[00:12] <jelmer> mwhudson, I can reproduce the determining revisions to fetch slowness
[00:23] <thumper> jelmer: awesome
[00:23] <thumper> jelmer: what was it
[00:24] <jelmer> thumper: building against svn 1.4 slows down bzr-svn significantly
[00:24] <jelmer> because we attempt to fetch all revision properties manually for each revision in the repository and don't cache the results
[00:25] <jelmer> something that can apparently take 7 hours for some repositories where it takes less than a minute when built against svn 1.5...
[00:26] <thumper> Ursinha: ping
[00:26] <thumper> jelmer: hmm.. interesting
[00:26] <thumper> jelmer: on a side note, has bzr-hg been fixed for the infinite recursion bug?
[00:32] <jelmer> thumper: I haven't had a chance to look into it yet.
[00:32] <poolie> hi jelmer
[00:33] <jelmer> 'morning poolie
[00:42] <wgrant> sinzui: Is there a good reason that admins cannot promote others to admins?
[00:46] <thumper> wgrant: admins of what?
[00:46] <wgrant> thumper: Teams.
[00:46] <thumper> wgrant: they can can't they?
[00:46] <thumper> wgrant: I thought they could
[00:46] <wgrant> No. Only owners can.
[00:46] <thumper> really?
[00:46] <wgrant> Admins can demote, but not promote.
[00:47] <thumper> hmm...
[00:47] <thumper> seems a bit screwy to me
[00:47] <thumper> if you trust them to be admins, let them be admins, not half-admins
[00:47] <wgrant> Right.
[00:50] <sinzui> wgrant: I am certain there is not *good* reason admins cannot promote
[00:51] <wgrant> It has been like that for as long as I remember, and I have always wondered about it.
[00:53] <thumper> wgrant: patches accepted :)
[01:06] <maxb> Submitting a patch loosening security policy can feel weird
[01:08] <wgrant> Yes.
[01:08] <wgrant> Particularly a security policy that has been around for perhaps 5 years and nobody knows the purpose of.
[01:09] <maxb> I don't suppose there's any light at the end of the tunnel w.r.t. removing xx-resetpassword-of-sso-account.txt ?
[01:09] <maxb> I miss being able to do a clean test run
[01:10] <wgrant> The OpenID RP stuff should cause that to vanish.
[01:10] <wgrant> But I'm not sure how far off that is...
[01:23] <Ursinha> thumper: pong
[01:24] <thumper> Ursinha: I was wondering if you had used (QAed for me) the API to get branches by date
[01:24] <Ursinha> thumper: not yet
[01:24] <thumper> ok
[01:24] <Ursinha> thumper: but will soon
[01:24] <thumper> :)
[01:34] <maxb> wgrant: I don't suppose you've already debugged ""GpgmeError: (32, 176, 'Unknown error code')"" on lucid, have you?
[01:36] <wgrant> maxb: Try http://paste.ubuntu.com/379455/
[01:36] <wgrant> (in sourcecode/pygpgme)
[01:36] <wgrant> Other than that, most stuff seems to work with stock Lucid + ppa:launchpad.
[01:37] <wgrant> I haven't run the whole test suite lately, though.
[01:37] <wgrant> The need to downgrade python-setuptools has mysteriously vanished in the past week or so.
[01:37] <maxb> nifty
[01:38] <maxb> I'm actually not working on 2.6 or lucid right now, having got distracted by "No really, don't load my system bzrlib.plugins.*"
[01:38] <wgrant> Is there a codeimport queue visible somewhere?
[01:38] <wgrant> I'm wondering if I would be better to branch from Git locally rather than waiting for my new import to complete.
[02:05] <mwhudson> lolz
[02:05] <mwhudson> mwh@grond:unicod-branch-names-bug-449528$ ls lib/canonical/launchpad/images/*gray*
[02:05] <mwhudson> lib/canonical/launchpad/images/persongray.png  lib/canonical/launchpad/images/teamgray.png
[02:05] <mwhudson> mwh@grond:unicod-branch-names-bug-449528$ ls lib/canonical/launchpad/images/*grey*
[02:05] <mwhudson> lib/canonical/launchpad/images/edit-grey.png  lib/canonical/launchpad/images/link-grey-arrow.png
[02:06] <mwhudson> wgrant: no, it's not visible
[02:06] <mwhudson> (there's already a bug report about that)
[02:06] <thumper> mwhudson: looking for a grey tick?
[02:06] <mwhudson> should be getting more machines soon though...
[02:06] <mwhudson> thumper: well
[02:06] <mwhudson> thumper: playing around with the icon for "partial success", yeah
[02:07] <mwhudson> tried a few things, i think a grey tick works best of the things i've tried
[02:08]  * thumper nods
[02:08]  * thumper is working through today
[02:08] <mwhudson> i've made one in gimp
[02:08] <mwhudson> i was actually trying to decide what to save it as :-)
[02:09] <thumper> I thought we use american spelling in the code, so gray
[02:09] <mwhudson> wfm
[02:10] <wgrant> Partial success? Does that mean incremental imports are sufficiently close that UI is a concern?
[02:10] <mwhudson> wgrant: yes
[02:14] <mwhudson> the gimp still has a pretty weird ui
[02:15] <wgrant> mwhudson: The single window mode is coming in the next release.
[02:16] <thumper> just in time to be removed by default?
[02:16] <mwhudson> wgrant: hooray
[02:17] <mwhudson> thumper, wgrant: http://people.canonical.com/~mwh/partial-success.png
[02:17] <thumper> mwhudson: looks good
[02:17] <wgrant> thumper: It's still probably a year away.
[02:18] <thumper> mwhudson: did you just desaturate the image?
[02:18] <mwhudson> thumper: yeah
[02:18] <wgrant> mwhudson: Nice.
[02:18] <wgrant> It has a tooltip, I hope?
[02:18] <mwhudson> wgrant: yeah
[02:18] <mwhudson> "Partial Success"
[02:21] <thumper> mwhudson: where is our list of contributor agreement signers?
[02:21] <wgrant> We're mostly not in the team :(
[02:22] <mwhudson> thumper: private url given to you in private channel
[02:22] <thumper> :)
[02:23] <lifeless> also the public team - you should get added wgrant :)
[02:28] <wgrant> Does anybody actually use Empathy?
[02:31] <jelmer> wgrant: Yep
[02:31] <jelmer> mwhudson, thumper: Does the "determining revisions to fetch" step always take very long or just on the initial import?
[02:31] <mwhudson> jelmer: just on the initial import
[02:31] <jelmer> mwhudson: I don't think there's a lot we can do to work around that other than upgrade the svn on lp to 1.5
[02:32] <mwhudson> jelmer: ok, want to fire off an rt asking how much work that would be?
[02:32] <mwhudson> or i can
[02:32] <wgrant> jelmer: With MSNP?
[02:32] <jelmer> wgrant: yep
[02:33] <jelmer> mwhudson: how urgent is it? Lucid will have a recent enough version of svn
[02:33] <wgrant> jelmer: Odd. I keep trying to use it, but give up every time because it silently drops incoming messages once the conversation times out after a minute or so.
[02:33] <wgrant> It doesn't drop messages for you?
[02:34] <jelmer> wgrant: not that I have seen
[02:36] <mwhudson> jelmer: maybe not all that urgent, it's only a performance thing after all
[02:50] <wgrant> Subject: None
[02:50] <wgrant> Thankyou Launchpda.
[03:23] <maxb> FYI, I already have svn 1.6 *and* pysvn for hardy in a PPA
[03:26] <_thumper_> mwhudson: https://code.edge.launchpad.net/~ian-clatworthy/bzr/whats-new/
[03:26] <thumper> mwhudson: a bad error... know what caused it?
[03:26] <thumper> mwhudson: my guess is the scanner, I just don't know why
[03:26] <thumper> mwhudson: also, we should get bzr 2.1 final into LP
[03:26] <mwhudson> thumper: no, that's the puller
[03:27] <mwhudson> thumper: 2.1 is probably trivial, drop the tarball in, update versions.cfg and ec2 land
[03:27] <thumper> mwhudson: does the puller set the stacked branch in the DB?
[03:27] <mwhudson> thumper: yes
[03:27] <thumper> ah
[03:27] <mwhudson> that's a pretty nasty error indeed
[03:27] <mwhudson> maybe duelling pullers?
[03:28] <mwhudson> a remirror will probably clear it up
[03:32] <thumper> hmm..
[04:13] <thumper> mwhudson: it is past beer o'clock
[04:13] <thumper> mwhudson: are you tweaking the code import dispatcher?
[04:13] <mwhudson> thumper: yeah
[04:13] <mwhudson> thumper: i'm thinking about it
[04:14] <thumper> mwhudson: thinking about doing it Monday?
[04:14] <mwhudson> or maybe i'll stop instead and go and sit outside until emma calls and tells me she's finished work
[04:14] <mwhudson> thumper: yeah
[04:14] <thumper> you know if you try again, you'll miss something
[04:14] <thumper> do it Monday
[04:14] <thumper> have a beer
[04:14] <mwhudson> ok :-)
[04:14] <thumper> I hope it is sunnier in Wellington than here
[04:14] <thumper> we used to have sun
[04:14] <thumper> but it has gone again
[04:15] <mwhudson> it's a very very nice day today actually
[04:15] <thumper> do you think the description should go above or below?
[04:15] <mwhudson> sunny, not too windy
[04:15] <thumper> I'm thinking above
[04:15] <mwhudson> bit cool perhaps
[04:15] <mwhudson> thumper: did you make that screeny?
[04:15] <thumper> yeah
[04:15] <thumper> 6
[04:16] <thumper> http://people.canonical.com/~tim/description6.png
[04:16] <mwhudson> oh right
[04:17] <mwhudson> i'm not really sure
[04:17] <mwhudson> i guess it makes sense for it to go high up the page
[04:25]  * mwhudson EOWs
[04:25] <poolie> cheerio
[04:26] <poolie> thumper: really a description and a commit msg and a comment?
[04:26] <thumper> poolie: the editable description is instead of the initial comment
[04:27] <poolie> oh ok
[04:27] <poolie> more like a bug then?
[04:27] <thumper> poolie: trying to be
[04:33]  * thumper EOWs too
[04:52] <Ursinha> thumper: are you still there?
[04:53] <Ursinha> thumper: if so, do you have an example of usage of the branches search by date?
[05:46] <kfogel> thumper: looking rocketfuel-setup, I see that it does 'bzr branch lp:~launchpad-pqm/launchpad/devel $LP_TRUNK_NAME' after attempting 'bzr launchpad-login $lpusername'.  This means that if the user has a Launchpad login (and successfully logs in) that the 'bzr branch' will happen over bzr+ssh://, right?
[05:46] <kfogel> thumper: oh, I see your answer in https://answers.edge.launchpad.net/launchpad/+question/101139
[05:46] <kfogel> thumper: which I think says that
[05:52] <poolie> hello kfogel
[05:53] <kfogel> poolie: hello
[06:22] <wgrant> BjornT: Can you please review https://code.edge.launchpad.net/~wgrant/launchpad/sprbu-columns-to-sprb/+merge/18995 reasonably soon?
[08:18] <henninge> Morning!
[08:18] <henninge> How do I find out the current database user in a unit test?
[08:27] <adeuring> good morning
[08:57] <BjornT> wgrant: sure. sorry, i had missed that one, thanks for reminding me.
[08:59] <wgrant> BjornT: Thanks.
[09:12] <mrevell> Hi
[11:01] <deryck> Good morning, all.
[12:02] <wgrant> bigjools: The circular import may not be a good reason to not do it, but is there a good reason *to* do it?
[12:10] <bigjools> because Attribute sucks and it takes 60 seconds to do the right thing
[12:11] <bigjools> and on that note, I am heading out for a liquid lunch while my power is off
[14:09] <wgrant> Why is db-devel so ancient?
[14:44] <deryck> BjornT, ping
[14:47] <BjornT> deryck: pong
[15:15] <kfogel> deryck, adeuring: just fyi (re my mail about failing test): the EC2 run that I submitted right after that has now passed, so I'm going to pqm-submit my changes for bug #255868.  sinzui already reviewed.
[15:15] <mup> Bug #255868: Project summary page should show links to patches <story-patch-report> <ubuntu-upstream-relations> <Launchpad Bugs:In Progress by kfogel> <https://launchpad.net/bugs/255868>
[15:16] <deryck> kfogel, sounds good.  Was going to look later, but glad it's passing now.
[15:16] <deryck> kfogel, it looked to me like your ec2 run was against devel instead of db-devel, based on the log messages you sent.
[15:19] <adeuring> kfogel: I haven't received your mail (yet)...
[15:19] <kfogel> adeuring: was to bugs team members (all)
[15:20] <kfogel> deryck: based on the two outputs (failure and then success), you're right, and yet the command line was the same and specified db-devel.  hmmm.
[15:20] <adeuring> kfogel: our mail server seems to be a bit slow sometimes...
[15:21] <kfogel> deryck: I'll keep the records; would like to figure out what happened here -- a lost EC2 run costs a lot.
[15:22] <deryck> kfogel, you use `utilities/ec2 land` to make sure the syntax is good, too.
[15:22] <deryck> kfogel, I think most of us use that now, rather than ec2 test.
[15:22] <kfogel> deryck: yeah, should just do that
[15:23] <kfogel> ok, submitted.  out for a while
[15:24] <BjornT> bigjools: ping?
[15:25] <bigjools> BjornT: yo
[15:26] <BjornT> bigjools: we currently have both DistributionSourcePackage and DistributionSourcePackageCache in the db. it feels like we should combine those two, but i'm not sure what the latter is used for. can you enlighten me?
[15:28] <bigjools> BjornT: it's for searching; the distro search page uses it
[15:29] <bigjools> gets updated every day
[15:29] <bigjools> there's an fti on the various columns
[15:30] <BjornT> bigjools: would  make sense add settable columns to it? i.e., have it represent a distribution source package in the db, which can have bug reporting guide lines for example.
[15:31] <bigjools> BjornT: I don't follow what you mean
[15:35] <BjornT> bigjools: basically, i think that the DistributionSourcePackage python class should be connected to a table in the db. this table should both have some properties for the package, for example bug reporting guidelines, which is settable by users, as well as acting as a cache for certain properties, like the latest publication record.
[15:36] <BjornT> bigjools: would it make sense to use DistributionSourcePackageCache for this? is would it be better to keep it for searching only?
[15:37] <bigjools> BjornT: there is already a DistributionSourcePackageInDatabase !
[15:37] <bigjools> I would keep it for searching only
[15:37] <bigjools> DistributionSourcePackage and DistributionSourcePackageInDatabase should be merged; that was always the intention
[15:38] <bigjools> and you'll notice the latter has the bug_reporting_guidelines on it
[15:38] <BjornT> bigjools: yes, i know. my question was whether it would make sense to merge DistributionSourcePackageInDatabase and DistributionSourcePackageCache
[15:39] <bigjools> BjornT: I don't think it does
[15:39] <bigjools> partly because the latter is also archive-specific
[15:39] <bigjools> i.e. it gets generated with knowledge about publications
[15:44] <BjornT> bigjools: ok, makes sense. DistributionSourcePackageCache should probably get renamed as to not cause that much confusion.
[15:44] <bigjools> yeah, that's cool
[15:44] <bigjools> naming is the hardest part of programming, after all :)
[15:57] <BjornT> bigjools: do you have time to have a quick call to talk me through https://code.edge.launchpad.net/~wgrant/launchpad/sprbu-columns-to-sprb/+merge/18995? i'm interested in the model's PoV
[15:57] <bigjools> BjornT: yeah can do
[15:58] <bigjools> call me on skype when you're ready
[15:59] <BjornT> cool
[16:02] <jtv> BjornT, can I bounce something off you?
[16:02] <jtv> BjornT: we have a bunch of fake transaction managers in tests.
[16:03] <jtv> Would it be a good idea to make that a test helper?
[16:10] <BjornT> jtv: could be, if they are similar enough.
[16:10] <jtv> BjornT: looks like.  In fact this is one reason why I still like to pass transaction managers around: it's a reasonably clear injection point for tests.
[17:08] <adiroiban> gmb: hi. Any news regarding the landing of this MP https://code.edge.launchpad.net/~adiroiban/launchpad/bug-522188/+merge/19395 ?
[17:08] <gmb> adiroiban: !??! It appears to have disappeared again. This is vexing. I'll go and see if I can find it.
[17:09] <adiroiban> :))
[17:11] <gmb> adiroiban: It looks like the ec2 run just sort of died and didn't email anyone to say that it had. Starting a new run now; I'll keep an eye on it.
[17:12] <adiroiban> gmb: np. I will remind you on monday if it will fail again
[17:13] <gmb> adiroiban: If that happens I'm going to take a very large axe to our ec2 utility.
[17:20] <adiroiban> :)
[17:24] <jtv> abentley: preparing to land your branch-url branch
[17:25] <abentley> jtv, rock.
[17:26] <jtv> abentley: that's a whopping 4-branch landing.  :)  Can I just land the last branch (after pumping) and the whole pipeline will be included?
[17:29] <abentley> jtv, yes.  As long as all your stuff is fully merged into mine.  You can check that with bzr missing.
[17:31] <jtv> abentley: it says "you have 15 extra revision(s):"
[17:31] <jtv> followed by... yes, 15 revisions.
[17:34] <jtv> abentley: does that mean I'm good to go?
[17:35] <abentley> jtv, well, it depends whether the extra revisions are in my branch or yours.
[17:35] <jtv> abentley: I merged yours into mine.
[17:37] <abentley> jtv, okay, never mind then.
[17:37] <jtv> abentley: and all those 15 revisions are mine.
[17:37] <abentley> jtv, the general answer to your question is yes.
[17:37] <jtv> \o/
[17:38] <abentley> If you want to land multiple branches, and all the branches have been merged directly or indirectly into one of them, you can just land that one.
[17:39] <abentley> pumping is just a convenient way of ensuring all the branches are merged into the last pipe.
[17:42] <jtv> convenient indeed... it's been fun!
[18:37] <sinzui> EdwinGrubbs: I like your proposal. I'll reply with some questions about how communities complete for the involvement portlet.  Let's assume we are going to extend the Involvement portlet instead of creating a new portlet
[20:05] <deryck> Have a good weekend, everyone.
[20:42] <maxb> I have a MP that is "review approve"d but not "merge approve"d. Would someone possibly be around who could flip that setting and ec2 land it?   https://code.edge.launchpad.net/~maxb/launchpad/bug-497731/+merge/19732
[21:20] <rockstar> gary_poster, hi
[21:20] <gary_poster> hi rockstar
[21:21] <gary_poster> (I'm about to turn into a pumpkin btw, but can reply using cell phone in a bit)
[21:22] <rockstar> gary_poster, so, if I need to create a new script in bin/ then I need to create a .in file in buildout-templates/bin, but where do I tell buildout about it?
[21:25] <gary_poster> rockstar, if it is one of those kinds of templates--free form--then just putting it in the directory is sufficient.  rerunning bin/buildout will notice and build.  Note that there are other kinds of scripts that let you keep your code in the tree (see lp-windmill, for instance, and how it is hooked up in buildout.cfg [scripts] entry-points, but I imagine you've seen those and rejected them with reason.
[21:41] <wgrant> Can somebody please ec2 land https://code.edge.launchpad.net/~wgrant/launchpad/sprbu-columns-to-sprb/+merge/18995 for me?
[21:43] <lifeless> lp-project-upload looks nearly-useful
[21:43] <lifeless> but I really don't want to depend on ubuntu-dev-tools for an upstrea project
[21:43] <lifeless> any suggestions on where I should ask pitti to move the code to ?
[21:49] <wgrant> lifeless: Maybe launchpad-tools, the collection of useful Launchpad-related scripts?
[21:49] <lifeless> wgrant: I'm thinking lptools
[21:49] <lifeless> https://bugs.edge.launchpad.net/lptools
[21:49] <lifeless> is that what you were meaning?
[21:50] <lifeless> dobey: btw
[21:50] <wgrant> lifeless: I meant a hypothetical thing, the lack of existence of which was a bug.
[21:50] <lifeless> dobey: _puhlease_ beat statik into giving you some work time to do QA [we do allow for this when we open source something, so that we're not doing abandon-ware]
[21:50] <lifeless> wgrant: hah. Well it exists but is named differently.
[22:06] <wgrant> Why has db-devel not been merged into for a couple of days?
[22:28] <dobey> lifeless: was off today, so was out doing other things in the nice warm sun :)
[22:28] <lifeless> fair 'nuff
[22:29] <lifeless> dobey: I'd love to see you do an upstream release :)
[22:30] <dobey> lifeless: During this cycle, I don't even really have enough work time to finish the stuff for Lucid. But I think after Lucid, I want to switch to our ops+ team for 6-12 months, and do a lot of work on infrastructure/tools, and lptools would certainly fall into that realm
[22:30] <dobey> lifeless: yeah, I know. I would if I had any time to fix the bugs and do it :)
[22:31] <lifeless> dobey: just doing a release would be helpful
[22:31] <lifeless> it would mean that the packaging would be a little more sane.
[22:31] <dobey> lifeless: I'll see if I can't do that. I need to do a release for changeup also
[22:33] <lifeless> changeup ?
[22:34] <dobey> A fairly simple thing to make it possible to automatically restart ubuntuone-syncdaemon, gwibber-daemon, and similar 'behind the scenes' user-level services, on upgrades
[22:34] <lifeless> dobey: also if you can merge the non-packaging parts of bzr+ssh://bazaar.launchpad.net/~lifeless/debian/sid/lptools/sid/ that would be really nice.
[22:34] <dobey> what's non-packaging in there?
[22:35] <lifeless> dobey: e.g.: bzr merge lp:~lifeless/debian/sid/lptools/sid/; bzr revert debian; bzr revert --forget-merges; bzr commit -m "whatever"
[22:35] <lifeless> dobey: have a look :>
[22:36] <dobey> hrmm
[22:36] <lifeless> a bunch of work from ted
[22:36] <lifeless> desktop files
[22:36] <lifeless> a preferences dialog
[22:37] <dobey> sounds frightening
[22:40] <lifeless> https://code.edge.launchpad.net/lptools/+activereviews
[22:40] <lifeless> would be good to merge
[22:40] <dobey> I wish I could get the same list of reviews from the API, as is shown on http://code.launchpad.net/~dobey/+activereviews
[22:40] <lifeless> it has most of the changes I'm talking about
[22:40] <lifeless> dobey: you can now I believe
[22:41] <lifeless> or not quite; you can get reviews-requested-for-an-object
[22:41] <dobey> yeah, I just haven't had time to look at his branch because it's so big
[22:41] <lifeless> which is not the same
[22:41] <lifeless> dobey: JFDI, honestly, everyone is running his branch.
[22:49] <james_w> dobey: you can now get those two reviews on your +activereviews page with lp.me.getRequestedReviews() or similar
[22:50] <lifeless> james_w: I thought that didn't handle group membership yet?
[22:50]  * lifeless really wants a direct match to +activereviews too
[22:51] <james_w> but those reviews are requested of dobey n'est pas?
[22:51] <lifeless> in this case
[22:52] <james_w> hence "those two reviews"
[22:52] <lifeless> it may be that +me/activereviews isn't showing what it should either ;>
[22:52] <james_w> it's a start
[22:53] <james_w> so, one API request to get them all would be good
[22:53] <james_w> and just returning a list of merge proposals is fine, as you can sort client side
[22:54] <james_w> so we need a parameter to getRequestedReviews() to do a recursive lookup, then a modification to the internal method it calls
[22:54] <dobey> There's a 3rd review on a private branch that isn't directly requested of me, but is via group membership
[22:54] <james_w> do you know if there is a way to get the transitive closure of team memberships for a person?
[22:55] <james_w> dobey: ah, I can't see that one, indeed you can't get that easily with this API yet
[22:55] <james_w> I'd like to fix that, but it's more than just exposing something over the API like the current version
[22:55] <dobey> I don't really care about sorting. That's a trivial problem. :)
[22:55] <dobey> I just want all the reviews pertaining to me