[00:49] blr: What's the status of the loggerhead and go get branches? [00:51] Oh [00:51] The loggerhead branch landed a minute before I asked? [00:51] Between loading +activereviews and the actual MP... [00:51] Nice timing! [01:01] wgrant: just had a look at BranchRefNavigation, so presumably I want context.development_focus.unique_name? [01:03] blr: Yep, I think that's right. [01:03] The aliases aren't resolved by the codehosting HTTP server, unlike the git case. [02:07] blr: Is the go branch ready now? [02:22] wgrant: yep, I think so, want to have a look? [02:23] blr: I'd suggest reworking the tests to verify that it actually ends up on the page. [02:24] blr: Create branches and repos with known paths and then check that the meta tag appears in the HTML. [02:24] wgrant: ok, worth keeping the unit tests as well? [02:25] will add a soupmatcher [02:25] blr: I'd probably do both. Construct the branch/repo with a known path, create the view, check the property, and then render the view and check the content to verify that the template worked. [02:25] Yep [03:17] wgrant: do you have a sec to look over those tests again please? [03:21] hmm perhaps would be good to have a negative test with no vcs default set. [03:24] Indeed, worth checking the tag isn't there at all. [03:24] And default VCS but no default repo or no default branch depending on the VCS. [04:38] wgrant: added those tests, thanks. [04:38] mwhudson: would you like to review the branch? [07:32] blr: er, seems like i'm too late? [07:37] it looks ok to me [07:38] i'll have a poke at it when it hits qastaging [07:38] i forget how often it updates... [07:41] mwhudson: Automatically, and it will try every 15 minutes [07:41] ah [07:41] from stable? or db-stable? [07:42] qastaging from stable [07:44] half an hour or so then [08:16] mwhudson: great, thanks [08:17] blr: hrm, shouldn't there be appropriate tags on view-source:https://qastaging.launchpad.net/pydoctor ? [08:17] or is the "qastaing doesn't really have the branches" thing [08:26] mwhudson: You'll need to set the Product.vcs property until we backfill it. Hit the Code config link in the sidebar and hit save without making any changes. [08:26] That should make it appear. [08:27] ah, i loaded the page and saw bzr already selected [08:27] cool [08:28] hm :( [08:28] mwhudson@glamdring:linux_arm64_dynlink$ go get -v qastaging.launchpad.net/pydoctor [08:28] Fetching https://qastaging.launchpad.net/pydoctor?go-get=1 [08:28] Parsing meta tags from https://qastaging.launchpad.net/pydoctor?go-get=1 (status code 200) [08:28] import "qastaging.launchpad.net/pydoctor": parse https://qastaging.launchpad.net/pydoctor?go-get=1: no go-import meta tags [08:28] package qastaging.launchpad.net/pydoctor: unrecognized import path "qastaging.launchpad.net/pydoctor" [08:28] doesn't look happy [08:30] Oh [08:30] content="qastaging.launchpad.net/~mwhudson/pydoctor/github bzr http://bazaar.qastaging.launchpad.net/~mwhudson/pydoctor/github" /> [08:30] there's the problem [08:31] hm? [08:31] The first bit should be just qastaging.launchpad.net/pydoctor [08:31] ah [08:31] yeah [08:31] blr: ^^ [08:31] oh right [08:32] confusing message :/ [08:33] hey, um, the configure code link doesn't appear after you change the default to git [08:33] is that a known bug? [08:37] wgrant: hmm, so it should be, will fix. [08:38] mwhudson: sorry, check again tomorrow morning :0 [08:39] mwhudson: Yep, for now you have to get there from the sidebar on +index [08:52] wgrant: mwhudson: fix here once the diff renders, https://code.launchpad.net/~blr/launchpad/golang-meta-import-fix/+merge/263625 [08:53] blr: Does it fix the series case too? [08:54] wgrant: yes [08:58] wgrant: if there's a bug for the missing link, could you assign it to me (sounds like a regression from my configure-code branch?) [09:04] mwhudson: will ping you when the fix is on qastaging. [09:14] blr: thanks [09:26] blr: Nope, I just never added the link to +git. [09:26] (because of the awkwardness around showing portlets for the default git repo too) [09:31] Ooh, first trimmed-to-context inline comment mail I've seen. Nice. [09:32] Hm, looks slightly off though [09:33] blr: Would you like me to have a look at the dodgy quoted patch headers near the top of http://paste.ubuntu.com/11809068/ ? Must be near your EOD [09:33] There's an extra "=== modified file 'lib/lp/registry/browser/productseries.py'" for some reason [09:40] Oh dear, that looks like a bzrlib bug [09:40] http://paste.ubuntu.com/11809104/ [09:50] But it doesn't throw the line numbers off? Weird. [09:51] Just working on setting up a bzr test case [10:09] wgrant: btw [10:09] wgrant: have you noticed the Canonical internal channel references on the sso recovery help page [10:10] wgrant: https://login.launchpad.net/+device-help [10:10] lifeless: Yes [10:10] We don't maintain SSO, and there is a bug reported AIUI. [10:10] ok cool [10:10] I know you don't [10:10] but you're closer to the issue :) [10:47] blr,wgrant: https://code.launchpad.net/~cjwatson/bzr/fix-keep-dirty/+merge/263632 [10:47] If you're OK with that, I can stuff it into 2.6.0.lp.2 for the time being [10:53] cjwatson: lgtm, thanks [10:58] wgrant: I don't know the bzr landing process. Is it supposed to wait for Richard? [10:59] (for trunk, that is, I'll go ahead with .lp.2 nowish) [11:00] cjwatson: Anyone with PQM access (eg. vila) can land it. [11:01] vila: oh hai :-) [11:04] cjwatson: hey ! /me looks [11:19] cjwatson: reviewed, minor nit and it's good to land [11:30] Ah, assertLength, nice. [11:31] vila: Fixed, thanks [11:59] cjwatson: thanks, sent to pqm [12:00] Yay, thanks [12:00] * cjwatson gets that into LP [12:03] Hopefully pqm actually works at the moment === frankban_ is now known as frankban [15:15] cjwatson: err, pqm... lacks hope ? [15:15] http://pqm.bazaar-vcs.org/ is up though [15:15] vila: Richard Wilbur reported that it didn't work the other day; we sent him to the public RT [15:16] cjwatson: I missed that :-/ What's the public RT ? [15:16] rt.ubuntu.com, but I don't see a ticket from him there [15:16] oh, https://rt.ubuntu.com/Ticket/Display.html?id=26672 [15:17] vila: failing that, looks like you'll need to ask webops to have a look, perhaps referencing that ticket [15:19] cjwatson: thanks ! will do [16:13] Laney: I'm sure that, absent a heatwave, it would have taken me a considerably shorter time to sort out this dep-wait bug [16:13] I think I have code that will do the job now, but need to figure out tests [16:13] cjwatson: Oh nice, you were working on it? [16:13] Yeah [16:13] I saw some chatter about dose-debcheck so thought you may be going all out straight off [16:14] I tried using dose but it ended up being more trouble than worth [16:14] Since we explicitly only want to care about the top level of build-deps, it's just as easy to check them directly using python-debian [16:15] (I'd love to have something like Debian's dose-builddebcheck setup, but it's hard to do efficiently with lots of archives) [16:19] That's an interesting problem [16:20] I suppose you should be able to make the overlay archive case more tractable by reusing earlier analysis [16:36] In principle [16:36] In practice I fear it's a fair bit of work in dose ... [16:36] Anyway, not needed for this bug [20:31] cjwatson: hmm, did you happen to find the bug in bzrlib? Potentially something I introduced with the changes for preserving dirty headers [20:34] cjwatson: ah I see you did, thank you. [20:39] wgrant: one last thing with the goimport that I believe I missed, in the series case the content should be "{base}/product/series bzr branch_path"? [20:49] thanks vila [20:52] blr: ;) o/ [20:52] blr: not landed on lp:bzr yet as pqm is broken :-} [20:53] err, broken as in someone is fixing it, not broken and nobody can land in lp:bzr anymore, it's an LTS package after all :o) [20:54] hah right [22:39] blr: Yeah, all sorted modulo the PQM problem, landed on qastaging [22:55] blr: Right, the first segment needs to match the request, and the third segment needs to be the full branch path. [22:59] wgrant: yep. that should be looking better now. [23:01] wgrant: cjwatson approved earlier, but before that fix, want to have another look? [23:08] blr: lgtm, thanks. [23:08] woot [23:08] oh yeah, i was trying to test the git support last night [23:08] get "qastaging.launchpad.net/pyrepl": found meta tag main.metaImport{Prefix:"qastaging.launchpad.net/pyrepl", VCS:"git", RepoRoot:"https://git.qastaging.paddev.net/pyrepl"} at https://qastaging.launchpad.net/pyrepl?go-get=1 [23:09] looks good? [23:09] the subsequent bits aren't going to be happy because that isn't a go project [23:09] mwhudson: yes, the git case was fine, I just mangled the bzr case :) [23:09] blr: seems good to check :) [23:09] just landing the fix branch now [23:09] cool [23:10] so qastaging will get it in an hour or so? [23:10] mwhudson: yep! [23:10] Yep, about an hour. [23:10] cool [23:10] then what's the eta until production? [23:10] want to get the special case removal change upstream asap [23:11] mwhudson: We should deploy today for the MP email fix anyway, so two or three hours? [23:11] wgrant: AWESOME [23:11] wgrant: is that reliant on the pqm fix? [23:12] blr: No, we have a patched version of bzr anyway, which Colin's upgraded devel to in the latest revision. [23:12] ah good, okay. [23:12] bzr's PQM is broken, not ours. [23:12] right [23:12] vila: do you think there will be another bzr release? [23:14] nice to see the less verbose MP comment email coming through, modulo the wonkiness. [23:14] Indeed, it is a big improvement.