=== cyphermox_ is now known as cyphermox
=== doko_ is now known as doko
[11:05] <jibel> Are alternate images ready for publication on the tracker ?
[11:12] <cjwatson> I think we have a bit of work yet before that's the case
[11:13] <cjwatson> at a minimum they'll need rebuilt for apt 0.8.14.1ubuntu5
[11:16] <jibel> cjwatson, thanks
[11:28] <cjwatson> jibel: and we really ought to have a new ubiquity (waiting for bug 790547)
[11:28] <ubot4`> Launchpad bug 790547 in python-xklavier (Ubuntu) "[MIR] python-xklavier (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/790547
[11:36] <Laney> is it easier for me to say 'please sync everything you can from http://people.canonical.com/~ubuntu-archive/transitions/ghc.html', or to file bugs? :-)
[11:38] <Laney> I mean just the 'bad' packages
[11:38] <cjwatson> easier to file bugs if you don't mind, then our tools help
[11:38] <Laney> yep
[11:38] <cjwatson> don't worry too much about the descriptions though
[11:39] <Laney> doing then
[11:39] <Laney> I remember giving james_w the input to the sync script once, but I forgot the syntax now
[11:47] <Laney> sync spam ahoy
[11:49] <cjwatson> ok, processing
[11:49] <Laney> ty
[11:50] <Laney> 3 more left that are blocked on other people
[11:52] <pitti> whoops, seems I left this channel since my last server reboot, sorry
[12:12] <doko> cjwatson: python-xklavier in main now (except for arm)
[12:13] <ev> yay, thanks doko
=== chuck_ is now known as zul
[13:47] <cjwatson> https://wiki.ubuntu.com/IncidentReports/2011-05-31-pam-security-update-breaks-cron
[13:49] <ev> cjwatson: might be worth adding a "please RT" in the future to maximize the spread of such tweets
[13:50] <cjwatson> ev: didn't have room
[13:50] <ev> ah, fair enough
[13:53] <cjwatson> note that this means that changes to oneiric will not be published at the moment
[13:54] <cjwatson> this will probably affect alpha-1 publication until we get it sorted out
[13:54] <cjwatson> mdeslaur: do you need any further assistance for the moment?
[13:55] <mdeslaur> cjwatson: I've prepared updated packages which I am about to test...but my hardy one fails to compile...could you take a look?
[13:55] <cjwatson> sure (modulo sorting out my hardy chroot)
[13:57] <mdeslaur> cjwatson: the issue is the security update introduced an ABI change, a couple of new symbols, which in other updates is acceptable, but in pam, not good
[13:57] <cjwatson> that's the issue that broke cron?
[13:57] <mdeslaur> cjwatson: I've gotten rid of the ABI change, but it doesn't want to link in hardy
[13:57] <cjwatson> ok
[13:58] <cjwatson> debootstrapping now
[13:58] <mdeslaur> cjwatson: yes, cron has libpam loaded already, and a dynamically loaded pam module from the new version can't find the symbols in the old one cron has loaded
[13:58] <cjwatson> that makes sens
[13:58] <cjwatson> e
[13:58] <mdeslaur> I'm kicking myself for not thinking that that could happen
[13:59] <cjwatson> cf. https://blueprints.launchpad.net/ubuntu/+spec/foundations-o-pam-restarts
[13:59] <cjwatson> (sorry, I don't have a local hardy mirror any more so a little slow)
[13:59] <mdeslaur> cjwatson: so, look at the change to Linux-PAM/libpam/Makefile.am in my debdiff, that's the new private lib
[14:00] <mdeslaur> cjwatson: I _very_ much appreciate your help
[14:01] <cjwatson> you said you couldn't get in touch with Kees; have you tried Rick?
[14:01] <cjwatson> both west coast though
[14:02] <mdeslaur> cjwatson: jamie is my manager, he is on vacation today, and Rick doesn't appear online yet
[14:02] <cjwatson> or Matt for that matter
[14:02] <cjwatson> oh, sorry, manager not tech lead.  I knew that *headdesk*
[14:02] <cjwatson> I'd notify mdz if I were you
[14:03] <cjwatson> standard rules, escalate immediately until you find a victim :)
[14:03] <mdeslaur> cjwatson: yes, thanks for that, I just pinged him
[14:07]  * cjwatson has never had to go all the way up to Jane ... yet
[14:07] <mdeslaur> cjwatson: stop scaring me :P
[14:10] <doko> the incident report doesn't mention any distro series
[14:10] <cjwatson> feel free to edit that in
[14:10] <mdeslaur> I'll add it
[14:13] <ev> anyone else ending up with the initramfs bailing with no log?
[14:14] <jibel> ev, yes, i'm investigating.
[14:14] <jibel> ev, I can't reproduce on i386, I'm installing a fresh amd64
[14:14] <cjwatson> try debug=
[14:15] <cjwatson> or just 'debug' and then look in /dev/.initramfs/initramfs.debug
[14:16] <ev> will do
[14:16] <ev> (was unaware of that trick)
[14:21] <ev> ah, can't find the root device (either by uuid or by block device)
[14:21]  * ev digs
[14:27] <jibel> the problem comes from udev 171
[14:28] <jibel> pitti, ^
[14:30] <pitti> jibel: hi
[14:31] <pitti> ev: you get a boot failure with udev 171?
[14:31] <pitti> ev: what kind of boot device do you have?
[14:31] <pitti> I only have a fairly standard system here, ext4 /, oneiric du jour
[14:31] <jibel> pitti, hi, yes, reproducible on a fresh installation of Oneiric in VM.
[14:32] <ev> pitti: just confirming that udev 170 fixes it
[14:32] <jibel> pitti, just upgrade udev, libgudev and libudev
[14:33] <pitti> ev: to confirm, if you keep libudev/libgudev at 171 and only downgrade udev, does it still work?
[14:34] <ev> pitti: testing exactly that now
[14:37] <jibel> pitti, still broken with udev at 170 and libs at 171
[14:37] <cjwatson> /usr/bin/ld: cannot find -lpamprivs
[14:37] <cjwatson> mdeslaur: is that the same failure you're referring to?
[14:37] <pitti> jibel: ok, as expected; thanks!
[14:37] <mdeslaur> cjwatson: yes
[14:37] <jibel> :)
[14:37] <pitti> I'll start an install in a VM now and see what happens
[14:38] <cjwatson> mdeslaur: can you pass me a working debdiff for some other release?  I'd like to compare
[14:38] <mdeslaur> cjwatson: sure, one sec
[14:39] <mdeslaur> cjwatson: pam_1.1.1-2ubuntu5.3.debdiff in ~/mdeslaur
[14:40] <pitti> jibel: OOI, how do you test in kvm? the current dailies don't really get me to a working desktop or ubiquity
[14:40] <pitti> jibel: with the server isos?
[14:40] <jibel> pitti, alternate
[14:40] <pitti> ah
[14:41]  * pitti goes to download that the
[14:41] <pitti> ...n
[14:41] <ev> yeah, for what it's worth, downgrading all the udev packages fixes it for me too.  Thanks jibel!
[14:44] <cjwatson> cocoplum -> archive/ports syncing re-enabled now
[15:01] <cjwatson> mdeslaur: (still working on this, but I have a call now)
[15:01] <mdeslaur> cjwatson: thanks
[15:38] <ev> doko: if you have time for one more, I'd appreciate a look at MIR bug 790725
[15:38] <ubot4`> Launchpad bug 790725 in pyflakes (Ubuntu) "[MIR] pyflakes (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/790725
[15:47] <mdeslaur> cjwatson: any progress?
[15:48] <cjwatson> sorry, still otp
[15:50] <pitti> FTR, 20110532 alternate amd64 installs just fine in kvm; caveat: need to select "GNOME classic" as a session, unity-2d is crashing
[15:50] <pitti> erm, 20110531 obviously
[15:54] <mdeslaur> cjwatson: ah, I think I got it
[15:55] <doko> ev: done
[15:55] <ev> doko: nice, thanks!
[15:56] <cjwatson> mdeslaur: oh, good, I was stuck in the guts of libtool
[15:56] <jdstrand> mdeslaur: I'm here now. is there something I can do to help?
[15:57] <mdeslaur> jdstrand: ah! hello
[15:57] <mdeslaur> jdstrand: sorry for doing this on your day off :)
[15:57] <mdeslaur> jdstrand: I think everything is under control
[15:57] <mdeslaur> jdstrand: packages for lucid, maverick, natty are built, and I'm satisfied that they solve the issue
[15:57] <mdeslaur> jdstrand: I'm putting the final touch on the hardy package which I'll upload in a few minutes
[15:58] <mdeslaur> jdstrand: once I'm satisfied, I'll publish a -2
[15:58] <cjwatson> mdeslaur: what did you do, just hit it with -Wl?  I'd got as far as noticing that libtool wasn't passing -L through to gcc, but hadn't figured out the proper fix
[15:58] <cjwatson> and my libtool is rusty
[15:58] <mdeslaur> cjwatson: I'll let you know in a sec if my crowbar worked :P
[16:00] <jdstrand> mdeslaur: thanks. I have read all backscroll. do you need help with testing? with usn publication?
[16:01] <mdeslaur> jdstrand: I'm ok at this point
[16:01] <mdeslaur> jdstrand: thanks for the offer!
[16:01] <jdstrand> mdeslaur: ack
[16:02] <jdstrand> cjwatson: I know mdeslaur already did, but I'd like to also thank you for all the help :)
[16:02] <mdeslaur> jdstrand: cjwatson rocks!
[16:02] <jdstrand> yes, he does :)
[16:03] <cjwatson> np
[16:03] <cjwatson> http://www.welikeballs.com/2010/04/misue-of-language-pt-372-rock-star.html
[16:04] <mdeslaur> lol :)
[16:05] <jdstrand> hehe
[16:06] <skaet> :)
[16:07] <mdeslaur> cjwatson: crowbar:
[16:07] <mdeslaur> -	-L$(top_builddir)/libpam -lpam
[16:07] <mdeslaur> +	-L$(top_builddir)/libpam $(top_builddir)/libpam/libpamprivs.a -lpam
[16:13] <cjwatson> OK, that was one of the crowbars I was considering
[16:13] <cjwatson> cool
[16:13] <cjwatson> Laney: ooh, big pile of libghc6-* just fell out of NBS
[16:14] <Laney> \o/
[16:15] <Laney> are you still doing new-source runs? there's a lot of new stuff being uploaded to Debian currently
[16:15] <cjwatson> from time to time, yes
[16:15] <Laney> seems clint is becoming a haskell fan
[16:15]  * cjwatson processes 'new-source | grep haskell'
[16:16] <Laney> but for the transition, gitit is getting some new depends
[16:16] <cjwatson> ah, that would be actually important then
[16:16] <cjwatson> flushed a load through for you now; let me know if you care about stuff that isn't haskell-*
[16:17] <Laney> thanks
[16:17] <Laney> not aware of anything else currently :-)
[16:17]  * Laney hopes joeyh follows through on his ikiwiki-haskell threat
[16:23] <jdstrand> fyi, I took the liberty of updating https://wiki.ubuntu.com/IncidentReports/2011-05-31-pam-security-update-breaks-cron a little
[16:27] <cjwatson> that's fine, I didn't go back through all the history
[16:29] <cjwatson> mdeslaur: give me a shout when it's published enough to be worth updating ubuntustatus
[16:30] <mdeslaur> cjwatson: sure, I've got about a half-hour of testing left to do, and then I'll let you know and release the updates
[16:36] <cjwatson> mdeslaur: does any of this affect oneiric (for alpha 1)?
[16:36] <jdstrand> mdeslaur: so, pam got new symbols which prevented cron from working (module is unknown)
[16:37] <jdstrand> mdeslaur: question-- people who do *not* have auto updates should have been prompted for the services restart, correct? it is only people who have auto updates (and those that chose not to restart) that are affected. correct?
[16:37] <mdeslaur> jdstrand: nobody should have been prompted for the services restart
[16:38] <jdstrand> huh, I thought that was something I always saw when doing an upgrade (at least, do-release-upgrade)
[16:38] <mdeslaur> jdstrand: yes, pam has that in the postinst conditional on certain versions, but the update didn't update that and trigger it
[16:39] <jdstrand> ah, I see
[16:39] <mdeslaur> jdstrand: so, desktop users won't get an update-manager popup until they reboot, server users who installed it won't get automatic updates until they either reboot, restart cron, or do an apt-get dist-upgrade
[17:02] <jdstrand> mdeslaur, cjwatson: since things seem to be well in hand (thanks again!) I am going back to my day off. I am available by phone/email in case you need me. I've asked kees to keep an eye on things. We can do the postmortem tomorrow/later this week
[17:02] <jdstrand> cjwatson: that last part was more for mdeslaur than you :)
[17:04] <cjwatson> mkay
[17:07] <ev> ubiquity 2.7.2 uploaded which should hopefully, finally fix the ftbfs issues
[17:09] <skaet> cjwatson,  was talking to pitti, and it looks like udev is going to need some bisecting.   Any downsides of falling back to last version of udev for the alpha-1 images?
[17:09] <pitti> no, we don't really depend on 171 right now
[17:13] <cjwatson> skaet: how would we do that?
[17:13]  * cjwatson doesn't like rolling back libraries
[17:14] <cjwatson> although at least libudev0's shlibs aren't strict
[17:17] <skaet> cjwatson,  not sure how else to work around the fact that udev is failing for some, and not for other,  and based on the backscroll - prior version seems to make it stable.   Recommendations?
[17:18] <cjwatson> pitti still seems to be debugging on #ubuntu-devel - we have a little while yet
[17:19] <pitti> I'm preparing/testing a revert right now
[17:19] <pitti> and ask people to bisect in parallel
[17:19] <pitti> being able to reproduce would of course help :-(
[17:20] <cjwatson> would it help to have CD builds with 171?
[17:20] <pitti> cjwatson: it would be interesting to see whether it makes a difference
[17:20] <pitti> cjwatson: but having fresh desktops would be interesting in its own right, now that apt is fixed
[17:20] <pitti> udev won't hit the archive for another 100 minutes, 'nuff time to build a set in between?
[17:20] <cjwatson> I've kicked some off on x86, let's see what happens
[17:21] <pitti> I asked in #udev, kay is not aware of problems with 171, so we need to debug this ourselves
[17:21] <pitti> they have a vastly different initrd/systemd/etc, of course
[17:33] <pitti> ok, reverted udev builds, installs, upgrades, boots
[17:35] <pitti> uploaded
[17:40] <pitti> I bumped build score on amd64, but it's still a bit crowded; so amd64 will likely miss the next publisher
[17:42] <pitti> oh, building now
[17:43] <skaet> :)
[17:44] <pitti> ok, amd64 done, i386 well into the build, so they should hit the archive in 70 mins
[17:46] <skaet> thanks pitti
[17:47] <jibel> udev 171-0ubuntu2 is OK on amd64
[17:48] <ScottK> It looks like all the Kubuntu LiveFS builds failed last night.  All but one due exclusively to problems fetching packages files.  Is this a known problem or is there something I need to do about it?
[17:48] <pitti> jibel: *phew*, thanks muchly for bearing with me!
[17:48] <ev> yet another ubiquity making its way through
[17:48] <pitti> good enough for a1
[17:49] <cjwatson> ScottK: known, fixed in latest apt
[17:49] <ScottK> cjwatson: Thanks.
[17:49] <cjwatson> it broke all flavours
[17:50] <ScottK> Did you do respins after the fix?
[17:50] <cjwatson> no
[17:50] <cjwatson> there were other things in progress anyway ...
[17:50] <ScottK> OK.  Still are, so I guess no point.
[17:50] <cjwatson> I'm respinning Ubuntu right now, but that's just in case it helps with debugging udev
[17:51] <mdeslaur> cjwatson: I've just released the updated packages
[17:51] <ScottK> We've got quite a number of uploads in progress at the moment, so I suspect it'll be about the time the next spin happens anyway when stuff calms down.
[17:52] <cjwatson> mdeslaur: OK, are all the binaries in place?
[17:54] <pitti> good night everyone!
[17:57] <mdeslaur> cjwatson: at the next publisher run they will be
[18:37] <mdeslaur> cjwatson: publisher has published the binaries
[18:43] <cjwatson> mdeslaur: great, thanks.  want to close the main task on that bug?
[18:46]  * cjwatson dents
[18:49] <mdeslaur> thanks cjwatson
[19:05] <cjwatson> would anyone like to figure out what's up with all the KDE stuff on http://people.canonical.com/~ubuntu-archive/testing/oneiric_probs.html?
[19:08] <ScottK> Sure
[19:09] <cjwatson> looks like an argument between kdepim-runtime and akonadi at first glance, but I only looked at one of the uninstallables
[19:13] <ScottK> That wouldn't explain all of it.
[19:13] <ScottK> -pim runtime was just uploaded though.
[19:16] <ScottK> I'm at a conference for $WORK this week and I think we just killed the hotel Internet when everyone signed on the wifi.
[19:16] <ScottK> Not nearly the service lamont gives here.
[19:17] <ScottK> (so it may take awhile)
[19:20] <lamont> ScottK: I haven't had much to do with the UDS wifi layout in forever
[19:20] <ScottK> lamont: OK.  Canonical IS then.
[19:21] <lamont> +1
[19:49] <cjwatson> pitti: new CDs built
[19:50] <pitti> cjwatson: ah, thanks; with the new udev 171-0ubuntu2?
[19:51] <slangasek> cjwatson, pitti: I've seen lots of mail about image build failures due to not finding /var/lib/apt/lists, but no discussion about it in scrollback; is this problem understood / already solved, or is it something I should dig into for alpha-1?
[19:51] <cjwatson> pitti: no, 0ubuntu1
[19:51] <pitti> slangasek: yes, it was a regression in apt, fixed today
[19:51] <cjwatson> pitti: deliberately, so that you can see if the problems people were reporting are reproducible with that
[19:51] <slangasek> ah, ok
[19:51] <pitti> cjwatson: right, thanks
[19:52]  * pitti -> off again, see you tomorrow!
[19:53] <slangasek> 'night, pitti :)
[20:06] <slangasek> cjwatson: skaet says you're working on fixing a ubiquity problem yet; anything else I need to know about image status for a1?
[20:06] <slangasek> skaet: should there be an email reminder of the a1 freeze, btw?
[20:06] <skaet> slangasek,  was sent, but appears blocked waiting for an approver to ubuntu-devel-announce ;)
[20:07] <slangasek> hmm!
[20:10] <slangasek> skaet: message accepted
[20:10] <skaet> lol
[20:10] <skaet> thanks!
[20:10] <cjwatson> slangasek: that's the only thing I *know* of right now
[20:10] <ScottK> cjwatson: I think I figured it out (KDE packages).  Solution will take a bit of work though.
[20:10] <cjwatson> oh, wait, I need to do a d-i rebuild to pick up a netcfg fix from earlier
[20:11]  * cjwatson nods
[20:15] <cjwatson> well, I can at least reproduce the ubiquity failure in pbuilder
[20:17] <cjwatson> new d-i uploaded
=== yofel_ is now known as yofel
[20:36] <charlie-tca> cjwatson: I heard I owe more than a beer now. Thanks for helping us out.
[20:36] <cjwatson> np, but don't say that until you see a successful build :P
[20:52] <cjwatson> OK, I see the ubiquity problem
[20:57] <skaet> cjwatson,  thats good - will it be a straight forward fix?
[20:57] <cjwatson> one-liner
[20:57] <skaet> :D
[20:57] <cjwatson> uploading now
[20:57] <skaet> excellent.  :)
[20:59] <cjwatson> http://bazaar.launchpad.net/~ubuntu-installer/ubiquity/trunk/revision/4740
[21:02] <cjwatson> beer o'clock, back later
[21:04] <skaet> single character even...  enjoy the beer.
[21:48] <jibel> udev problem is reproducible with latest desktop iso: dropped into a busybox shell after installation and reboot.
[21:48] <cjwatson> pitti: ^-
[21:49] <cjwatson> we can have another build in a bit with current udev, now that ubiquity has built
[22:48] <cjwatson> slangasek: I'm done here - ubiquity is published on amd64/i386, so you can probably start spinning images at this point and see what breaks
[22:49] <cjwatson> ubiquity/armel is accepted, ubiquity/powerpc is building and looks nearly finished
[22:49] <cjwatson> (though powerpc in general is behind at the moment)
[22:49] <cjwatson> I don't think I have any other pending status
[22:50] <slangasek> ok
[22:50] <slangasek> will take it from here :)
[22:52] <cjwatson> thanks!
[23:03] <ScottK> slangasek: No rush on any Kubuntu images.  Still beating stuff into shape.
[23:09] <slangasek> ack