r1280 ubuntu/ (build/config/armel/omap/netboot.cfg debian/changelog): add boot.scr to easily boot OMAP netinst kernel and initrd from SD card
[13:31] <CIA-3> ubiquity: evand * r3929 testing/tests/run: Run tests under xvfb.
[14:09] <ev> cjwatson: the "ellipsis as .*" you mentioned the other day comes standard with the doctest module that launchpad subclasses.
[14:10] <cjwatson> ah right, good
[14:33] <ev> ugh, fakechroot fakeroot deboostrap ... is being entirely uncooperative
[14:34] <ev> (for individual ubiquity component tests)
[14:54] <cjwatson> --variant=fakechroot?
[14:56] <ev> indeed, fakeroot fakechroot debootstrap --variant=fakechroot --arch=i386 lucid /tmp/test-chroot http://archive.ubuntu.com/ubuntu
[14:57] <ev> err, ignore the --arch
[14:57] <ev> I seem to be running into dbo 561991
[14:57] <ev> err bugs.debian.org
[15:40] <cjwatson> superm1: so, I fixed some bugs in the debconf/apt integration you committed to casper a while back that was getting in the way of bug 557011 ;-)
[15:40] <cjwatson> superm1: but the rest of it is proving difficult and I'm having to sit and think about it fairly hard
[15:51] <cjwatson> I *think* that what we need to do is: (a) run casper's debconf-communicate with a read-only templatedb; (b) run apt-get with the same templatedb but read-write (i.e. remove DEBCONF_DB_REPLACE); (c) continue running apt-get with a throwaway read-write configdb which passthroughs to casper/debconf-communicate's real read-write configdb, which is the one that ends up on disk after boot
[15:51] <cjwatson> very delicate though
[16:49] <cjwatson> hmm, well, that sort of worked but created other problems
[16:52] <cjwatson> I think I'm going to need to run casper's debconf-communicate with an entirely separate configdb, and debconf-copydb things over at the end
[16:52] <cjwatson> ugh
[16:52] <ev> ouch
[19:22] <superm1> cjwatson, great on finding those other debconf/apt integration issues from before, thanks :)
[22:07] <amichair> ev: u still there?
[22:07] <ev> amichair: indeed I am, what's up?
[22:07] <amichair> ev: feel like trying out a fix to the mouse freeze issue?
[22:08] <ev> definitely
[22:09] <amichair> oki, so - if necessary, first make sure u can recreate the issue in ur setup (for verification)
[22:09] <amichair> then, slip in and edit kde_ui.py by adding 'self.app.processEvents()' as the first line of debconffilter_done method
[22:09] <amichair> then try to recreate again
[22:10]  * amichair crosses his fingers...
[22:12] <ev> will do
[22:13] <amichair> ev: btw, u familiar with that FIXME bit in this method? it doesn't look right.
[22:36] <ev> amichair: I *think* that did the trick, just confirming.
[22:37]  * amichair uncrosses his fingers and puts on a party hat
[22:45] <amichair> ev: as for the FIXME there, when I comment it out I don't get the crash it claims to solve - any idea what that's about?
[22:52] <ev> amichair: not sure and bzr blame/log isn't much help.  We best leave that as is for lucid though.
[22:53] <amichair> ok, though even if necessary, that doesn't seem like a clean solution (the other two events should be disconnected as well, or better yet, all should be setEnabled(False))
[22:53] <amichair> but best if it can just be deleted :-)
[22:54] <amichair> but indeed, we'll leave it for now
[22:54] <ev> indeed, I'm just trying to avoid risk this late in the cycle :)
[22:54] <amichair> ok, so I'll prepare the mouse fix in my branch for merging
[22:54] <ev> thanks
[22:57] <amichair> btw I noticed the Yes/No buttons in quit confirmation dialog aren't translated (at least to Latvian), and lacking default keyboard focus (try quitting using only keyboard - it's awkward)
[22:57] <amichair> unless Yes/No is valid Latvian, of course. I wouldn't know :-)
[22:57] <ev> hrm, confirmed with Spanish
[22:58] <ev> though yes was focused as soon as I hit alt
[22:59] <amichair> yeah, maybe I did something wrong there.
[23:00] <amichair> Should the default to quit be Yes? isn't No the safer choice, since it's not irreversible?
[23:02] <ev> yeah, No should be the default
[23:03] <amichair> should I open bugs for all these?
[23:04] <ev> please do
[23:06] <amichair> okily dokily
[23:07] <ev> haha, neighboreno
[23:12] <amichair> Bug #561876
[23:16] <amichair> and Bug #561881
[23:16] <ev> 561876> just testing a fix for that now
[23:16] <ev> thanks
[23:16] <amichair> u beat me to it ;-)
[23:18] <amichair> pushed the hang fix to branch at lp:~amichai2/ubiquity/fixes/
[23:18] <amichair> should I find someone else to confirm, or is it safe enough?
[23:19] <amichair> (seeing the event model there is quite intertwined, might be good to play it safe)
[23:27] <ev> amichair: I think it's safe.  Committed.
[23:27] <amichair> ev: ok
[23:28] <amichair> Well, that wraps up my good deed for the day... time to get some rest
[23:28] <cjwatson> superm1: just a reminder re my information request in bug 550694 in case you didn't see it
[23:28] <ev> amichair: enjoy, and thanks again for your help!
[23:28] <amichair> ev: Thanks, and good night (whenever it hits your tz!)
[23:29] <CIA-3> ubiquity: evand * r4069 ubiquity/ (debian/changelog ubiquity/frontend/kde_ui.py): Fix hang unless mouse is moved (LP: #556376)
[23:32] <superm1> cjwatson, yeah i just saw it.  been a busy day, i've been away from bug mail and IRC, i'll try to get it up there before tomorrow
[23:37] <cjwatson> thanks
[23:38] <cjwatson> my network speeds still haven't come back up to what they're pleased to call normal, so I may have to call my ISP again tomorrow
[23:39] <cjwatson> on the upside I think I've finally defeated casper
[23:46] <superm1> that's great news :)