/srv/irclogs.ubuntu.com/2011/07/06/#ubuntu-installer.txt

evlifeless: hi08:02
lifelessev: wondering about 2 things08:04
lifelessev: the crash dump discussion, and how horrible my patch was08:04
cjwatsonev: wubi> looks basically fine; could you (a) rename the hook to ubuntu_binary_build-wubildr.sh, which I think fits better into the naming scheme (b) make sure that the live-build patch is forwarded as a Debian bug report?08:41
cjwatsonlifeless: sorry, not quite got round to it yet ... BTW do you have a GNU copyright assignment on file?08:41
evcjwatson: sure08:43
sorenIsn't bzr a GNU project these days?08:44
evnow to figure out how to wire this into cdimage.  Dug through the code a bit last night, but not being able run it on my local system will make this somewhat delicate.08:44
evlifeless: what were you wondering about the crash dump discussion?08:44
lifelesscjwatson: yes08:45
lifelesscjwatson: (I have a GNU assignment from years ago, libtool, binutils, smalltalk)08:46
lifelesscjwatson: I haven't done one for work@canonical, but this was pretty personal :)08:46
lifelesssoren: it is a GNU project, but it doesn't use the FSF assignment atm08:47
sorenlifeless: Oh, ok.08:47
lifelessev: is the project resourced? who will be hacking on it? what help can I give them?08:48
cjwatsonlifeless: would that assignment cover a patch to grub?08:48
cjwatsonev: *cough* I never bother running cdimage locally :-)08:48
lifelesscjwatson: I don't know08:49
evlifeless: I have a work item to create a job profile for an engineer to hack on it from the foundations team.  Matthew has written up the initial design specification: https://wiki.ubuntu.com/CrashTracker08:50
lifelessev: I'll look at that this weekish and offer feedback if I have any08:51
evlifeless: brilliant, that would be much appreciated08:51
evthanks08:51
evcjwatson: :)08:51
lifelesscjwatson: if it doesn't cover one, I'm certainly happy to sign another assignment etc08:52
evcjwatson: oh, there was one thing I wanted to ask you.  Do you feel that lzma'ing of the ext3 disk imagine and wubildr belongs more in live-build or lp:ubuntu-cdimage?09:16
cjwatsonnot sure I hugely care - putting it in live-build might involve adding a new option to lb config, which is more work and more coordination with Debian, but might be easier for users09:20
cjwatsone.g. if we want live-build's iso building mode to ultimately work for this case, it'll need to go in lb anyway09:21
cjwatsonbut I don't mind that much09:21
evI'm not sure I follow how this relates to the iso building mode?  Would people want to generate an ISO of just the Ubuntu base image along with wubildr?09:27
cjwatsonwho knows09:33
cjwatsonbut ok, I guess it wouldn't be that common, this is mainly non-iso09:33
cjwatsonso I guess I don't care, do whatever's easiest09:33
evyay, my favorite answer09:34
* ev tries to remember the magic runes to check if a livefs buildd is up to date10:34
ogra_ev, check the logs ?10:34
ogra_it used to print the versions at the top10:34
evI mean the version of livecd-rootfs and live-build10:35
ogra_yes, it used to print that at the top of the log10:35
evah10:35
ogra_but i just checked, seems thats gone with live-build10:35
cjwatsonfeel free to fix (probably in livecd-rootfs/live-build/auto/build)10:36
ogra_if your upload is published to the archive it should auto-update the builder on startup though (at least it did in the past, i hope that dindnt change as well)10:36
cjwatsonthe livefs buildds are dist-upgraded before starting10:37
ogra_good10:37
evah awesome10:37
evthanks10:37
cjwatsonsee BuildLiveCD, which does this10:37
evwill do, thanks10:37
cjwatsonev: could you take bug 806349?  looks like it should be straightforward11:51
ubot2Launchpad bug 806349 in ubiquity "OEM Install fails with - KeyError: "The cache has no package named 'python2.6-minimal'" - without network connection." [High,Triaged] https://launchpad.net/bugs/80634911:51
cjwatsonhm, actually, that's not fair, it's my code :)11:51
CIA-37ubiquity: cjwatson * r4768 trunk/ (debian/changelog scripts/plugininstall.py):11:57
CIA-37ubiquity: Handle the case where one of the packages returned by 'pyversions -s' or11:57
CIA-37ubiquity: 'py3versions -s' is not installed (LP: #806349).11:57
evheh, would've grabbed it but was out at lunch12:30
CIA-37installation-guide: cjwatson * r494 ubuntu/ (build/entities/common.ent debian/changelog): Bump kernelversion to 3.0.12:30
CIA-37installation-guide: cjwatson * r495 ubuntu/ (3 files in 3 dirs): Remove some 2.6 kernel specifics.12:32
CIA-37installation-guide: cjwatson * r496 ubuntu/ (debian/changelog en/welcome/what-is-ubuntu.xml): Bump GNOME version to 3.0.12:32
CIA-37installation-guide: cjwatson * r497 ubuntu/ (build/entities/common.ent debian/changelog): Bump release version and names for Oneiric.12:33
cjwatsonev: where have you been committing your livecd-rootfs changes to?  I just got an archive reject after committing to lp:livecd-rootfs and uploading.14:01
evoh bum14:01
evI didn't realize it was in version control14:01
evshould've checked14:01
evI'll merge them in now14:01
davmor2ev: I redirect you to my twitter reply to your I'm cowboying the livecd-rootfs system14:02
cjwatsonthanks.  I think the Vcs-Bzr field is correct14:02
evit is, and I even have a local copy14:03
evI just completely forgot about its existence :-/14:03
evcjwatson: so I see that BuildLiveCD dist-upgrades the chroot before running live-build, but how does one notify the buildds to update their copy of livecd-rootfs to get the latest BuildLiveCD?16:35
evis this just done in a cron job or is it a "bug lamont" thing16:35
cjwatsonev: the latter, via an RT ticket17:02
cjwatsonI try to avoid changing BuildLiveCD unnecessarily, but sometimes it's needed17:03
evit was seemingly necessary here, but noted17:03
evgiven that this requires the RT runaround17:03
evwould you mind just reviewing the small change I made to it in trunk?17:03
evjust to make sure that's the interface you're happy with17:04
cjwatsonseems fine17:07

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!