[03:07] <chrissbx> Hi. I've recently imported a git repo to launchpad, using the official import functionality. I choose the Git master branch and called it trunk locally.
[03:07] <chrissbx> Now I've got a second branch in the Git repo, and would like to import that, too. But somehow I just can't find that import functionality in the web UI anymore.
[03:09] <wgrant> chrissbx: Launchpad only gained support for importing non-master git branches a couple of weeks ago, and the UI isn't exactly obvious yet. If you go to the Code index for your project, you should see an 'Import a branch' link on the right.
[03:10] <wgrant> chrissbx: Until we have a special UI for this, you can import a non-master branch by adding ',branch=somebranchname' to the end of the git URL.
[03:15] <chrissbx> Hm, I can't find that "import a branch" link anywhere. This is the project that's already being imported. On code.launchpad.net/~me/myproject/trunk it shows me stats on the import.
[03:15] <wgrant> chrissbx: code.launchpad.net/myproject
[03:16] <wgrant> code.launchpad.net/~me/myproject/trunk is the page for that particular branch, not the project as a whole.
[03:17] <chrissbx> Hm, problem is that I'm having a fork; https://launchpad.net/redshift is showing all the forks, but isn't "my page" afaict, anyway there's no "import" link.
[03:17] <wgrant> chrissbx: Click the "Code" tab at the top.
[03:18] <chrissbx> Ah, there.
[03:31] <wgrant> chrissbx: Imports from GitHub work better over git:// than https://. I've changed the URL, and it has now imported successfully.
[03:33] <chrissbx> Ah, thanks.
[15:23] <wzssyqa> when dput to ppa Changes file must be signed with a valid GPG signature: Verification failed 3 times: ['General error', 'General error', 'General error'] : Permission denied.
[15:23] <wzssyqa> why?
[15:24] <bigjools> wzssyqa: known problem, it will be ok, there's a bug at the server end
[15:24] <wzssyqa> bigjools: many thanks
[15:24] <bigjools> wzssyqa: are you uploading to ppa.launchpad.net?
[15:25] <wzssyqa> bigjools: yes
[15:25] <bigjools> ok thanks
[15:36] <geser> bigjools: is there a bug for this I can point people at?
[15:37] <bigjools> yes, one sec
[15:37] <bigjools> bug 798957
[16:06] <abentley> deryck[lunch]: I relieve you.
[16:11] <ahasenack> timeis somebody working on that bug? It also broke all our automated pppa uploads
[16:11] <ahasenack> s/^time//
[16:11] <bigjools> ahasenack: it can't have broken anything, it doesn't affect the upload
[16:12] <ahasenack> bigjools: I was told our jenkins upload caught the error and stopped
[16:12] <ahasenack> bigjools: the job is "red", so dput exits with non-zero
[16:13] <ahasenack> easy enough to workaround for a while, of course (|| /bin/true)
[16:13] <bigjools> ahasenack: ignore "general error"  I think
[16:38] <Darxus> https://code.launchpad.net/~vcs-imports/spamassassin/trunk "The import has been suspended because it failed 5 or more times in succession." - the error is "ValueError: too many values to unpack", from http://launchpadlibrarian.net/81067671/vcs-imports-spamassassin-trunk.log
[16:38] <Darxus> Looks like a launchpad bug?
[16:39] <Darxus> This failed for 5 days before I was notified, is there a way I can get notified if automated imports have failed for something more like 24 hours?
[16:47] <Darxus> Can I be emailed every time an import for lp:spamassassin fails?
[16:55] <deryck> thanks, abentley
[18:01] <abentley> deryck: Could you please run https://pastebin.canonical.com/53397/ on staging?
[18:01] <deryck> abentley, sure.
[18:03] <deryck> abentley, https://pastebin.canonical.com/53399/
[18:06] <abentley> deryck: Could you run this variation please? https://pastebin.canonical.com/53401/
[18:07] <deryck> abentley, https://pastebin.canonical.com/53402/
[18:08] <abentley> deryck: That is the number of branches in loom format that need an upgrade.  AFAICT, loom doesn't natively support upgrades.
[18:09] <dobey> sinzui: ping
[18:10] <sinzui> hi dobey
[18:11] <dobey> sinzui: hey. bit of a technical question about having a "distribution" added in launchpad; what would the consequences be of having say "Microsoft Windows" being a "distribution" in launchpad be?
[18:12] <sinzui> dobey, lots of screwups along the lines of gentoo and redhat because LP only knows Debian package names
[18:12] <benji> that and the universe would explode from an irony overload
[18:13] <dobey> sinzui: is there any better way to have an OS in LP that isn't debian? so we can separate "windows" bugs from all our ubuntu/ppa/upstream bugs?
[18:14] <sinzui> dobey, no, we do not support it. The gentoo and redhat registered distros are pretty crippled
[18:14] <niemeyer> Hey folks
[18:14] <niemeyer> Can someone confirm that this regexp is correct for LP usernames:
[18:14] <niemeyer> "[a-z0-9][a-zA-Z0-9+.-]+"
[18:15] <Darxus> niemeyer: If that's a perl regex you need to escape the .
[18:17] <dobey> sinzui: crippled is fine. i just want to assign windows bugs to windows in a somewhat sane way. tags do not seem the sane way to me
[18:17] <sinzui> dobey, setting aside that Lp does not know how to handle this. Consider the case of the gimp master branch that clearly supports linux and windows. Lp will honour that branch as a source package branch for a "windows-distro" but it could never be built. the a bug in the code can obviously be in windows and linux
[18:17] <dobey> would be really great if a bug had a list of platforms it affected, like bugzilla does
[18:18] <dobey> to make distinction between platform and distribution
[18:18] <sinzui> dobey, No it would not
[18:18] <sinzui> we allowed users to report bugs in any distro, that lead to user thinking they were actually reporting bugs to users who could fix them
[18:18] <sinzui> We do not permit users to report bug against distros that do not officially use Lp
[18:18] <dobey> i do not mean as a distro
[18:19] <sinzui> right, as a record of affected somethings
[18:19] <dobey> i mean "this is affects only 'linux amd64'" for example
[18:20] <dobey> hrmm
[18:20] <tumbleweed> are tags not sufficient for that?
[18:20] <sinzui> dobey, Lp will require the package to be official as well as the distribution officially uses Lp to track bugs
[18:21] <niemeyer> Darxus: No, that wasn't the question..
[18:21] <niemeyer> Darxus: The question was whether this matches the implementation of Launchpad
[18:21] <niemeyer> Darxus: The regex is actually correct for perl/python/etc
[18:22] <dobey> tumbleweed: tags suck as a way for doing that; at least in the case of windows.
[18:22] <niemeyer> Darxus: dot doesn't have to be escaped within a char set
[18:22] <Darxus> Huh.
[18:26] <dobey> sinzui: i suppose we could create an "Ubuntu One Windows" distribution, and just use it for our stuff?
[18:28] <sinzui> dobey, That will work if you set it to use Lp bugs and you set the official package branches
[18:28] <sinzui> dobey, distro's use an API script to set the branch of the official package
[18:30] <dobey> sinzui: ok, thanks.
[23:28] <Sweetshark> so why does launch pad not let me make 807759 an duplicate of 745836? and why does it not tell me what it doesnt like about it?
[23:33] <wgrant> Sweetshark: Bug #745836 doesn't exist, or is private.
[23:33] <wgrant> Ah.
[23:33] <wgrant> Or I can't type.
[23:35] <wgrant> Sweetshark: 807759 has duplicates, some of which you possibly can't see, so you can't change their duplicate marking.
[23:37] <Sweetshark> wgrant: ok -- they are dupes of 745836.
[23:37] <Sweetshark> Just trust me ;D
[23:37] <wgrant> I think I have access to all of them...
[23:37] <wgrant> Let me try.