[11:44] <cjwatson> please nobody touch cdimage for a while; we're switching to a new machine
[11:44] <cjwatson> I'll mail ubuntu-release with details when it's done
[13:16] <brendand> cjwatson - by don't touch you mean don't try to download from it?
[13:21] <cjwatson> brendand: downloading is fine, I meant operators
[13:21] <cjwatson> sorry to confuse
[13:21] <brendand> cjwatson - np
=== chuck__ is now known as zul
[14:30] <jdstrand> skaet: hey. I just filed bug #888568. I understand mdeslaur spoke to you about adding ubuntu-security to ubuntu-release so we can manipulate nominations
[14:30] <ubot4> Launchpad bug 888568 in launchpad "ubuntu-security should be able to target to release (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/888568
[14:32] <jdstrand> skaet: I'm not sure how you want to move forward, but right now my team can't fully participate in the kernel cadence because of this bug. this might need you to say this bug is important for ubuntu engineering, or it might be not super important for now, but add ubuntu-security to ubuntu-release until this bug is fixed
[14:39] <cjwatson> OK, nusakan is live now as the antimony replacement
[14:39] <cjwatson> mail sent to ubuntu-release
[16:56] <micahg> jdstrand: I think we could get away with being added to -drivers instead of -release and still have the same result
[16:56] <cjwatson> no, please don't re-bloat -drivers
[16:56] <cjwatson> we only just finished cutting it down
[16:57] <micahg> is bloating release any better?
[16:57] <jdstrand> micahg: yes, but we aren't a driver, and that group is trying to have reduced privs and membership. release isn't correct either. I'm hopeful security contacts will just get the perms to add tasks as part of the pending driver updates
[16:57] <jdstrand> micahg: but skaet and I may have a temporary workaround depending on what happens with that bug
[16:58] <jdstrand> micahg: see the comments in that bug. there seems to be a reasonable path forward, and it may even be timely :)
[16:58]  * skaet nods
[17:00] <micahg> jdstrand: yeah, but it seems that there's some incorrect assumptions in that bug, I'll comment
[17:03] <micahg> jdstrand: ah, no, the info is fine
[17:04] <jdstrand> cool
[17:16] <infinity> cjwatson: Oh my, that's a significantly beefier machine.
[17:28] <cjwatson> micahg: yes, -release has less privilege
[17:28] <cjwatson> -drivers is even more of a god team
[17:32] <jdstrand> skaet: mdeslaur will be taking over release status meetings for me. is it enough to let you know?
[17:33] <skaet> jdstrand, yup. That's fine.
[17:34] <cjwatson> infinity: isn't it just
[17:34] <skaet> jdstrand,  mdeslaur,   no meeting tomorrow.  ;)
[17:34] <cjwatson> 12 cores, SAN storage ...
[17:34] <jdstrand> heh
[17:35] <cjwatson> infinity: when you write your master build script, you probably want to err towards a bit more parallelism
[17:35]  * skaet is looking forward to seeing the release build statistics from the new machine.  :)
[17:36] <cjwatson> rough estimate from what I've seen so far, the non-livefs parts of builds are about two or three times faster
[17:36] <infinity> cjwatson: By which, you mean, "start everything at once, and cackle maniacally"? :P
[17:36] <cjwatson> infinity: pretty much
[17:37] <micahg> cjwatson: sorry, forgot about the milestoning privs and such, was thinking more about the possibility of release being able to handle freeze exceptions at some point (but I guess that's not even in the queue to be done yet)
[17:37] <cjwatson> and then wait forever for livefses, for now ;-)
[17:39] <cjwatson> ubuntu-server 20111107: mirror 9 minutes, build 22 minutes.  ubuntu-server 20111110: mirror 2 minutes, build 11 minutes
[17:41] <cjwatson> ubuntu alternate 20111107: mirror 10 minutes, build 16 minutes.  ubuntu alternate 20111110.1: mirror 6 minutes, build 6 minutes
[17:41] <cjwatson> there's moderate variance but even so
[17:41] <cjwatson> note that the various parts of cdimage's local mirror are now highlandered and it syncs straight from cocoplum
[17:42] <cjwatson> so it does have a bit more work to do in mirroring, and ultimately deploying lmirror would probably save us a bit more on full build passes
[17:43] <cjwatson> but this should be well out of the critical path for now
[18:07] <slangasek> I did not know highlander was a verb
[18:10] <cjwatson> 'tis now, English has productive verbing :)
[18:29] <skaet> cjwatson,  nice stats. :)
[20:16] <infinity> cjwatson: Hahaha.  "Highlandered".
[20:22] <slangasek> :)
[23:20] <GrueMaster> Not sure what is happening, but my preseeded netinstalls on omap4 (armel) are giving me a precise-restricted/Package file corruption warning.  I get this both on my local mirror and on ports.u.c.
[23:21] <GrueMaster> Can someone check that?