[12:29] <scott-work> should the ubuntu studio alternate install still be experiencing problematic images?
[12:33] <scott-work> yesterday's image failed installation directly after the first menu
[13:40] <ScottK> cjwatson: I would appreciate it if you would respin kubuntu daily live i386.  I made some changes to help with CD size yesterday, but they didn't make it in and if it's not inconvenient, it'd be nice to know today how much more we have to do.
[14:50] <skaet> ScottK,  cjwatsons out today,  I've kicked off the kubuntu daily-live again for you.
[14:50] <ScottK> skaet: Thanks.
[15:51] <skaet> ScottK,  first attempted failed,  retrying now.   If this fails, will ask for help figuring out what went wrong.
[15:51] <ScottK> skaet: Don't bother.
[15:51] <ScottK> I know what went wrong.
[15:51] <ScottK> I just uploaded a fix.
[15:51] <ScottK> Try it in an hour.
[15:51] <ScottK> (I think)
[15:51] <skaet> ok,  will do.
[16:06] <ScottK> skaet: If you would kick them off in ~40 minutes, that'd be great.
[16:09] <skaet> ScottK,  will do.
[16:09] <ScottK> Thanks.
[16:19] <charlie-tca> I have an email saying today's images (2011-07-22) are oversize, but the images don't seem to be on http://cdimages.ubuntu.com/
[16:27] <skaet> charlie-tca, I don't see them there either.  hmm...    will look into it after my next meeting (if someone else doesn't solve the mystery before me,  ;) )
[16:27] <charlie-tca> Thanks!
[16:28] <charlie-tca> Seems like all images are missing today, not any particular ones
[16:29]  * skaet wondering if its going to be a space issue.... :P
[16:48] <charlie-tca> no build logs for today, either
[16:49] <charlie-tca> They may not have been built
[17:30] <hggdh> skaet: sorry, I had already pressed the hang-up button you you said my name
[17:31] <hggdh> so all I got was "Carlos, is"
[17:31] <skaet> hggdh,  no worries.   what time would suit you to talk this afternoon?
[17:31] <hggdh> skaet: anytime
[17:31] <hggdh> your call, even now
[17:31] <skaet> I'll call you back now then.
[17:32] <hggdh> waiting
[17:58] <ScottK> skaet: Did you kick off the kubuntu live rebuilds we discussed?  If so I'm seeing neither failure mails nor new images.
[18:03] <skaet> ScottK,  something strange is going on....
[18:04] <skaet> it looked like the builders suceeded - but i'm not seeing the images either.
[18:06] <ScottK> OK.  Time to ask for help, I guess.
[18:06] <ScottK> Given that charlie-tca saw the same.
[18:06] <skaet> yup,  put in the ping to IS.
[18:06] <ScottK> Thanks.
[18:06] <charlie-tca> no logs, no images for Ubuntu, Xubuntu, Ubuntustudio
[18:07] <ScottK> I have reason to believe the automatic image builds didn't go.
[18:07] <ScottK> When skaet did manual builds for kubuntu live it failed.
[18:07] <ScottK> I think it would have failed last night too.
[18:38] <skaet> ScottK,  charlie-tca; the obvious space issue has been ruled out.   trying some sample builds now with IS to see if that sheds any light...
[18:41] <charlie-tca> Thanks for trying
[18:48] <ScottK> smart: I got a powerpc livefs failure mail (expected )
[18:48] <ScottK> So ... something is happening.
[18:50] <ScottK> Still no Jul 22 images though.
[18:50] <charlie-tca> xubuntu images appeared
[18:50] <charlie-tca> no ubuntu or ubuntustudio, though
[18:52] <infinity> skaet: I'm going to assume that everything is commented out of cdimage's crontab for the purpose of this testing, right?
[18:52] <infinity> skaet: Not because someone forgot to turn it all back on? :)
[18:52] <ScottK> Progress then...
[19:05] <skaet> charlie-tca - there's Xubuntu images up now for you.
[19:06] <skaet> or at least I can see they've built.
[19:06] <charlie-tca> yup, they are there
[19:06] <charlie-tca> Thank you very much
[19:06] <skaet> ScottK,  final stages for a set of Kubuntu ones.
[19:07] <jdstrand> skaet: do you have a moment to talk about dbus/apparmor as it applies to oneiric and 'p'?
[19:07] <ScottK> Great.
[19:11] <skaet> heya all,  if you want any images today,  please let me know here,  we had a glitch with the cronjob last night (still not sure how it all got commented out... :P )    Hopefully we'll be back to full set of dailies tomorrow.
[19:13] <skaet> jdstrand, now that our build mystery is solved (thanks infinity), sure.
[19:13] <charlie-tca> skaet: studio needs a set if we can get them. They missed yesterday, too
[19:13] <jdstrand> skaet: ok, this shouldn't take long. just a heads up, and then a question or two
[19:13] <jdstrand> skaet: so, you may be aware of https://blueprints.launchpad.net/ubuntu/+spec/security-o-apparmor-dbus
[19:13] <ScottK> jdstrand: Just make it perfectly secure and work flawlessly.  What is there to discuss?
[19:13] <jdstrand> this is something that the security and kernel teams are working together on
[19:14] <jdstrand> ScottK: hehe
[19:14] <jdstrand> skaet: so light background
[19:14] <jdstrand> skaet: we have apparmor to confine applications. it lets you saw 'you can read that file, execute that one, use networking, etc'
[19:15] <jdstrand> skaet: the problem is, with dbus it is too coarse
[19:15] <skaet> charlie-tca, ubuntu studio triggered now.
[19:15] <charlie-tca> Thank you
[19:15] <jdstrand> skaet: ie, you can either talk to dbus or not
[19:16] <jdstrand> skaet: so if you can talk to dbus, you can talk to everything on the dbus. this is a problem as more and more applications are providing services
[19:16] <jdstrand> skaet: so, we are tasked with making dbus acces more finegrained
[19:16]  * skaet nods
[19:16] <jdstrand> skaet: "this confined application can talk to this application via dbus, but not that application"
[19:17] <jdstrand> ok, so this bp is all about that
[19:17] <jdstrand> there is kernel work and userspace work
[19:17] <skaet> ok, have it openened,  looks like its mostly in progress.
[19:17] <jdstrand> the userspace consists of updating libapparmor and dbus itself
[19:17] <jdstrand> the changes are big, and much is yet to be written
[19:18] <jdstrand> but we *really* want it for 'p', so the plan is this:
[19:18] <jdstrand> get a libapparmor and a dbus up that has the patches
[19:18] <jdstrand> the dbus apparmor code paths will be disabled by default
[19:19] <jdstrand> but can be turned on via a conffile
[19:19] <jdstrand> nothing other than dbus will use the new ipc/dbus API in libapparmor
[19:20] <jdstrand> what this will allow is for us to have the code in oneiric for people to play with, with very low risk to regular users
[19:21] <jdstrand> it also, you permitting, would allow us to push functionality into these unused-by-default-users functionality/bug fixes/changes in past feature freeze
[19:21] <jdstrand> the idea being that when 'p' opens, we turn it on for real and fix bugs/iterate through 'p' so it is solid for release in april
[19:22] <jdstrand> skaet: what do you think?
[19:23] <jdstrand> skaet: tbh, we aren't going to be adding features past feature freeze (unless we miss it, which is conceivable, but I'd use the freeze exception process appropriately)
[19:23] <skaet> jstrand, seems a reasonable staged approach.    When would you plan on landing libapparmor and dbus changes?
[19:23] <jdstrand> skaet: but there will likely be a lot of code changes
[19:23] <skaet> heh
[19:24] <skaet> The best interval to land this is right after A3 before feature freezed.
[19:25] <jdstrand> skaet: we are actively working on it. so, I have to say looking at the dates that we will likely need a freeze exception
[19:25] <skaet> Is there a good way to back out if it opens up some unforseen interaction issues?  (I think disabled by default should handle it.... but?)
[19:25] <jdstrand> skaet: yeah totally-- it is just a quilt patch
[19:25] <jdstrand> (for dbus)
[19:26] <jdstrand> libapparmor wouldn't be a big deal, cause it's api nothing uses
[19:27] <jdstrand> I can most likely have that be a set of patches on top of apparmor as well
[19:27] <skaet> jdstrand,  ok, lets check point around A3 timeframe,  and see what the ETA is
[19:27] <skaet> I'll hopefully have a better picture about the rest of churn around Feature freeze at that point.
[19:28] <jdstrand> skaet: so looking at that, I can't do between a3 and ff-- holiday. it is possible, that quite unlikely that I could have a dbus and libapparmor in by next friday
[19:28] <jdstrand> skaet: basically, I'm thinking it is almost certainly a freeze exception
[19:29] <skaet> heh,  that explains it then.
[19:29] <jdstrand> skaet: but, one thing you can count on is I do thorough testing :) we have a lot of tests for dbus and apparmor, and I will be adding more. I will certainly be adding some for the default configuration of 'off'
[19:29] <jdstrand> s/that quite/though quite/
[19:29] <skaet> Yup.  :)   As long as it doesn't get too close to beta,  we should be ok.
[19:30] <jdstrand> skaet: ok. thanks! I will certainly keep you posted
[19:30] <skaet> If we don't have it in, and stable 2 weeks before beta lets regroup.
[19:31] <ScottK> I see we have Kubuntu images now.  Thanks.
[19:31] <jdstrand> skaet: incidentally, this is more than just something the security team would like to have-- it is something several projects need/want, including ARB
[19:32] <skaet> yup,  am sympathetic and can see it is going to be a good thing.   Just balancing it with some other things that I'm cynical about landing late. ;)
[19:32] <skaet> and it will be best for P,  if we can indeed get it in and kick the tires a bit now.
[19:33] <jdstrand> skaet: sure. well, like I said, this will be totally opt-in and should be very low risk for oneiric
[19:33] <jdstrand> skaet: but yeah, you are rightfully cautious and we'll talk more about it when there is more to discuss :)
[19:33] <skaet> jdstrand, sounds good.   and thanks for explaining now rather than waiting until Feature Freeze :)
[19:33] <jdstrand> skaet: sure thing! :)
[19:40] <charlie-tca> Thank you, skaet. We have UbuntuStudio images now
[19:44] <skaet> For completeness,  have just triggered off a set of ubuntu dailies (desktop and server) as well.