[01:12] <StevenK> wgrant: https://code.launchpad.net/~stevenk/launchpad/less-public-bmp/+merge/150731
[01:17] <wgrant> StevenK: Sounds good
[01:17] <wgrant> Thanks for fixing those security adapters
[01:17] <wgrant> By deleting them
[01:20] <StevenK> Heh
[03:05] <StevenK> wgrant: Trying to render create_initialized_view(branch, '+index', rootsite='code')
[03:05] <StevenK> results in a large and angry traceback
[03:05] <StevenK> Due to breadcrumbs, from what I can see
[03:26] <StevenK> wgrant: http://pastebin.ubuntu.com/5572431/
[03:43] <wgrant> StevenK: Some views won't render like that
[03:43] <wgrant> I don't quite recall why
[03:44] <wgrant> webops: As lp_publish@pepo, could you run `LPCONFIG=ftpmaster-publish scripts/ftpmaster-tools/obsolete-distroseries.py -d ubuntu -s jaunty`?
[03:47] <StevenK> Hah
[03:47] <StevenK> wgrant: The pastebin is my success
[03:47] <wgrant> Oh, wrong chan
[03:47] <wgrant> StevenK: Right
[03:48] <wgrant> That's the usual workaround
[04:09] <StevenK> wgrant: https://code.launchpad.net/~stevenk/launchpad/moar-preload-people-branch-index/+merge/150953
[04:10] <wgrant> StevenK: Why not just make hasSubscription not completely insane?
[04:11] <wgrant> Hm
[04:11] <wgrant> Also, is that the right method?
[04:12] <wgrant> Ah, it is
[04:12] <wgrant> But I think your change is still insane.
[04:12] <StevenK> wgrant: http://pastebin.ubuntu.com/5572472/
[04:13] <wgrant> StevenK: see DecoratedBranch.hasSubscription
[04:14] <StevenK> Argh
[04:14] <StevenK> Argh
[04:15] <StevenK> Hm
[04:15] <StevenK> I can probably delete that, then
[04:15] <wgrant> Delete what?
[04:15] <StevenK> DecoratedBranch.hasSubscription
[04:16] <wgrant> Not really.
[04:16] <wgrant> For one, your change to Branch.hasSubscription is wrong
[04:16] <wgrant> For another, it's not clear that it's valuable.
[04:16] <wgrant> I'd change your branch to just make DecoratedBranch.hasSubscription not crap
[04:17] <StevenK> wgrant: http://pastebin.ubuntu.com/5572479/
[04:17] <wgrant> StevenK: How does that change anything?
[04:18] <StevenK> self.subscriptions is a cached list, apparently
[04:18] <wgrant> Sure
[04:18] <wgrant> But it's exactly the same as what the method was doing before, except slower because there's no short-circuit
[04:18] <wgrant> So it'll just execute *more* queries.
[04:19] <StevenK> So I should do the personID check, then?
[04:19] <wgrant> Right
[04:19] <wgrant> That's how you would avoid making person queries.
[04:20] <StevenK> wgrant: http://pastebin.ubuntu.com/5572483/
[04:20] <wgrant> That looks more sensible.
[04:21] <StevenK> Tempted to remove the last three lines of the docstring for being wrong and causing more queries
[04:21] <StevenK> Well, the lines were wrong, they are not any more ...
[04:27] <StevenK> wgrant: MP updated.
[04:27] <wgrant> StevenK: r=me
[04:29] <StevenK> wgrant: And BMPs are very very unhappy on qas
[04:29] <StevenK> But I suspect that was known
[04:29] <wgrant> What do you mean?
[04:29] <StevenK> I tried to look at a few MPs on qas, and they OOPS due to librarian content
[04:30] <wgrant> Yes
[04:30] <wgrant> That's always been the case
[04:30] <wgrant> You have to create new ones
[05:06] <StevenK> wgrant: What do you think the plan is for bug 736014 ?
[05:06] <_mup_> Bug #736014: BugTask:+huge-vocabulary timeout <target-picker> <timeout> <vocabulary> <Launchpad itself:Triaged> < https://launchpad.net/bugs/736014 >
[05:06] <StevenK> We've discussed it a few times, but then Branch and BMP ate my brain.
[05:09] <wgrant> StevenK: DistributionSourcePackageCache/DistroSeriesPackageCache
[05:09] <wgrant> (I believe those two parallel class names were designed to have the same acronym just to make me angry, since there's no other reason for one to have Source and one not)
[05:09] <StevenK> Those two classes make me sad
[05:21] <wgrant> z,cb 3
[05:21] <wgrant> bah
[05:22] <StevenK> Cat?
[05:22] <wgrant> Accidental Dvorak
[05:24] <lifeless> so thats what, 'f*ck u' ?
[05:25] <wgrant>  /win 3, but close enough
[08:16] <noodles775> Hi! Is anyone up for a small-ish review (190l): https://code.launchpad.net/~michael.nelson/launchpad/include-binary-size/+merge/150831
[08:34] <czajkowski> wgrant: StevenK what is the story with on call reviewer nowadays ?
[08:45] <lifeless> on what :P
[08:46] <lifeless> noodles775: you know buildbot is 30m or something now :)
[08:46] <czajkowski> lifeless: mornin'
[08:50] <noodles775> lifeless: no? wow. fwiw, the tests didn't seem to run correctly on my instance: http://paste.ubuntu.com/5572841/ (that's using `time ./bin/test`)
[08:55] <lifeless> noodles775: http://blog.launchpad.net/performance/parallel-testing-is-live
[08:56] <lifeless> noodles775: what size instance (machine RAM) ?
[08:56] <cjwatson> About 40m I think
[08:58] <noodles775> lifeless: it's an m1.small (2G), nothing else running on it (I just set it up for LP development yesterday)
[08:59] <cjwatson> I know wgrant said that adding hashes would be "probably overcomplicated", but cf. bug 1088527
[08:59] <noodles775> Wow - 40mins from 6hrs.
[08:59] <noodles775> Looking
[08:59] <_mup_> Bug #1088527: Needlessly asks for exact package name, file size, uploaded version <ca-escalated> <Developer registration portal:New for michael.nelson> <Launchpad itself:In Progress by michael.nelson> < https://launchpad.net/bugs/1088527 >
[09:00] <cjwatson> Err, sorry
[09:00] <cjwatson> cf. bug 1007195
[09:01] <cjwatson> is what I meant to say
[09:01] <_mup_> Bug #1007195: Librarian-backed HostedFile objects do not expose SHA-1 hash <soyuz-build> <Launchpad itself:Triaged> < https://launchpad.net/bugs/1007195 >
[09:01] <cjwatson> (Doesn't specifically have to be the SHA-1 of course; I think that was just what the librarian keyed off at the time or something
[09:01] <cjwatson> )
[09:02] <cjwatson> Anyway, don't let me get in the way by expanding the scope of your work, it was just in case it was trivial to allow for it
[09:03] <noodles775> cjwatson: sure - it should be trivial. So s/include_sizes/include_meta/ and add the the sha1 to the dict?
[09:07] <cjwatson> I guess something like that
[09:31] <lifeless> noodles775: sounds like there is a leak in there somewere, I'd use a bigger instance
[09:31] <lifeless> noodles775: lp's cloud test regime uses x1.xlarge IIRC
[09:32] <noodles775> lifeless: will do next time - I just fired this one up for this bug and normaly only use small for dev work. Thanks.
[10:17] <noodles775> lifeless, cjwatson: updated with your suggestions (except for HasLength). Thanks!
[11:07] <lifeless> noodles775: why not HasLength?
[11:09] <noodles775> lifeless: I commented on the MP, but http://paste.ubuntu.com/5572967/ (and I haven't updated lp's deps) (or am I missing something obvious?)
[11:13] <lifeless> 0.9.14 is quite old
[11:13] <lifeless> 0.9.29.is current, and 30 is about to come out
[11:13] <noodles775> Nice
[11:13] <lifeless> so should be as simple as dropping a dep in launchpad-dependencies and updating versions.txt
[11:13] <lifeless> gnight!
[11:13] <noodles775> G'night :)
[11:14] <lifeless> I can't see any incompat warnings in those releases
[11:15] <lifeless> there's a bunch of shiny improvements though, new matchers and more
[11:15] <lifeless> right, -> sleep.
[14:17] <noodles775> Would someone be able to send this branch to land for me? I've done updates after comments from RobC and ColinW: https://code.launchpad.net/~michael.nelson/launchpad/include-binary-size/+merge/150831
[14:18] <czajkowski> gmb: any idea how we take care of stuff like this now awadays
[14:22] <noodles775> Np, I can wait for one of the LP engineers - the branch isn't that urgent :)