[05:58] <smoser> :-(
[05:58] <smoser> https://code.launchpad.net/~cloud-init-dev/+recipe/cloud-init-daily
[05:58] <smoser> trying to set up a git recipe
[05:58] <smoser> "Recipe Contents" shows : # bzr-builder format 0.4 deb-version {debupstream:ubuntu-pkg}-{debversion}+trunk
[05:58] <smoser> I edit it, change that to
[05:59] <smoser> # git-build-recipe format 0.4 deb-version {debupstream:ubuntu-pkg}-{debversion}+trunk
[05:59] <smoser> hit save
[05:59] <smoser> it goes back to bzr-builder.
[06:11] <wgrant> smoser: Hm, I'm not sure it's possible to change a bzr recipe into a git one or vice-versa.
[06:12] <wgrant> It seems like you've recreated it now?
[06:12] <smoser> yeah. i think it is sorted now
[06:12] <smoser> i had eve deleted it
[06:12] <smoser> and made a new one
[06:12] <smoser> but it was sticky
[06:12] <smoser> i dont know. maybe luser error too
[06:13] <smoser> wgrant, is my build log failure obvious to you ?
[06:13] <smoser> https://launchpadlibrarian.net/277898914/buildlog.txt.gz
[06:14] <smoser> (and second, a feature request for {git_describe} as a substitution variable)
[06:15] <wgrant> smoser: The type of the recipe is determined by whether you register it from a Bazaar or Git branch page.
[06:16] <wgrant> ValueError: Invalid version string '{debupstream:ubuntu-pkg}-0.7.7~git1029-gb56d7a1-0ubuntu1+trunk~ubuntu16.10.1'
[06:16] <wgrant> Looks like debupstream isn't being expanded, hm.
[06:17] <wgrant> The only slightly unusual thing is the hyphen in "ubuntu-pkg"
[06:17] <wgrant> It'd normally just be "packaging".
[06:17] <smoser> right. i thought htat might be what was tripping it up.
[06:17] <smoser>  it is valid
[06:17] <wgrant> It's possible there's a bug there, but it'd be odd.
[06:18] <smoser> i'd like to use 'git describe's output for my "upstream version" when i make a snapshot and such.
[06:18] <smoser> 0.7.7~git1029-gb56d7a1-0ubuntu1
[06:18] <smoser> is a valid albeit odd version.
[06:18] <smoser> but my plan is to get to
[06:18] <smoser>  0.7.7-1-gHASH-0ubuntu1
[06:19] <wgrant> Yeah, it's similar to the bzr-builder latest-tag substitution
[06:19] <wgrant> What's that -1 doing in there?
[06:19] <smoser> thats commits since last tag
[06:19] <wgrant> Ah
[06:19] <smoser> LASTTAG-COMMITS_SINCE-gHASH
[06:19] <smoser> that is what git describe will output
[06:20] <smoser> and that just seems quite sane as a thing to use when you make a snapshot upload
[06:20] <wgrant> https://launchpad.net/git-build-recipe/+filebug
[06:20] <wgrant> latest-tag is already implementing using git-describe, albeit with --abbrev=0
[06:27] <smoser> wgrant,
[06:27] <smoser>  Then I'd do something
[06:27] <smoser> bah
[06:27] <smoser> https://bugs.launchpad.net/git-build-recipe/+bug/1611222
[06:27] <smoser> and
[06:27] <smoser>  https://bugs.launchpad.net/git-build-recipe/+bug/1611219
[07:33] <cjwatson> wgrant: phase A> that will be interesting, thanks.
[07:57] <wgrant> cjwatson: Doesn't solve everything, but removes most of the scaling by package volume and count.
[12:01] <quyhoang> hello
[13:56] <teward> should I be concerned that most of the arm builders are at 'cleaning'?
[14:08] <teward> oop there it went
[14:08] <teward> (it looks like it was a long clean cycle0
[14:13] <cjwatson> The reset operation fairly often times out or otherwise fails, but the nature of the problems is predictable enough that we have a watchdog script to try again.
[14:13] <cjwatson> So they generally sort themselves out.
[14:15] <teward> indeed.
[14:16] <teward> looks like it's fine now, it was just sitting for a while in 'cleaning' and had a 30 minute queue, when i had last refreshed the builders page.  WOrking now.