[04:04] Good morning [06:57] good morning [06:57] guten Morgen dholbach [06:58] salut jibel :) [07:06] dholbach, I prepared some tools and documentation for the testing hackfest [07:06] ah nice [07:06] dholbach, https://launchpad.net/auto-package-testing [07:06] dholbach, http://bazaar.launchpad.net/~auto-package-testing-dev/auto-package-testing/trunk/view/head:/doc/USAGE.md [07:07] awesome - I'll take a look at it in a sec :) [07:07] dholbach, would you have time for a review? [07:07] ack [07:19] jibel, that looks great [07:21] dholbach, cool, thanks! [07:22] dholbach, I think we should ask people to prepare a base VM before the event, so we don't waste time with that during the session [07:23] in the announcement maybe [07:29] jibel, I think pitti said he had prepared a base vm which people could copy [07:31] right, but doing that costs quite a lot of time, too [07:32] I'll still bring it, of course [07:32] (USB stick) [07:32] and we should put it on the posters [07:44] jibel: ohhhh, very nice it creates the test vm automatically now and runs automatically in it? very cool [07:46] mvo, right, and you can even login and keep the VM after the run for investigation, cool isn't it? [07:47] !!! [07:47] yes [07:48] jibel, do we have plans to put this into the archive? :) [07:49] dholbach, when I'll know how to create a package ;) [07:50] I was asking because I was wondering if I should put something about it into http://developer.ubuntu.com/packaging/html/auto-pkg-test.html [07:50] right now it just has this little blurb: http://developer.ubuntu.com/packaging/html/auto-pkg-test.html#executing-the-test [08:04] jibel, ^ so once it's in the archive I think we should advertise it in the guide [08:04] thanks a lot for your work on this [08:07] dholbach, I'm working on a juju charm for autopkgtest not sure we'll want to maintain this tool and the charm [08:07] dholbach, unless I can share the scripts [08:10] jibel, to package the scripts they might need to be made aware of different paths [08:11] jibel, if you want we can look at this during the cycle [10:38] $ cat ~/.adtrc [10:38] BASEDIR=/home/martin-scratch/images/ [10:38] jibel: but when I run ~/ubuntu/auto-package-testing/bin/run-adt-test -r quantal -d -a amd64 -b lp:~pitti/ubuntu/quantal/coreutils/adt [10:38] it still looks for stuff in /tmp/adt [10:40] pitti, looks like a bug. I finish something, then lunch and I'll have a look right after. [10:40] I'm looking into it [10:44] ooh, I know why === yofel_ is now known as yofel [10:51] jibel: http://paste.ubuntu.com/1306887/ fixes it; want me to commit? [10:51] it still doesn't actually start kvm (looking into that next), but it does fix the paths [10:52] right, qemu-img fails, -b doesn't exist [10:54] odd, the backing file option is documented, but the manpage (nor --help) mention it as valid for "create"; only for "rebase" [10:55] quantal regression? [10:55] ooh, -o [11:04] jibel: nevermind, that patch is obviously bogus; I'll think of something better === _salem is now known as salem_ [11:40] jibel: better fix: http://bazaar.launchpad.net/~pitti/auto-package-testing/fixes/revision/74 [11:44] pitti, thanks. merging [11:45] pitti, do you have any other fix? [11:45] jibel: I can push for cleaner history [11:45] pitti, ack [11:46] jibel: yes, I'm working on making the script run with set -e, fixing the qemu-img call, and whatever else comes up [11:46] jibel: if you want, I'll keep pushing to my branch, and you take a look when I'm done? [11:46] batch mode [11:46] pitti, sure, [11:47] pitti, I removed -e because it the script exited on return of get_free_port, it'd be nice to have it back [11:48] ack, will do [12:09] jibel: oh, seems qemu-img failed for a different reason; it seems it still accepts -b, but it's not documented [12:10] jibel: I think we should change it to -o backing_file anyway (it's that way since at least lucid) [12:11] yay, seems to run happily now [12:12] pitti, what was the problem with -b ? [12:12] -b? [12:12] :) [12:13] jibel: it's not a documented option [12:13] jibel: oh, I moved /tmp/adt to /home/martin-scratch/adt, but that didn't update the symlink; that confused it [12:13] I re-ran prepare-image with my BASEDIR fix now, working fine now [12:14] hm, it's building coreutils in the VM, is it supposed to? [12:15] no, it is not supposed to build coreutils [12:18] $ ~/ubuntu/auto-package-testing/bin/run-adt-test -r quantal -a amd64 -b lp:~pitti/ubuntu/quantal/coreutils/adt succeeded now \o/ [12:18] (except that I didn't expect it to build/install, just run the tests) [12:20] jibel: ok, top three commits in http://bazaar.launchpad.net/~pitti/auto-package-testing/fixes/changes, ok for me to push? [12:21] * pitti deletes his old manual adt VM, this is so much nicer [12:23] pitti, thanks for all the fixes. Feel free to push. [12:24] jibel: ooh, that would be with -b, that doesn't supply --no-built-binaries [12:24] well, I guess that's kind of correct [12:25] jibel: mind if I add a --no-build option? [12:26] pitti, I don't mind, I was not certain which options were best when testing from a branch [12:26] jibel: I think defaulting to build is the safe option, as you might have actual code fixes in the barnch [12:26] but if your branch only adds tests, this is a faster short-cut [12:27] and I didn't want to expose all of adt-run options from the script otherwise it's easier to run adt-run directly. [12:34] jibel: oh, nevermind; I didn't specify a package name; when I append "coreutils", it works [12:34] * pitti read bin/testbed/run-adt more carefully now [12:34] ♥ [12:34] this looks great now [12:35] pitti, when the user just specifies a branch we suppose that he wants to build the branch [12:36] when there is a branch and a package, we suppose that the user wants to tests the binary with the tests from the branch [12:36] documentation patchs are also accepted [12:36] :) [12:40] jibel: http://paste.ubuntu.com/1307065/ ? [12:45] jibel, gema, dholbach: ok, so for Tue evening I now have a live demo how to adt-ify coreutils [12:46] great :) [12:46] I guess we have a projector there, we'll need it for the lightning talks anyway [12:46] Awesome! [12:47] and we have three copy&paste-able commands from the wiki page to create a VM [12:48] let's get ready to rumbleeeee! [12:52] mmrazik, there's something in the slides about autopilot training [12:53] mmrazik, do we have times/dates/rooms for that already? [12:54] dholbach: you mean this one: http://summit.ubuntu.com/uds-r/meeting/21249/qa-r-autopilot-in-ue/ ? [12:54] do we need to supply slides beforehand? [12:54] dholbach: or you are talking about the plenary on Monday? [12:54] the plenary is on monday [12:54] yes, the slides still have TBD on them :) [12:55] mmrazik, so that's one session? [12:55] * thomi was summoned [12:55] thomi: dholbach is curious about the plenary slides. Are they ready? [12:55] thomi: they seem to be tagged TBD [12:55] mmrazik, if it's just one session I can add it to the last slide [12:55] just wanted to make sure [12:56] dholbach: the session on Monday is a workshop/session where people can have questions and where we can talk in more depth about stuff [12:56] dholbach: yes. thats it. Plenary and then this single autopilot session. [12:56] uhhh, they're ready, except that the date & time of the autopilot session might change, and I wanted to make sure they were accurate [12:56] http://summit.ubuntu.com/uds-r/meeting/21249/qa-r-autopilot-in-ue/ says 2012-10-31, which would be wednesday [12:57] dholbach: we were just wondering with thomi yesterday how much is the schedule set in stone [12:57] thomi, if you say you will take care of updating the slide I'm happy :) [12:57] we can still change the last slide on Monday morning if you want [12:57] dholbach: sounds good to me. slides 17 & 18 are the ones that need to be updated maybe [12:58] I'll leave that to you guys :) [12:58] thanks a bunch for all your work everyone [12:58] UDS will be great for automated testing [19:53] can everyone with a ppc computer please see if the netboot Bug #1071880 affects them so we can see what hardware causes it [19:53] Launchpad bug 1071880 in debian-installer (Ubuntu) "Netboot PPC crashes with"Fixing recursive fault but reboot is needed!"" [Undecided,New] https://launchpad.net/bugs/1071880 [19:54] also whats the command to get a hardware profile so i can upload mine [19:57] Noskcaj: it's better to email the l-qa mailing list. Ensure you put PPC in the subject heading so that people know it is PPC related, thanks :) [19:58] phillw: i was about to, just putting it in the irc first [20:00] Noskcaj: wxl did email about that, I'll ask that he put it onto https://wiki.ubuntu.com/Lubuntu/Testing/PPC%26Mac64 so it easily found, I'm still restoring piglet at the moment but will have a quick look through my emails to see if I can find it. [20:08] phillw: thanks [20:11] Noskcaj: http://pastebin.com/7vMCqsTq [20:11] I'll try and get the wiki area done tonight (piglet willing). [20:13] i will run the profiler as soon as the lubuntu cd downloads [20:13] i have just sent off the email too [20:17] yeah, just seen the email. I'm still battling with permissions / ownership issues from my restore... Well, it was a CentOS backup of my /home partition so I did expect some 'fun' importing it into Lubuntu 12.10 :D [20:17] and we have another PPC tester :) [20:20] i saw that, Yay [20:20] and centOS to lubuntu, sounds very fun [20:24] it's okay.. just a lot of chown -R / chgrp -R going on :P [20:25] I know phil from ubuntu-classroom-backstage (he's one of the 'staff' on there) - He must have like my presentation the other day :D [20:27] I' going to actually get around to burning an A3 12.10 CD and try it on the G3 (I think) upstairs over the weekend. It did run Xubuntu okay a couple of years ago, so fingers crossed :P [20:32] kewl! the automatic passwords part of chromium is working! (CentOS uses Chrome, so I was not sure if they would actually transfer).... Happy Bunny :) [21:09] wxl: please ping me next time you're about, thanks. [21:12] phillw: crazy day today and looking like it might last for the rest of the weekend [21:17] wxl: I know that feeling, just about got /home restored to Lubuntu 12.10 - next up is to the same for my CentOS system. anyways, do you have in mind a structure for people to post their PPC profiles up as per http://pastebin.com/7vMCqsTq onto the PPC wiki area? [21:30] phillw: not necessarily [21:33] wxl: I was thinking along the lines of "Model", "CPU", "GPU", "RAM", "Link to full Spec.", "Person who owns it".... ? [21:35] phillw: see mailing list message about it [21:36] wxl: that just states putting the pastebin up. for people to browse, I think it would be better to have the basic details on the page. === salem_ is now known as _salem [21:47] phillw: sure we can do that. i think what you mentioned above works === Ursinha is now known as Ursinha-afk [23:39] wxl: my rough draft for PPC hardware is at https://wiki.ubuntu.com/Lubuntu/Testing/PPC%26Mac64/HardWare please feel free to refine it! [23:41] looks good phillw but i will edit it. not sure nv34m necessarily equals geforce fxgo5200 which is what that card is more likely known as === Ursinha-afk is now known as Ursinha