[09:48] cjwatson: https://code.launchpad.net/~twom/launchpad/+git/launchpad/+merge/379201 is ready (rebased and resubmitted on master) [09:49] although I've not actually checked it with the buildd branch... [09:52] * tomwardill tries to work out how to do that [11:46] Folks, I'll top-approve https://code.launchpad.net/~pappacena/launchpad/+git/launchpad/+merge/379021, removing auditor (and django) from LP's code. Any objection? [11:48] dooooooo iiiiiiitttt [11:49] pappacena: no objection, but for the future, I'd personally love to see the kind of message you've got in the LP MP, in the commit message instead [11:50] oh though if it ends up in the merge commit I guess that works [11:50] (Mostly I'm thinking from a browsing-git perspective "ok this definitely removes auditor but _why_") [11:53] I guess only the "Commit message" field from LP MP goes to the git history, and not the description. In fact, I'm adding way more details in the field that *doesn't* go to the git history today. I'll start putting more details in the "commit message" field. [11:54] Oops yes [11:54] That feels like a slightly odd distinction [11:54] I think it throws me slightly every time [11:55] But yes, thank you [11:57] * tomwardill attempts to work out how buildd-manager works [11:58] Idle! [11:58] woo [12:02] tomwardill: reviewed [12:03] cjwatson: ta. Will look after lunch [12:04] commit message: I've taken to including full details in the actual git commit messages that are part of the branch I'm pushing, and normally only the first line in the MP's commit message that ends up in the merge commit, as otherwise it gets pretty duplicative [12:06] Ah; I've found myself using `git log --merges` a lot in a few places, so really liked the duplication [12:06] (but that's a purely personal foible) [15:35] cjwatson: do you know of existing examples of TrialTestCase with pushConfig and similar LP test methods? [15:41] oh,. might just be able to lift the pushConfig method, I think that's the only one I need [15:56] tomwardill: MakeSnapBuildMixin.makeSnap does exactly this [15:57] ah, so it does [15:57] I was looking for instances of the `pushConfig` method [16:01] Could I ask for some py3-related reviews? [16:01] https://code.launchpad.net/~cjwatson/lazr.restful/newer-grokcore-martian/+merge/379034 and https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/379036 are various dependency upgrades [16:01] https://code.launchpad.net/~cjwatson/lazr.restful/py3-division/+merge/379205 is an easy bit of porting [16:02] https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/378460 is a rather larger bit of porting; the top commit in that branch is very large and formulaic, but https://git.launchpad.net/~cjwatson/launchpad/commit/?id=7e44f2849205ce68c9d6e2987874f3d7536c7429 is worth reviewing separately [16:05] cjwatson: how silly an idea would be to use the 'current' series of the distribution as the distro_series of an OCIRecipeBuild? (Ideally, I think just selecting the latest LTS would be a reasonable approach, but I can't figure out if that's possible) [16:08] (Are people generally getting MP review requests by email? I feel people mostly don't look unless I explicitly ask) [16:09] tomwardill: It's not ideal because it usually won't be an LTS and because it couples OCI builds to a not-very-related piece of process. It would be OK for a first pass to unblock you, as long as it has a great big XXX on it and we commit to figuring out something better soon [16:10] I get them by email, but I think I need to make my filter smarter to avoid them getting lost in the snapstore ones [16:10] Mm [16:11] I should also check that folder more often tbh [16:12] * tomwardill pokes at filter [16:13] Yeah - I dislike nagging too much and I figure my responsibility is more to review than to be reviewed, but at the same time I don't want things I do to be stalled [16:13] yeah, I'll fix my filter and make a point of checking it more often [16:14] Cool [16:17] well, I would, if I could make gmails UI behave [16:31] ooh, it works, and it even told me about your comment reply [16:45] cjwatson: that error handling change makes it much nicer [16:46] * tomwardill has run into that urlopen thing before [16:47] tomwardill: which change, sorry? [16:47] oh, in six-urllib? [16:47] https://git.launchpad.net/~cjwatson/launchpad/commit/?id=7e44f2849205ce68c9d6e2987874f3d7536c7429 [16:48] that is much nicer than parsing the length of args [16:48] Yeah, it looked less awful this way [16:48] requests would probably be better still, but one thing at a time [16:48] heh