[04:56] <pitti> Good morning
[08:38] <jibel> balloons, all the ubiquity tests are failing because the environment is not setup correctly when ubiquity is started with --autopilot. There has not been any change in ubiquity recently so it might as well completely fail to start. I'll have a closer look today.
[09:11] <pitti> jibel: we got a new autopilot into utopic today
[09:17] <jibel> pitti, I don't think it's the problem because /var/log/installer/autopilot doesn't exist and is created when ubiquity starts with --autopilot enabled. So apparently ubiquity completely fails to start.
[09:18] <pitti> jibel: btw, I asked on d-devel@ and autopkgtest-devel@ about autopkgtest-xenlvm users, and that I'll drop it if there's no interest
[09:20] <jibel> pitti, good. I think I fixed most of shstring support and it's now working with ssh. I did some tests with lxc and I'm currently testing with qemu to make sure I didn't broke them and will propose the fix for review.
[09:22] <pitti> \o/
[09:22] <pitti> jibel: I'll upload 2.17.1 today to fix some bugs and the autopkgtest, so please let me know if you have shstring fixes to review/commit
[09:24] <jibel> pitti, http://paste.ubuntu.com/7586202/
[09:24] <jibel> I'll propose a proper MP
[09:25] <pitti> jibel: don't worry about an MP, the package is in git anyway
[09:26] <jibel> pitti, how do I do the equivalent of a rebase with qit?
[09:26] <jibel> git
[09:26] <pitti> jibel: you mean you have your local tree checked out and want to update it to what I committed recently? git pull --rebaes
[09:26] <pitti> git pull --rebase
[09:26] <pitti> which is a shortcut for "git fetch origin; git rebase origin"
[09:27] <jibel> ty
[10:35] <pitti> jibel: saw the Debian bug, thanks!
[11:00] <pitti> jibel: would you mind fixing doc/README.virtualisation-server along with the new intended usage, or want me to?
[11:04] <pitti> jibel: what I don't understand is why in the case of specifying a shstring the auxverb is the same?
[11:05] <pitti> jibel: or does it just happen to work with ssh and adt-xenlvm-on-testbed?
[12:01] <jibel> pitti, because everywhere downs['auxverb'] is used shell quoting is already done, so extra quoting in the auxverb command is not necessary. Re-reading the code, we might need to add quote cmd_string in execute() too
[12:03] <jibel> I'll fix the doc
[12:20] <pitti> jibel: thanks, I'll wait for these fixes then
[12:20] <pitti> so far it looks good, all tests still pass
[12:22] <pitti> jibel: I'll also ponder moving execute_raw() into the protocol and thus into lib/VirtSubproc.py; it would be much better suited there, and then I think we don't need to expose print-*-command at all, right?
[12:58] <jibel> pitti, you mean execute_raw from the class Testbed in adt-run?
[12:58] <pitti> jibel: right
[12:59] <pitti> jibel: if that makes things easier for you (as that entire VirtSubproc API can then go away) I can look into that now
[13:00] <jibel> pitti, indeed, it'd be better and less confusing, but it's working as it is now so it's probably low priority
[13:00] <pitti> jibel: I recently added that execute_raw(), but it probably wasn't thought out very well
[13:01] <pitti> with the thing your patch changes it'll be much easier
[13:01] <pitti> jibel: right, I can look at that after your patch, too
[13:20] <cgoldberg> robotfuel, hey... do you know subunit well?  I'm working with result details (file attachments) in subunit streams
[13:21] <robotfuel> cgoldberg: I don't know a lot, you probably have to convert to bits?
[13:22] <cgoldberg> robotfuel, i've got it working with attaching images (as bytes).. I'm javing a little trouble in parsing the output
[13:23] <robotfuel> cgoldberg: do you have a branch I can look at? you need to encode and decode the results.
[13:23] <cgoldberg> robotfuel, 1 sec.. I'll paste a snippet to give you context
[13:25] <cgoldberg> robotfuel, http://paste.ubuntu.com/7587461/
[13:26] <cgoldberg> robotfuel, i have one test case that attaches an image ... it looks to me like as I retrieve it, i only get a chunk of the file at a time... so I prob need to accumulate all the chunks to get the full image
[13:29] <cgoldberg> robotfuel, what I'm trying to do is take a saved subunit stream, parse any attachments that contain data, save bytes as individual files
[13:54] <cgoldberg> robotfuel, i think I figured it out... subunit has "chunked.Decoder(output)" to decode chunked streams
[14:07] <pitti> jibel: I pushed your shstring fixes, so you can rebase
[15:49] <senan> balloons, danchapman : hey
[15:49] <balloons> senan, hey how are you!
[15:49] <DanChapman> hey senan, how are you?
[15:50] <senan> balloons,DanChapman : I'm good thank you
[15:50] <senan> how about you ?
[15:50]  * DanChapman waves to balloons
[15:50]  * balloons waves back :-)
[15:50] <balloons> I'm a little under the weather.. had a bit of an accident and still recovering
[15:51] <senan> balloons, accident ?
[15:52] <balloons> senan, yes a kiting accident. I got picked up and dropped
[15:53] <senan> balloons, I got transferred to Kerala
[15:53] <elfy> balloons: should have let go ...
[15:53] <elfy> #justsaying
[15:54] <balloons> senan, ahh.. just in time for monsoon season
[15:54] <balloons> elfy, hah! I eventually pulled the pin to separate
[15:54] <elfy> :)
[15:54] <elfy> well I hope you're recovering ok :)
[15:54] <elfy> I prescribe a nice bottle of red
[15:55] <senan> balloons, Nops :) . Permanent transfer. Kerala is my home state
[15:55]  * slickymasterWork thinks that balloons was hearing that music "I'm learning to fly, but I ain't got wings..."
[15:55]  * brendand has a glass in hand!
[15:55] <brendand> balloons, ya ok now?
[15:56] <elfy> see - we do care :p
[15:59] <senan> balloons, the calender bug you asked me to look into is resolved right
[16:09] <senan> balloons, I would like to work on gnumeric test
[16:10] <balloons> brendand, really bruised and sore.. And still a little sick from last week, heh.. My body is not happy with me!
[16:10] <balloons> elfy, brendand thank you for the well wishes!
[16:10] <balloons> senan, a gnumeric test.. cool.. Are you able to introspect gnumeric?
[16:11] <senan> balloons, I didnt try. But in the  bug details its written it introspect ok
[16:12] <balloons> awesome.. should be good then :-)
[16:12]  * elfy wonders if there is a bug for that gnumeric thing
[16:13] <elfy> also just mentions in passing that lderan was looking at xubuntu a/p tests - gnumeric being one
[16:15] <balloons> elfy, ahh yes good point
[16:15] <senan> balloons,you only wrote that comment :P
[16:15] <elfy> someone appears to be working on it
[16:15] <elfy> bug 1115588
[16:16] <senan> elfy,ya I saw it just now :(
[16:17]  * elfy nonchalantly mentions in passing bug 1115601 and then whistles tunelessly as he wanders off again
[16:17] <elfy> that's of course if senan likes the idea of not just looking at Ubuntu stuff :)
[16:19] <brendand> balloons, i might have a crack at something along the lines of 'writing a good test'. or maybe we already have guidelines for that?
[16:21] <balloons> senan, hah.. funny I wrote the comment. I was just trying to help myself out in the future too
[16:21] <senan> elfy, balloons, ok then I'll look into Abiword
[16:22] <balloons> brendand, excellent. As far as what we have available now, there's the AP docs talking about AP tests, there's guidelines for manual testcase writing, and there's some other tidbits as well. Are you thinking of making it generic?
[16:22] <balloons> brendand, you can always turn it into a contribution as well.. We can host it on developer.u.c
[16:22] <brendand> balloons, show me all the tidbits and i'll see if it's worth doing
[16:23] <balloons> brendand, http://unity.ubuntu.com/autopilot/tutorial/good_tests.html
[16:24] <balloons> brendand, https://docs.google.com/a/canonical.com/presentation/d/12szRruujCFmnaJgDh01S4maicfLLW0DBQ-mqClazPd4/edit#slide=id.p
[16:25] <balloons> and finally, http://developer.ubuntu.com/apps/platform/guides/acceptance-testing-using-the-page-object-model/
[16:25] <balloons> that's probably the most useful of the bunch.. There's also thomi's blog posts. Let me dig those up
[16:26] <balloons> brendand, http://www.tech-foo.net/on-test-levels-and-coverage.html
[16:31] <brendand> balloons, i'll save the autopilot bits for someone else
[16:31] <brendand> balloons, but i can roll some of the other stuff up into one presentation
[16:33] <balloons> brendand, I think something generic could work well.. awesome, thank yu
[16:34] <brendand> balloons, i have a week to prepare right ?
[16:36] <balloons> brendand, yep. It starts next Tuesday.. you can delay until June 12th if you wish which is the last day :-)
[16:36] <balloons> just propose the session and I can get you the timeslot and day you want
[16:48] <brendand> balloons, which track?
[18:45] <brendand> balloons, i missed your reply - which track should i propose the session for?
[18:49] <elfy> there was no reply :)
[19:45] <balloons> brendand, the users tracl
[19:51] <elfy> I thought it was - but wasn't quite sure enough to say :)
[20:20] <brendand> balloons, you think it's a useful topic for users?
[23:07] <balloons> brendand, if there's a better track, pick it :-)