[13:23] <mvo> can anyone suggest a workaorund for https://code.launchpad.net/~snappy-dev/snappy/trunk-github ? it fails right now with: bzrlib.plugins.git.errors.UnknownCommitExtra: Unknown extra fields in <Commit ace271dbbf103122fd56f186c1e3fd0708b6a905>: ['gpgsig'].
[13:24] <mvo> I was wondering if a code-import from github->lp-git would work but it seems like I can not ask for such a code import directly
[14:22] <cjwatson> mvo: We'd like to implement git-to-git mirrors, and there's some hope of getting it onto the schedule soon, but the only possible workaround right now is to mirror manually.
[14:23] <mvo> cjwatson: hm, I guess I could remove the tag with the gpgsig also? would that help?
[14:24] <cjwatson> mvo: I think that's a signed commit, not a signed tag.
[14:26] <cjwatson> mvo: It has to be nowhere in the history of the branch you're trying to import, so it may not be sensibly possible for you to undo ...
[14:26] <cjwatson> Unless it's like the most recent commit or something.
[14:28] <mvo> cjwatson: its probably in between, but I will try to figure out which one. I was suspecting a signed tag because that is what I do when I release. I will try to find out if there is a signed commit somewhere
[14:29] <mvo> cjwatson: thanks for your help, I will scratch my head a bit over this
[14:30] <clivejo> hi guys, Launchpad seems to be taking an hour plus to publish, is there something wrong?
[14:30] <wgrant> Just investigating that.
[14:31] <cjwatson> The PPA publisher seems to be chugging along.
[14:31] <wgrant> Very slowly.
[14:31] <wgrant> I'm getting 100KB/s haetae to carob.
[14:32] <cjwatson> Yeah
[14:32] <cjwatson> 2016-02-12 14:31:35 DEBUG   Added /srv/launchpad.net/ppa-archive/ci-train-ppa-service/landing-000/ubuntu/pool/main/m/mir/mir-demos_0.19.2+16.04.20160212.1-0ubuntu1_amd64.deb from library
[14:32] <cjwatson> 2016-02-12 14:31:48 DEBUG   Added /srv/launchpad.net/ppa-archive/ci-train-ppa-service/landing-000/ubuntu/pool/main/m/mir/mir-doc_0.19.2+16.04.20160212.1-0ubuntu1_all.deb from library
[15:17] <sophomeric> I have some packages in pending publication status. How long does it typically take for this to complete?
[15:21] <dobey> typically, not very long
[15:21] <dobey> currently, the publisher is running a bit slower than typical
[15:22] <sophomeric> It's been a little shy of an hour. I know I'm being impatient but I thought I'd ask.
[15:24] <Laney> There's some kind of network trouble going on which seems to be causing things to back up
[15:25] <sophomeric> I'll just wait until Monday then since it's nearly the end of the day here. Thanks for the update.
[15:26] <cjwatson> It'll hopefully be sorted out soon
[20:52] <clivejo> any update on the publisher slowdown/stop?
[20:55] <dobey> it's still slow
[20:56] <dobey> incredibly slow
[21:22] <Saviq> is there +junk support for git in launchpad?
[21:22] <cjwatson> Saviq: yes, but it's spelled +git
[21:23] <cjwatson> Saviq: https://help.launchpad.net/Code/Git#Repository_URLs  see the bit about personal repositories
[21:23] <Saviq> cjwatson, thought ~saviq/+git/project would be the equivalent of ~saviq/project in bzr...
[21:23] <cjwatson> Saviq: ~saviq/project doesn't mean anything in bzr in isolation, please clarify :)
[21:24] <cjwatson> Saviq: in "~saviq/git/repository", the "repository" bit has no connection to any project name except by coincidence
[21:24] <Saviq> cjwatson, if I have a ~saviq/project/branch in bzr, wouldn't a corresponding branch in git be "branch" in ~saviq/+git/project?
[21:25] <cjwatson> Saviq: a project repository would be "~saviq/project" (your default repository for that project; use that unless you have a reason not to) or "~saviq/project/+git/repository" (separate named repository; this is for special cases)
[21:25] <Saviq> cjwatson, ok, misunderstood "personal" in this case
[21:25] <cjwatson> it's a little odd but we needed to do it this way for URL consistency with other parts of LP
[21:26] <cjwatson> Saviq: I've added a mention of +junk to the help page, to assist people familiar with bzr
[21:26] <cjwatson> modulo wiki saving ever getting back to me
[21:27] <Saviq> cjwatson, can a git repository ever be ~saviq/project/foo?
[21:27] <cjwatson> Saviq: no
[21:28] <Saviq> or ~saviq/project, or ~saviq/project/+git/foo
[21:28] <cjwatson> it can be either of those last two
[21:28] <Saviq> ok
[21:28] <Saviq> so we got one more level of depth compared to +junk, works for me
[21:28] <cjwatson> ~saviq/project/foo was dodgy URL design with bzr codehosting, it achieved brevity at the expense of permanently colonising an important swathe of namespace
[21:29] <Saviq> cjwatson, agreed
[21:29] <cjwatson> we couldn't use it because it would have been ambiguous
[21:29] <cjwatson> or else the webapp URLs would have had to differ from the git.l.n URLs
[21:29] <Saviq> cjwatson, that clears it up, thanks!
[21:29] <cjwatson> np
[21:31] <Saviq> cjwatson, oh one more question, I was always more used to pull, as opposed to merge, requests with git, is that a design decision to keep them MPs, or just TODO?
[21:32] <dobey> that just seems like nomenclature differenct to me
[21:32] <cjwatson> Saviq: well - you can only pull if you've rebased first, which is a project decision
[21:33] <cjwatson> Saviq: merging is still perfectly legit in git, and many projects do that
[21:33] <Saviq> cjwatson, sure, which is why I said "was more used to" :)
[21:33] <cjwatson> Saviq: e.g. if you look at linux, most commits by Linus are merges :)(
[21:33] <cjwatson> :)
[21:34]  * Saviq always liked the flat history, but it's probably unwieldy in big projects
[21:35] <Saviq> or plain not useful - better to be able to drill down from a merge
[21:35] <Saviq> instead of ploughing through all the commits on one level
[21:35] <dobey> i prefer to call it "fake history" than flat history ;)
[21:36] <Saviq> better than "merge trunk" being half of that history ;P
[21:37] <Saviq> but I guess one does not cause the other
[21:37] <dobey> if merge trunk is happening too much, then that should signify there is something wrong with the development process, or the understanding of the developers about how branches work
[21:39] <Saviq> well, s/trunk/prereq-branch/
[21:42] <dobey> sure. i think my statement works for either case. :)
[21:43] <Saviq> it's just we're touching the same areas of code a lot these days, and can't land often enough, meaning we get into conflict dance all the time
[21:48] <dobey> sure. not blaming you. just saying i think flattening commits is a non-solution to the process issues.
[22:04] <cjwatson> PPA publishing issues should hopefully improve very shortly
[22:04] <dobey> sweet
[22:51] <jose> 'ello! I was wondering, when creating a Launchpad account/ubuntu one id, does it require you to verify your email address?
[22:57] <wgrant> A Launchpad account does, an Ubuntu One account does not.