[01:58] <ScottK> pitti: https://jenkins.qa.ubuntu.com/view/Vivid/view/AutoPkgTest/job/vivid-adt-marble/lastBuild comes up 404 and I don't see any link at the top level to Vivd results to try and navigate to it.  Is the public instance set up for Vivid results yet?
[02:07] <RAOF> ScottK: I'm pretty sure it is - I've got failure notifications for colord with resolvable public URLs.
[04:16] <ScottK> Weird.
[04:17] <ScottK> https://jenkins.qa.ubuntu.com/view/Vivid/ is 404 for me as well.
[04:44] <hyperair> argh unity-settings-daemon leaks memory so bloody much
[04:52] <darkxst> pitti, https://code.launchpad.net/~darkxst/gnome-session/3.14/+merge/240212
[07:21] <Noskcaj> doko_, Could you please check if enblend-enfuse still needs our arm optimisation changes?
[07:26] <pitti> Good morning
[07:27] <pitti> RAOF, ScottK; /view/Vivid/ doesn't exist yet, I sent an RT last week; the individual jobs do exis (in the failure notification mails), just not the view
[07:27] <pitti> darkxst: \o/
[07:27] <ScottK> pitti: OK.  How do I get to the marble test results?
[07:28] <pitti> ScottK: https://jenkins.qa.ubuntu.com/job/vivid-adt-marble/
[07:28] <pitti> those URLs always work
[07:29] <ScottK> Thanks.
[07:29] <pitti> ScottK: and yes, I retried it once or twice and it seems to genuinely fail
[07:29] <ScottK> pitti: It might be a good idea to teach the excuses page to point at those always existing URL instead of the sometimes not existing URLs it uses now.
[07:30] <pitti> ScottK: oh, indeed; I'll prepare a britney fix
[07:30] <ScottK> Thanks.
[07:30] <pitti> we'll hopefully soon switch away, so I don't want to see the next release on this, but still..
[07:37] <pitti> ScottK: https://code.launchpad.net/~pitti/britney/britney2-ubuntu-jenkins-url/+merge/240221
[07:37] <ScottK> Thanks.
[07:38] <pitti> infinity, cjwatson: can you please merge and roll out https://code.launchpad.net/~pitti/britney/britney2-ubuntu-jenkins-url/+merge/240221 so that notifications have working URLs?
[07:40] <infinity> pitti: Pretty sure you can do that yourself.
[07:40] <infinity> pitti: Unless you were asking for a review. :P
[07:40] <infinity> pitti: Oh, britney is owned by -release not -archive.  Curious.
[07:41] <pitti> infinity: yeah, I can't push there
[07:44] <infinity> Dear bzr, I'm pretty sure that giving up on locks after 177 hours MIGHT BE SAFE TO DO ON YOUR OWN WITHOUT MY HELP.
[07:44] <infinity> Unable to obtain lock  held by cjwatson@bazaar.launchpad.net on taotie (process #24409), acquired 177 hours, 15 minutes ago.
[07:44] <infinity> ^-- Lolz.
[07:45] <infinity> pitti: Rolled out.
[07:46] <pitti> infinity: cheers!
[08:21] <LocutusOfBorg1> seb128, Laney thanks, don't forget to ping me if something else is needed :)
[08:21] <LocutusOfBorg1> and don't forget the vino patch in the bug report ;)
[08:24] <seb128> LocutusOfBorg1, sure
[08:27] <LocutusOfBorg1> and for gdcm I'm rebuilding the rdeps without the two patches, just keeping the arm64 delta
[09:26] <seb128> LocutusOfBorg1, what needs sponsoring?
[09:39] <LocutusOfBorg1> upnpc rdeps? :)
[09:39]  * mgedmin wonders if it would be impolite to draw pitti's attention to https://code.launchpad.net/~mgedmin/apport/fix-lp1387328
[09:40] <pitti> mgedmin: no, not at all; I saw the MP, just have my hands full
[09:40] <pitti> mgedmin: will get to it ASAP
[09:40] <mgedmin> no rush, I won't get to enjoy the fix until vivid anyway
[09:40] <seb128> LocutusOfBorg1, is there things ready for "sponsoring"? your ppa has packages that have older versions that vivid
[09:40] <mgedmin> I just don't want it to be accidentally forgotten
[09:44] <LocutusOfBorg1> seb128, I don't get, do I need to reupload with "ubuntu1" suffix rather than ~ubuntu?
[09:44] <seb128> LocutusOfBorg1, no, just tell me what needs to be sponsored
[09:44] <seb128> LocutusOfBorg1, like your ppa has "0ad (0.0.17-1~ubuntu15.04.1~ppa1)" which is older than the vivid version
[09:45] <seb128> LocutusOfBorg1, does that just need a no-change rebuild?
[09:46] <LocutusOfBorg1> yes, all of them
[09:46] <LocutusOfBorg1> I used backportpackage because I don't know about a tool that has something like "grab-addubuntuN-upload"
[09:47] <LocutusOfBorg1> so all of them uses the library, and only vino has the need of a fix (in the bug report)
[09:47] <seb128> LocutusOfBorg1, k
[09:47] <seb128> LocutusOfBorg1, I already uploaded vino, uploading transmission now
[09:48] <seb128> LocutusOfBorg1, the other ones I'm not sure there is much to "sponsor", if it's only bumping the changelog I can just get/run dch/upload?
[09:48] <LocutusOfBorg1> yep
[09:49] <seb128> thanks
[09:49] <LocutusOfBorg1> seb128, sorry, it is my first transition in ubuntu :)
[09:49] <LocutusOfBorg1> BTW vino picked up the old miniupnpc stuff
[09:50] <seb128> LocutusOfBorg1, no worry!
[09:50] <seb128> LocutusOfBorg1, you mean?
[09:50] <LocutusOfBorg1> yep, failing to build
[09:50] <LocutusOfBorg1> https://launchpadlibrarian.net/188771363/buildlog_ubuntu-vivid-amd64.vino_3.8.1-0ubuntu3_FAILEDTOBUILD.txt.gz
[09:50] <LocutusOfBorg1> so hold on about uploads, maybe the new miniupnpc isn't installed yet?
[09:51] <seb128> right
[09:51] <seb128> but since it stop building with the old one it should have an updated build-depends
[09:51] <infinity> LocutusOfBorg1: FWIW, for no-change rebuilds, "dch -r" usually gives you the sane version number (bumps ubuntuN if modified, adds buildN if not).
[09:51] <seb128> let me fix that
[09:54] <LocutusOfBorg1> seb128, yes, I can update maybe the patch if the older miniupnpc is used, but seems too much useless effort
[09:55] <seb128> right
[09:55] <LocutusOfBorg1> and I'm trying to rebuild gdcm dependencies ;)
[09:56] <cjwatson> dch -R not dch -r :)
[09:56] <infinity> cjwatson: Err, yes.  I always get the wrong one first. ;)
[09:56] <LocutusOfBorg1> still not picking up the new one in my vivid pbuilder
[09:56] <LocutusOfBorg1> thanks both :)
[10:09] <LocutusOfBorg1> seb128, seems to be available now
[10:09] <LocutusOfBorg1> Get:1 http://archive.ubuntu.com/ubuntu/ vivid-proposed/main libminiupnpc10 amd64 1.9.20140610-2ubuntu1 [24.0 kB]
[10:09] <LocutusOfBorg1> Get:2 http://archive.ubuntu.com/ubuntu/ vivid-proposed/main libminiupnpc-dev amd64 1.9.20140610-2ubuntu1 [28.4 kB]
[10:09] <seb128> LocutusOfBorg1, great
[10:23] <LocutusOfBorg1> 0ad needs another rebuild :(
[10:25] <seb128> I know
[10:25] <seb128> stop stressing over it, things take a bit
[10:26] <seb128> now that the binaries are there we can do the other rebuilds
[10:28] <LocutusOfBorg1> thanks ;)
[10:49] <LocutusOfBorg1> ps warzone3000 can be synced right?
[10:51] <LocutusOfBorg1> the delta was autoconf and miniupnpc related, both fixed in debian
[10:54] <Saviq> pitti, hey, is it currently possible to use apport-retrace on rtm-sourced .crash files?
[10:58] <pitti> Saviq: yes, sure; http://bazaar.launchpad.net/~ubuntu-archive/apport/lp-retracer-config/view/head:/Ubuntu%20RTM%2014.09/sources.list is the matching config for that
[10:58] <Saviq> pitti, the one in lp:daisy not good enough?
[11:00] <pitti> Saviq: ah, that looks identical indeed; I had thought that daisy woudl also use lp-retracer-config
[11:00] <Saviq> pitti, any pointer on why I would get http://pastebin.ubuntu.com/8759234/ then?
[11:01] <pitti> oh! 14.09 archive is missing Contents.gz
[11:01] <pitti> http://derived.archive.canonical.com/ubuntu-rtm/dists/14.09/
[11:01] <pitti> we need that to map a path to a package file
[11:01] <Saviq> oups
[11:02] <pitti> /home/michal/dev/canonical/retrace/sandbox/usr/bin/unity8
[11:02] <pitti> but aside from that, this path looks strange
[11:02] <pitti> probably just a confusing log message, though
[11:02] <Saviq> pitti, I am passing --sandbox-dir
[11:02] <Saviq> pitti, and -C and -S
[11:03] <pitti> Saviq: how old is that sandbox? does it work with a temp sandbox?
[11:03] <pitti> (i. e. without --sandbox-dir)
[11:03] <Saviq> pitti, there's nothing in there
[11:03] <Saviq> pitti, there's only packages.txt
[11:04]  * Saviq tries with a temp
[11:04] <pitti> ok, no immediate idea then :/
[11:06] <Saviq> pitti, here's the full log, if that could help with anything http://pastebin.ubuntu.com/8759314/
[11:07] <pitti> Saviq: it seems your .crash file doesn't have a Package:/Dependencies: fielD?
[11:07] <pitti> Saviq: and since RTM lacks Contents.gz, the "map file to package" fallback also doesn't work
[11:07] <Saviq> pitti, hmm Package: unity8 8.01+15.04.20141030~rtm-0ubuntu1
[11:08] <pitti> Saviq: did you run that unity crash through apport-cli or whoopsie-upload-all first?
[11:08] <Saviq> pitti, yes
[11:22] <cjwatson> pitti: Oh, heh, I guess we need to run the contents-generation job for 14.09 ...
[11:23] <ochosi> hey folks, quick question, when will status.ubuntu.com be bumped for V?
[11:23] <ochosi> pitti: and somewhat related, thanks for approving the xubuntu blueprints!
[11:28] <pitti> @pilot in
[11:49] <pitti> LocutusOfBorg1: erk, poor you (just reviewing runit merge)
[11:49] <pitti> this package is an utter mess
[11:52] <LocutusOfBorg1> pitti, why?
[11:52] <pitti> LocutusOfBorg1: err
[11:53] <pitti> LocutusOfBorg1: no debhelper, /etc/inittab, no dh_installinit, debian/rules is a mess, no SysV init script, etc.
[11:54] <pitti> TBH, this doesn't even look like something which we'd ever want to see on an Ubuntu system, so perhaps we should just kill it completely
[11:54] <LocutusOfBorg1> understandable :)
[11:54] <LocutusOfBorg1> so the problem is not the merge but the debian packaging?
[11:55] <pitti> there's also some problems with the merge (I'm following up to the bug)
[11:55] <LocutusOfBorg1> pitti, if you want to kill it don't loose your time in replying ;)
[11:56] <pitti> LocutusOfBorg1: ok, bug 1386581 updated
[11:56] <pitti> LocutusOfBorg1: well, it was mostly a sentiment; you apparently seem to care about the package :)
[12:00] <LocutusOfBorg1> dear pitti I don't know, users seems to be using it, see bug 1371290
[12:00] <LocutusOfBorg1> but if you want to drop it is fine for me, looking at the merge was honestly a PITA
[12:00] <pitti> yeah, like "I installed this wanna-be unsupported init system and now my stuff breaks"
[12:01] <pitti> LocutusOfBorg1: there's a number of reverse depends; many of them are only suggests: or have alternatives, but that needs to be checked
[12:01] <cjwatson> pitti: RT#76351 for Contents on 14.09.  It may well run into https://bugs.launchpad.net/launchpad/+bug/1384797 quite often, though.
[12:02] <pitti> cjwatson: ah, thanks
[12:05] <LocutusOfBorg1> pitti, what do you suggest?
[12:05] <LocutusOfBorg1> I don't care about runit, and if it makes users break stuff I think we just need to make it go
[12:06] <pitti> LocutusOfBorg1: if you want to fix the merge, go ahead; otherwise, if you don't care either, feel free to rename to "Please remove runit from vivid" and subscribe  ubuntu-archive
[12:23] <LocutusOfBorg1> the last runit merge :) my eyes are bleeding
[12:49] <Saviq> kirkland, hey, is there a way to copy text from byobu's F7 scrollback so that it's available in the X11 clipboard?
[12:57] <cjwatson> infinity: Could you merge libgcrypt20?  Blocks the imagemagick transition for reasons involving vdr plugins failing to build due to libgcrypt{11,20}-dev conflicts.  I checked that the ppc64el -O2 tweak is still required, at least in a utopic chroot.
[13:33] <apachelogger> stgraber: hey, ever heared of lxc refusing to get an ip address when multiple containers are starting at the same time? log doesn't seem very eventful http://kci.pangea.pub/job/utopic_unstable_sonnet/ws/lxc.log/*view*/ but lxc-info --ips didn't return anything over a 30 second period http://kci.pangea.pub/job/utopic_unstable_sonnet/70/console
[13:34] <apachelogger> thing is that this seems to be related to the fact that we are starting up to 30 containers at the same time as a retry outside that scenario will work just fine
[13:35] <apachelogger> also the user is allowed to have up to 128 veth devices, so I doubt it would be caused by that
[13:36] <stgraber> hmm. So I guess the things to check are 1) is the container getting bridged to lxcbr0 2) do you see a DHCP request in the host /var/log/syslog
[13:37] <apachelogger> stgraber: how would I best check 1)?
[13:37] <stgraber> apachelogger: lxc-info will get you the name of the veth device under Link, then check that against brctl show lxcbr0
[13:38] <apachelogger> stgraber: ok, thanks, I am going setup additional logging. I'll report back when it happens again ;)
[13:46] <kirkland> Saviq: sort of
[13:46] <kirkland> Saviq: presumably, you mean "other than by using your mouse and highlighting", right?
[13:47] <kirkland> Saviq: if so, then, you need 3rd party utilities (xclip);  see this post: https://unix.stackexchange.com/questions/131011/use-system-clipboard-in-vi-copy-mode-in-tmux
[13:51] <pitti> @pilot out
[14:01] <mdeslaur> So...the new dpkg in vivid-proposed is causing the apt autopkgtest to fail
[14:01] <mdeslaur> but, looks like dpkg is going to get promoted
[14:01] <pitti> mdeslaur: oh, dpkg breaks apt?
[14:01] <mdeslaur> pitti: it's making the test-bug-661537-build-profiles-support test fail
[14:02] <pitti> mdeslaur: we can still hold it back
[14:02] <mdeslaur> pitti: so, why isn't jenkins doing an apt autopkgtest for the dpkg upload?
[14:02] <pitti> mdeslaur: because apt doesn't depend on dpkg
[14:03] <pitti> (curiously)
[14:03] <pitti> well, it's essential
[14:03] <cjwatson> To be fair, causing a build-profiles test to fail is not the end of the world given that nothing is using those yet
[14:03] <cjwatson> And apt might genuinely need a tweak there
[14:03] <pitti> yeah, I agree
[14:03] <cjwatson> There may be a meta-problem that needs to be looked at here, but the actual proximate problem doesn't seem like cause for concern
[14:05] <mdeslaur> so how do we proceed? the apt failure is blocking my gnupg migration
[14:06] <mdeslaur> so, stuff that's in essential, and doesn't typically get listed as depends will never get tested properly?
[14:07] <pitti> mdeslaur: not as rdepends, indeed; the current logic only tests reverse build/binary depends
[14:07] <cjwatson> Maybe there's a call for some manually-inserted special cases here
[14:07] <pitti> (not saying that the behaviour is right/desirable, just how it works ATM)
[14:08] <mdeslaur> pitti: apt does list dpkg-dev as a build-depends, wouldn't that trigger it?
[14:08] <pitti> mdeslaur: hm, right, it should
[14:09] <pitti> mdeslaur: TODOed, will check with jibel on MOnday when he's back
[14:09] <mdeslaur> thanks pitti
[14:09] <mdeslaur> pitti: can gnupg get manually promoted in the meantime?
[14:10] <pitti> mdeslaur: yes, I can override the test result of apt for gnupg
[14:10] <mdeslaur> pitti: please do, thanks
[14:26] <Saviq> kirkland, great, thanks!
[14:27] <Saviq> kirkland, and yeah, selecting with mouse is great... until you start splitting vertically or scrolling up
[14:28] <kirkland> Saviq: right, so in that situation, what I usually do, is first press Shift-F11
[14:29] <kirkland> Saviq: which makes your current split "full screen"
[14:29] <kirkland> Saviq: and then highlight
[14:29] <kirkland> Saviq: Shift-F11 then toggles back out
[14:29] <Saviq> kirkland, indeed, was missing that, that definitely helps
[14:29] <Saviq> kirkland, thanks
[15:08] <cjwatson> doko_: Please could you merge plplot?  Indirectly blocking imagemagick transition (via gnudatalanguage).
[15:24] <cjwatson> xnox: Any progress with pythonmagick?
[15:25] <xnox> cjwatson: i must have been disconnected from my irc proxy. i have uploaded that. let me check if that got accepted properly.
[15:25] <xnox> https://launchpad.net/ubuntu/+source/pythonmagick/0.9.11-2ubuntu1
[15:25] <cjwatson> xnox: Ah, so you did, thanks
[15:26] <cjwatson> Note to self, reload transition page before complaining
[15:26]  * xnox is having all sorts of connectivity issues today.
[15:26] <xnox> cjwatson: well i pinged you on irc when i uploaded it. but it clearly didn't make it through the tubes.
[15:27] <cjwatson> It didn't, indeed
[15:30] <xnox> on the plus side, i have the perfect yubikey - it has all OTPs (lp, github, etc), U2F for google accounts, and gpg subkeys, and it can do NFC for otp authentication on the phone.
[15:30] <xnox> launchpad accepted subkey straight away, but debian infrastructure is taking its time to considering whether to accept the update into the keyring =)
[15:35] <dobey> xnox: how do you read via NFC on an ubuntu phone? :)
[15:41] <xnox> dobey: the hardware supports it, and the protocol is simple - it's just a url, thus shouldn't be hard to implement.
[15:42] <xnox> dobey: at the moment however, i'm only using it with Lastpass Android app - for two factor unlocking of my passwords store. And the Yubikey Auth Android App - which triggers HOTP (time and event based) generation on yubikey and displays the results.
[15:42] <dobey> xnox: just hard to access the hardware in an app in ubuntu :)
[15:43] <xnox> Yubikey Auth app is quite awesome, since all secrets are stored tamper proof on the yubikey and one can generate time-based OTP for Microsoft, Github, Facebook, Google accounts, and event based like for Launchpad.
[15:43]  * xnox ponders why we don't do timebased OTPs in launchpad, given that everyone else does as well by default.
[15:43] <xnox> no need to "resynchronise the key" if one presses the otps too much somewhere.
[15:44] <dobey> anyway, need to get some lunch, without using nfc payments
[16:01] <caribou> any quick trick to pass MULTI_OS_DIRECTORY to make using debhelper ?
[16:02] <cjwatson> like   override_dh_auto_build:\n\tdh_auto_build -- MULTI_OS_DIRECTORY=foo   you mean?
[16:03] <ricotz> pitti, hi, nearly forgot to mention that upgrading from precise to trusty postgresql-9.1 made a problem while is can't coexist with postgresql-9.3 due some dkpg-divert collision/incompatibility, so it needed to be removed, sorry i dont have a log
[16:03] <caribou> cjwatson: I need to replace the following cdbs statement :
[16:03] <caribou> DEB_MAKE_EXTRA_ARGS += MULTI_OS_DIRECTORY=$(DEB_HOST_MULTIARCH)
[16:04] <caribou> libnss-ldap upstream has removed cdbs & replaced by debhelper
[16:04] <caribou> I must admit that this is the first time I move a package from cdbs to debhelper
[16:04] <cjwatson> so then that's as I said above.
[16:05] <caribou> cjwatson: ok, thanks!
[16:05] <cjwatson> assuming by debhelper you mean the short dh(1) style
[16:05] <caribou> cjwatson: yes
[16:19] <bdmurray> xnox: Do you recall bug 1123798 at all? Not having whoopsie running on the Live CDs seems unfortunate.
[16:20] <pitti> ricotz: hm, I need logs, I'm afraid; postgresql-9.1 doesn't ordinarily get diverted
[16:28] <xnox> bdmurray: yes, i recall that bug. That shouldn't be happening, now that ubiquity-dm (for a few releases now) opens a pam session (which in turn opens a logind session)
[16:31] <bdmurray> xnox: would running whoopsie and installing be a good test?
[16:33] <xnox> bdmurray: so if there is no pam_systemd.so configured on the system, logind session will not be established and consolekit fallbacks may kick in. But I thought, by now, we even removed consolekit
[16:35] <xnox> bdmurray: i don't quite recall the reasons we disable whoopsie - it's installed, it has checks in the upstart job to not run when ubiquity is present.
[16:35] <xnox> bdmurray: so boot live cd, edit the whoopsie job, start whoopsie, restart lightdm - that should restart ubiquity-dm as well.
[16:38] <slangasek> xnox, bdmurray: aside from the fact that inotify doesn't work on the unionfs?
[16:39] <xnox> bdmurray: slangasek: "edit whoopsie job; initctl reload-configuration; start whoopsie"
[16:39] <xnox> that works.
[16:39] <slangasek> xnox: yes, but whoopsie itself uses inotify
[16:41] <xnox> slangasek: hm, it would work if /var/crash inode is updated (e.g. touch /var/crash/.poke, whould cause inode to be created on the overlay, and if whoopsie is started after that, monitoring will start working)
[16:41] <xnox> slangasek: should i make a patch for casper to setup /var/crash in the overlay?
[16:42] <slangasek> xnox: no opinion; was just responding to the question of why it was currently disabled, and I think this is why
[16:42] <xnox> slangasek: i think inode monitoring works correctly, because the reason why we disabled whoopise on the live images, is that when starting live session people say crash popups.
[16:42] <xnox> slangasek: so actually, i believe it all just works (if even because we crash before starting live session)
[16:43] <xnox> slangasek: no, that was not the reason. It was UX/rick pressure to disable. ev should know more details.
[16:43] <bdmurray> update-notifier definitely uses inotify (for crash file detection) and wouldn't work
[16:43] <slangasek> ok
[16:46] <bdmurray> and if whoopsie is running and you touch a .upload file the .crash file does get uploaded
[16:50] <ricotz> pitti, it was about "update-alternatives: using /usr/share/postgresql/9.1/man/man1/psql.1.gz to provide /usr/share/man/man1/psql.1.gz (psql.1.gz) in auto mode." or "update-alternatives: using /usr/share/postgresql/9.1/man/man1/postmaster.1.gz to provide /usr/share/man/man1/postmaster.1.gz (postmaster.1.gz) in auto mode." suggesting an incompatibility, although i am not able to reproduce it in a chroot -- it happened on do-release-upgrade
[16:50] <ricotz> pitti, sorry
[18:14] <bdmurray> xnox: So do you think the patch for casper is needed or not?
[18:15] <infinity> cjwatson: Can do.
[18:15] <infinity> Hahaha.
[18:16] <infinity> Dear MoM, congrats on failing to merge a 4-line delta.
[18:16] <infinity> No love, Me.
[20:49] <Noskcaj> cyphermox, IS there any chance you could help with bug 1387944 ? I don't think i understand the package well enough to do the main merge
[20:58] <bdmurray> slangasek: I'm working on an interim solution for bug 1365079 and I wonder if "title" in 'click info's output is too generic.
[20:58] <bdmurray> slangasek: for the package name that is
[20:59] <slangasek> bdmurray: title is indeed not guaranteed to be unique; should probably use name instead?
[20:59] <bdmurray> slangasek: yeah, that was my idea too. thanks
[21:04] <sergiusens> +1 on name
[21:29] <Noskcaj> Can someone please retry libgee-0.8 and modemmanager now vala 0.26 is in main?
[21:33] <Laney> yep
[22:11] <Noskcaj> Also please retry dballe, it had a out-of-data dep that has now synced
[23:29] <infinity> hallyn: Hey, why is libcgmanager0 not linked with libnih?
[23:30] <infinity> hallyn: Found entirely by accident while running "adequate libc\*" to check my own packages, and there was libcgmanager whining about a bunch of unresolved nih symbols. :P
[23:31] <infinity> stgraber: ^
[23:34] <infinity> ... and why is my Ubuntu upgrade trying to install gdm? :P
[23:35] <infinity> and gnome-shell.
[23:35] <infinity> Whee.
[23:37] <infinity> Oh, because I still had gnome-session installed.