[14:29] infinity: stgraber: so the nfs thing is blocking automated cdimage qa. Any updates, or is it blocked on the massive initramfs merge =/ [14:30] xnox: infinity said he'd take a look. In theory we can workaround it in casper by adding the 3 modules I mentioned the other day, so we could do that if it's urgent [14:31] stgraber: yes, please. It's not urgent-urgent, but high. Up to now we still don't have automated ISO testing on the same level as we had throughout quantal cycle. [14:32] xnox: do we have a bug number for this? [14:33] urgent> why not just add the relevant modules to initramfs-tools rather than either working around it in casper or waiting for the giant merge ... [14:34] stgraber: I don't believe we have one, apart from "jenkins bulb is red" [14:35] hmm, good point. Let me see how to add those to initramfs-tools directly, once I find a packaging branch that's up to date... [14:36] oh, it's just one revision behind, I can fix that easily [14:37] stgraber: note that the initramfs-tools version number is currently faked. [14:37] xnox: yeah, I just meant that ubuntu:initramfs-tools is at 0.103ubuntu0.1 instead of 0.103ubuntu0.2 [14:37] stgraber: oh _that_ revision =) [14:38] yeah QA are eager to get a working build. [14:38] oh fun, the branch is busted in a similar way as ifupdown... so won't be able to fix it [14:39] bug 1042396 [14:39] Launchpad bug 1042396 in bzr-builddeb (Ubuntu) "bzr-builddeb fails to import-dsc with native package on Ubuntu 12.04" [Undecided,New] https://launchpad.net/bugs/1042396 [14:39] will just do a manual upload [14:40] re urgent: it's blocking PS from running autopilot tests on hardware for raring. [14:44] * Laney wibbles [14:44] ./tests/run-pep8 gives a load of errors and fails the build for me locally, yet there aren't any commits (besides mine) since the release [14:44] (in ubiquity that is) [14:45] yeah, the new pep8 is complaining a lot more [14:45] Laney: get fixin ;-) [14:45] for example it now requires exact indentation when breaking a line [14:45] seems to be all about indentation [14:45] can it autofix those? [14:45] right, that's what I've noticed too [14:46] xnox: I doubt it, manually fixing those required quite a lot of magic for me in some other scripts [14:46] Perhaps --ignore= that for now [14:46] *sigh* [14:46] xnox: as you actually need to break strings in half now to make it happy [14:46] (you can no longer get away with doing print( and then the string at the beginning of the next line without indent as it'll report a no/or-missing indent) [14:47] hmm, the initramfs-tools versioning is terribly confusing... 0.103 was introduced in quantal with 0.103ubuntu0.2 which is a SRU looking version number, yet it never was a SRU [14:48] I guess I'll go with 0.103ubuntu1 then just to avoid messing with a potential SRU in quantal... [14:48] ah, or not, changelog suggests it was done on purpose... [14:49] * NOTE: Not an actual merge, just a cherry-pick and reving the [14:49] version number to avoid having to touch the Breaks in klibc-utils [14:49] Yeah, make it 0.103ubuntu0.3 or whatever it is please [14:49] ^^^ " stgraber: note that the initramfs-tools version number is currently faked." [14:50] alright, uploaded [15:19] is d-i/update-control the file to edit to change ubiquity BDs? [15:19] yes [15:19] * Laney got rather distressed when his changes were clobbered [15:19] ta [15:21] * cjwatson starts in on thegiant pile of pep8 corrections [15:22] just mentioning since there's probably no point in >1 person doing this [15:37] have fun :) [15:38] * stgraber gets back to fixing his other projects... [15:42] Laney: what are you coding for ubiquity? [15:42] =P [15:42] porting the webcam wotsit to gstreamer 1 [15:43] Laney: thanks a lot =) [15:43] yay, I won't be touched-it-last on that stuff anymore! [15:43] heh [15:43] it has Evan's name on it :P [15:43] * xnox is confused why ev is not all over the webcam plugin =)))))) [15:43] well, the last one who had to spend a week figuring out why the installer would hang on some laptops was me :) [15:44] * Laney fakes the commit logs [15:44] (tracked down to a regression in gstreamer causing some kind of infinite loop) [15:44] stgraber: because the rest of us do not click to hw enable webcam on laptops while testing =))))) [15:44] xnox: I don't actually have a single machine with a webcam builtin [15:44] just because I wrote it, that doesn't mean I should be held accountable for it ;) [15:44] xnox: I had to debug that one connected over ssh to London, working with gema's netbook [15:45] ev: yes it does ;) [15:45] shhhh you [15:45] * cjwatson tries to remember where the recommendation to indent closing brackets/braces in arrays/dicts came from [15:45] stgraber: 8-) [15:45] I always thought it was a bit bizarre but something somewhere enforced it and I thought it was PEP-8 [15:45] but I don't see it there any more [15:45] pylint maybe? [15:45] pyflakes ? [15:46] I recall that being used way, way, way back in the early days of ubiquity [15:46] or at least warnings being suppressed in comments [15:46] it will be hilarious if the new pep8 conflicts with pylint & pyflakes [15:46] everything conflicts with pylint [15:46] that's kind of the point [15:46] WARNING: THERE'S SOME CODE HERE [15:47] yeah, pylint just doesn't like you writting code, the only way to make something pylint-compliant is to not write it at all (IIRC pylint is happy when passed an empty file) [15:47] lol [15:47] I'm so amused that you've checked this [15:47] ev: maybe pylint should be one of the judges at the figure skating. [15:47] pyflakes doesn't seem to care [15:47] and I don't give a shit what pylint thinks :) [15:47] actually, no, it's not happy with an empty file, you need an empty file with a docstring apparently nowadays :) [15:48] C: 1,0: Missing docstring [15:48] pylint isn't very happy with pylint's code [15:48] stgraber: so what empty __init__.py is bad? [15:48] xnox: yep [15:48] lol [15:49] xnox: echo '""" This file is empty """' > __init__.py [15:49] then pylint is happy :) [15:51] aha, it's in the Launchpad Python style guide [15:52] that explains it [15:52] * cjwatson outdents a ton of stuff then [16:27] OK, ubiquity should be fine with pep8 now [20:18] cjwatson: hey, pep8 question for you as I'm guessing you've seen more of it than I did :) [20:18] parser.add_argument("--union-type", "-U", type=str, default="overlayfs", [20:18] choices=("overlayfs", "aufs"), [20:18] help=_( [20:19] "type of union (overlayfs or aufs), defaults to overlayfs.")) [20:19] pep8 now complains about the last line as it's outdented [20:19] the problem is that moving it to the line above will make the line length to be > 79 chars, so just trading pep8 errors really :) [20:19] did you get any of those in ubiquity? and if you did, how did you "fix" them? [20:20] an ugly hack would be to split into multiple strings and then just put them one per line below the help=_( block but that's ugly and I'm pretty sure gettext won't like it [20:26] (commited an ugly "fix" to LXC for now, but I'm really not fond of having to use string concatenation to workaround this...) [22:08] stgraber: splitting the string is the usual answer, and I don't see why gettext wouldn't like it as long as you concat by just putting the strings logically adjacent to one another rather than with + [22:08] I'm fairly certain I've seen that working with gettext all over the place elsewhere [22:09] you can also give yourself more space by using the hanging indent style rather than the visual indent style [22:10] which I initially found a bit ugly but now find much easier to read as it tends to be more consistent [22:10] and is definitely more practical with multiple nested indentations