[00:02] wgrant: you noticed# [00:02] ? [00:02] The middle one hasn't had any attention, AFAICT. [00:03] (try going to a distribution source package, distro series, distro arch series, or product series. Click on tabs. Be confused.) [00:07] :( [00:11] :(( [00:11] build failures [00:31] thumper, call? [00:32] rockstar: yes sir [00:36] * jml is back [00:45] wow, the template conversion summary is pretty exciting [00:45] lots of "almost dones"... who will be next [00:46] mwhudson, hello [00:46] jml: hello [00:47] mwhudson, do you want to talk at all about the entrypoints stuff? [00:47] jml: did i send my reply yet? :) [00:47] seems not [00:47] jml: i have a half written reply [00:47] mwhudson, oh cool. [00:48] mwhudson, well, if you'd like to talk, let me know. [00:48] jml: will do at some point today i'm sure [00:48] mwhudson, ok [01:01] jml: are we in testfix? [01:02] barry, I don't know. [01:03] https://lpbuildbot.canonical.com/builders/lp/builds/135/steps/shell_7/logs/summary seems to indicate some sort of failure [01:04] jml: blarg. let me look [01:06] find -name xx-*.txt | xargs rm -f [01:07] mwhudson: rs=me [01:07] barry: don't tempt me [01:07] sinzui, barry: should breadcrumbs have icons like the bug one? [01:08] thumper: i think they can. "should they" is a beuno question :) [01:08] hmm... [01:08] mwhudson: where's the candy-like history erase button when you need it? [01:09] got some family stuff to do now. if no one else attacks those failures in the meantime, i'll take a look in a few hours. ping me if you do though so i don't step on your toes [01:11] barry: I want to be able to explicity say "don't add this view name to the breadcrumbs", can I? [01:14] jml: this could give us better defaults [01:14] jml: if we default to not adding the page name [01:14] jml: but allow people to add it in [01:19] * mwhudson lunches [01:29] maxb, btw, what happened with the first test run was that Launchpad was in testfix mode, thus rejecting new patches. [01:29] thumper, "this" being... explicitly saying "don't add this view name to breadcrumbs"? [01:30] jml: yes [01:30] cool. [01:38] thumper, barry, they should not [01:38] salgado will fix that [01:39] beuno: by "that" I take it you mean breadcrumbs should not have icons? [01:41] yes :) [01:49] thumper, that diff size was misleading. Lots of it was deletions, which I don't review. :) [01:52] rockstar: I said that :) [01:54] * thumper needs fun and caffeine [01:54] heh [01:54] typo: s/fun/fud/ [01:54] although fun would also be good [01:57] jml, mwhudson: i have a testfix branch to submit [01:57] bac: great! [01:59] bac: do you want one of us to review it? [01:59] (or rs it, more likely) [02:00] mwhudson: no. just didn't want to interfere with work you may be doing. i fix one test and revert a revision that caused lots of failures. [02:00] i will rs it [02:00] barry: ^^ [02:00] bac: thanks a lot [02:03] * thumper turns BranchHierarchy.objects into a generator [02:04] hmm [02:04] debugging generators is harder [02:16] thumper, jml, hi, i'm just going to put my 2.1 timeline mail into launchpad milestones [02:16] (silly old me) [02:17] and just thinking about aligning them with launchpad milestones [02:17] i see you don't actually use milestones? [02:17] ): [02:17] i mean :) [02:18] oh maybe they're on launchpad-project but that oopses [02:18] well, i'll just add the ones i want and then we can talk about them [02:18] if we want [02:21] rockstar: ping [02:22] poolie: we do, kinda [02:22] btw https://edge.launchpad.net/launchpad-project/+milestones times out fairly persistently [02:24] poolie: I think that is a known issue, but not entirely sure [02:24] poolie: there was one that was doing *a lot* of bug tag queries [02:24] That was the project group milestone page. [02:24] Not the milestone listing. [02:25] thumper: No breadcrumbs should not have icons. [02:25] thumper, pong [02:31] hmm [02:32] mwhudson: fwiw, my next ec2 submission took 7 minutes... [02:32] bac: that's rather better [02:33] bac: how long did the instance take to boot? [02:33] mwhudson: dunno, wasn't paying attention [02:33] bac: it's in the output if you still have that [02:33] after all the ........................................................ that are near the beginning [02:34] 4:12 [02:34] is there any way to get the branch specified in the Vcs-Bzr: tag in a source package, via the LP API? [02:34] bac: cool [02:34] dobey: No. [02:34] boot time + three minutes isn't too bad, i think [02:34] dobey: But... [02:34] mwhudson: no, much better than last. [02:35] dobey: You should be able to assume soon that eg. lp:ubuntu/hardy/dpkg is the canonical hardy dpkg branch. [02:35] but that is not what i want :) [02:36] What do you want to do? [02:38] i want to get the sour package branch for a particular source_package_publishing_history in the LP API [02:39] You can't rely on Vcs-Bzr to be reliable for that. [02:39] well i can't rely on anything to be reliable for it at the moment [02:39] because there's no API to get it via LP [02:40] LP doesn't have that information. [02:40] Nothing does. [02:41] i just want to automate the tedium of doing releases [02:41] rockstar: sorry, missed pong [02:41] because it's taking me a lot more time than i would like to spend, doing them [02:41] rockstar: got a minute? [02:43] and it's going to get more so, once i have to start dealing with stable branches vs. trunk [02:44] i also wish 'person' had methods to query for bugs [02:45] as i don't see any way to do a query for 'all bugs assigned to me' [02:46] dobey: it can be done, but it isn't trivial [02:47] well, the web does it, so i'm sure it's doable. i just don't see how to do it, and don't really want to go searching through web ui code to figure out how :) [02:48] There's no API for it. [02:48] There's a bug about that. [02:48] It's just not trivial. [02:48] thumper, sure. Skype? [02:49] yep [02:49] thumper, just call [02:49] trying to type while pinentry keeps popping up is an arse [03:31] why are so few of the exposed methods and properties on Person actually visible on the API docs? [03:33] jml: The attributes all show up on Team. [03:33] :( [03:34] grr my brain isn't working very well today [03:34] wgrant, that's unfortunate. [03:34] jml: It is. That's what you get for doing mildly hackish things, I suppose. [03:34] implementing Team in a Person is also unfortunate [03:36] hmm [03:36] lifeless: does txAWS use boto under the hood? [03:36] nope [03:36] boto free [03:37] cleaner api [03:37] twisted! [03:37] and a teeny gui :> [03:37] lifeless: 2.4 compatible, i guess? [03:37] dunno :) [03:37] lifeless, bram says Twisted is confusing :\ [03:37] jml: its coding standards sure are! [03:38] lifeless, heh heh [03:38] lifeless, now _that_ I agree with. [03:38] lifeless: how does it talk to the web service part? i thought it was some SOAP horror show [03:38] * jml needs to hack up a Twisted version of pep8.py [03:38] i guess i can find this stuff out myself... [03:39] mwhudson, are you thinking what I think you're thinking? [03:39] jml: probably [03:39] mwhudson: twisted.http.client thingy + xml [03:39] wee, no setup.py [03:39] mwhudson, excellent. :) [03:39] that's going to make building an egg nice and easy [03:39] use packages [03:39] they are better than eggs [03:39] lifeless: "hnnnnnnnnnnnnnnnnnnnngh" [03:40] in other news, loggerhead is terrible [03:41] graar so is paste [03:41] yes [03:41] DEB [20090917-03:40:51.803] [47331369258224] paste.httpserver.ThreadPool: Added task (128 tasks queued) [03:41] stop spawning threads, stupid software [03:41] mwhudson, btw, do you still want a call today? If so, can we arrange a time now? [03:41] jml: i think it would be a good idea, yes [03:42] jml: any time in the next say three hours works for me [03:42] mwhudson, ok. let's talk now. :) [03:43] jml: not so "now" eh? [03:44] mwhudson, skype did something [03:47] I hate the default reactor [03:51] \o/ landing last of the code page 3.0 fixes [03:51] now for those breadcrumbs [04:29] thumper, wooot [04:30] * jml away [04:31] mwhudson: the bzr.dev builder is showing weirdness loading the git plugin [04:33] thumper: i guess bzr.dev is already at api version 2.1 ? [04:33] thumper: a run against the 2.0 branch would be more interesting right now i suppose [04:38] hmm... [04:41] this brings up a question [04:41] for the next 6 months of bzr development, what version is launchpad going to run? [04:41] 2.0 [04:41] or 2.1b? [04:41] 2.1b please [04:59] thumper: yeah, i think we more-or-less have to follow the "beta" line [05:11] rockstar: breadcrumbs are up [05:11] hmm.. [05:12] thumper, I see that. [05:12] ta [05:16] mwhudson: got any idea how to run "make check" properly for cscvs and the new bin/py stuff? [05:17] thumper: does make check PYTHON=~/canonical/checkouts/trunk/bin/py work? [05:17] mwhudson: it won't here because I don't have that path [05:17] I'd hate to see that in the Makefile [05:18] thumper: i was assuming you'd change to point to a real bin/py [05:18] mwhudson: on the commandline? [05:18] :) [05:18] sorry [05:18] misunderstanding [05:18] (i'm not sure it would work either) [05:20] got it using the right python [05:20] which hopefully will bring in the right bzrlib [05:20] mwhudson: I'm looking at the 'unicode' object has no attribute 'branchName' [05:20] thumper: ah, cool [05:24] *hurl* [05:26] if branch.__class__ is not types.StringType: [05:26] branch = branch.branchName [05:26] * thumper tries to remember the base_string thing [05:27] thumper: kill it quick, before it breeds [05:27] isinstance(x, basestring) [05:28] mwhudson: this is in the cscvs branch [05:28] I don't think I have that set up :( [05:28] * thumper looks [05:28] thumper: well i guess i'm glad it's not in launchpad [05:28] mwhudson, rockstar: either of you got a few minutes for a trivail cscvs fix? [05:29] thumper: yeah [05:30] thumper: where is this obscenity? [05:30] CVS/StorageLayer.py:537 [05:30] I'm checking for others [05:31] just the one [05:32] mwhudson: that should account for the three failing CVS test_worker import failures [05:33] * thumper checks [05:33] thumper: https://pastebin.canonical.com/22200/ ? [05:34] mwhudson: that is exactly what I have, and it passes the three TestCVSImport failures with 2.0rc2, r=me [05:35] thumper: ok, i'm running the cscvs tests here, but if they pass i'll land it [05:35] cool [05:35] I don't think you'll have any problems [05:35] that line has to be more robust now :) [05:35] mwhudson: I was very tempted to replace all the 8 space indents though [05:36] mwhudson: I almost went blind! [05:36] thumper: :-) [05:36] welcome to hel^Wcscvs [05:38] mwhudson: an rs=me for any fixing of the whitespace in cscvs :) [05:38] thumper: you can't trick me that easily! [05:38] * thumper tweaks the stacking tests to use format="2a" [05:38] :) [05:48] thumper: the cscvs branch is in pqm's queue now [06:01] mwhudson: cool [06:01] mwhudson: I'm looking at the TestBranchPuller failures now [06:02] mwhudson: it seems pretty mechanical, removing format='1.6' from places [06:02] thumper: ah right, we use "1.6" to mean "some format that supports stacking" a bit i guess [06:03] yeah [06:09] mwhudson: is that going to take ages to get through? [06:10] thumper: don't know [06:10] mwhudson: I have a fix for the rest [06:10] v.trivial thankfully [06:10] thumper: happy to hear that [06:11] mwhudson: can you review? [06:11] * thumper is pushing now [06:11] thumper: sure [06:11] ta [06:18] mwhudson: proposed, email should be through shortly [06:18] mwhudson: I'm running it through ec2 now [06:19] mwhudson: I'm expecting the CVS failures, but the others should be fine [06:19] thumper: cool [06:21] mwhudson: review diff up [06:21] mwhudson: I'm going to make dinner now [06:21] thumper: ok [06:21] mwhudson: I'll check on the ec2s later [07:05] Argh [07:05] The lazr.enum import problems on karmic are back [07:08] :( [08:22] I'm done for the day. G'night all. [08:22] Night jml [08:31] good morning [08:40] mwhudson: with loggerhead, is generating a revision graph cache CPU intensive? === henninge_ is now known as henninge [09:06] Morning === jtv1 is now known as jtv [09:21] Who's dev-ing on karmic these days, I could use a fails-for-everyone-or-just-me confirmation [09:22] See LaunchpadOnKarmic for tests which are failing these days for me even with python2.4 [09:28] maxb: Both issues confirmed. [09:29] thanks [09:29] (and gah :-) ) [09:29] Something like that. [09:29] Does buildbot use Hardy? [09:31] On the plus side, as of last night, all of the interesting bits of my python2.5 branch have landed on devel [09:31] What remains? Just the s/2.4/2.5/? [09:33] that, and an increased startup timeout for the test librarian which may or may not still be relevant [09:33] That zope.sendmail fix ended up upstream? [09:37] Gary ran with that one and got it done :-) [09:38] Very good. [09:52] losa: ping [09:52] thumper: hi [09:53] mthaddon: the pqm seems wedged, the csvcs merge of mwhudson hasn't seemed to progress for several hours [09:53] mthaddon: is it the actual test run wedged, or just the ui? [09:53] checking [09:53] ta [09:54] yeah, going to kill the job - see a few zombied processes [09:54] thumper: I'm afraid he'll need to resubmit [09:54] mthaddon: ok [09:55] mwhudson: got that? [09:56] mwhudson: Do you know if generating the graph revision cache in loggerhead is cpu bound or io bound? [09:58] night all [10:03] stub: any idea what would cause this simple query to take so long? https://lp-oops.canonical.com/oops.py/?oopsid=OOPS-1355EC1268 [10:05] bigjools: Something else must have had a lock on that row [10:06] yeah that was the only thing I could think of too, but I can't think what would [10:09] bigjools: I think it can also lock if there have been inserted rows not yet committed - until the commit happens, PostgreSQL can't tell if your update is going to cause unique constraint violations or not. [10:10] good point [10:10] bigjools: Perhaps the trigger was another script hanging mid transaction? [10:12] no idea what would be updating stuff that Build references, unless ... [10:12] builder maintenance I bet [10:12] ok thanks stub [10:13] if rows have been updated in the builder table, that could block too since PG needs to validate the foreign key reference. [10:15] mmmm I get an error about diverged branches when rf-get is updating dulwich [10:19] bigjools: did you try removing it and getting it fresh? maybe you've made some changes locally by accident? [10:20] intellectronica: definitely not made local changes! [10:20] bigjools: anyway, i see that gavin just wrote about this to the list, so i guess you're not alone (although i just updated and didn't experience this) [10:21] I suspect someone uncommitted a revsion on LP [10:21] and those of us that had that revision are getting issues [10:30] bigjools: dulwich gets rebased sometimes. [10:30] I don't even know what it is :) [10:30] The Python git implementation. [10:30] Used by bzr-git. [10:30] ah [10:30] bigjools: it probably got replaced with another branch [10:31] bigjools: my local revno is 297, and lp:s one is at 413 [10:31] yeah, when I did pull --overwrite a lot of stuff was deleted [10:36] fabulous rant by Cody about reST :D [10:59] Morning, all. [11:01] Morning deryck [11:14] bigjools: I got the same error doing an update the other day. I pull --overwrited my dunwich and things went fine after that. === mrevell is now known as mrevell-dejeuner === matsubara-afk is now known as matsubara === bigjools-afk is now known as bigjools_ === mrevell-dejeuner is now known as mrevell [13:33] henninge: From your new home page screenshot, it looks to me like there is no longer a way to get to the app homes. [13:33] wgrant: you know what, I was just thinking about that ... ;-) [13:33] seriously [13:34] henninge: Why should it be locationless? [13:34] wgrant: hm [13:34] wgrant: I went by Martins mock-up [13:35] beuno' mock-up that is ... [13:35] Yeah. [13:36] just trying to rattle his cage but I guess he is really not here ... ;) [13:37] wgrant: maybe I'd put them in the list on the right, above the featured projects. [13:38] or linkify from the list on the left for not logged-in users... [13:38] But that list is gone for logged-in users, isn't it? [13:39] wgrant: yes, so it's either on the left or the right, depending on your logged-in state ... [13:39] the right list is not visible for logged-out users [13:39] Ah, forgot that bit. [13:40] normally I'd call this kind of inconsistency terrible, but usually people stay logged in once they have an account [13:40] and very rarely use the hompage. [13:40] Right. [13:41] at least, that is my expectation, don't have any figures on the logg-in-out behaviour of LP users ... ;-) [14:06] beuno: Hi; sorry to interrupt your vacation and all, but do you have a moment to take a look at lp:~gmb/launchpad/bugtask-index-conversion ? It's - well, it's the bugtask-index redesign branch. === gmb_ is now known as gmb [14:23] morning Launchpad! [14:23] Code, Translations are done! [14:23] Registry -> 2 to go, Bugs -> 3, Answers -> 3!!! [14:24] blueprints -> 19 [14:24] we are almost there [14:26] flacoste: the 2 in registry are going to be "fun" [14:26] right, person-edit and person-index [14:28] person-index is done! [14:31] salgado: you rock! === henninge is now known as henninge-bbl [15:02] salgado, do expect some test failures in your breadcrumbs branch, translations team just landed a few page_title introducing branches === danilo_ is now known as danilos [15:06] danilo-afk, ok, thanks for the heads up [15:12] BjornT, i'm being asked to clear out all my outstanding branches, which includes https://code.edge.launchpad.net/~leonardr/launchpad/remove-edwin-code/+merge/7253 [15:12] i never merged that because i thought you wanted to use it [15:12] is that still true? [15:19] Chex, gary_poster, rockstar, bigjools_, henninge-bbl, sinzui, intellectronica, Ursinha: LP production meeting in 40 min at #launchpad-meeting [15:19] leonardr: i don't need that branch to be merged. the important changes were lines 32-35 in the diff, and that change is already merged. [15:20] matsubara: allenap is the new intellectronica [15:20] bjornt: ok, i'll just merge it on my own [15:20] intellectronica, thanks. I'll update the MeetingAgenda page [15:20] matsubara: we're training him in the stanislavski method. he's still working on the accent. be easy on him ;) [15:21] matsubara: thanks [15:21] * allenap quickly looks up stanislavski so he can take part in the charade [15:21] in the API, is project_group.all_milestones sorted at all? [15:21] hi sinzui [15:21] hi bac [15:22] allenap: http://en.wikipedia.org/wiki/Stanislavski's_system [15:22] intellectronica: I'm there :) [15:22] sinzui: so the idea with person+edit is all of those g*&^%d(&ed lozenges go away, leaving a nice, clean edit page. in doing so, a bunch of tests will break. that's it? [15:22] hehe [15:23] bac: That sums it up [15:23] sinzui: what about the 'other actions'? will they stay at the bottom? [15:23] bac: I want them removed, but clearly that is a lower priority [15:24] ok [15:25] salgado: i've merged in your person-three-o branch into my work. please let me know if you update it [15:26] What? How can devel be out of date compared to my local copy? Has someone been uncommitting? [15:26] bac, will do [15:42] intellectronica, can we have a short conversation about lp/picker.js? [15:43] leonardr: sure. skype? [15:43] leonardr: maybe ask EdwinGrubbs to join too? he knows much more about it [15:43] sure, but your name is on the code that concerns me [15:43] basically i'm trying to merge https://code.edge.launchpad.net/~leonardr/launchpad/remove-edwin-code/+merge/7253 [15:44] which has been unmerged for a long time [15:44] basically it stops the javascript web service client from requesting an XML rendition of a resource and then parsing the xml [15:44] leonardr: right, and meanwhile i've inlined some hacks to parse the xhtml result into picker itself [15:45] EdwinGrubbs: coordinate with gary_poster. He has started some of the tempaltes. He know which ones need love. [15:45] intellectronica: so when you do Y.DOM.create(entry) and all the code after that, that's parsing the xhtml? [15:46] leonardr: yeah, that's a nasty hack for parsing the result [15:46] well, parsing the result and getting the dom structure for the embedded value [15:46] Thanks sinzui. Hi EdwinGrubbs :-) thanks for review. Trying to DTRT for those templates now. I'm going to make a pastebin for me to share the info I have. [15:47] intellectronica: ok, the point we want to get to is where save() sends picker_result.api_url to picker._resource_uri + '/' + attribute_name [15:48] basically we tell launchpad to modify that single field [15:48] and the xhtml response we get back will be that single field [15:48] no need to parse anything [15:48] leonardr: awesome === danilos is now known as danilo-afk2 [15:49] leonardr: we still need to DOM.create to be able to use the result in a page, but there's no need to do all the complicated field by field iteration. that's really great [15:50] well, i suppose we can use innerHTML too, that may be more efficient [15:50] intellectronica: ok, how would you feel if i simply put my new code into place and let you make the change to picker.js? [15:51] leonardr: this is not a very good time to have broken code in trunk. maybe try to land both changes together? [15:51] gary_poster: it sounded like sinzui was suggesting that you had unconverted templates that I could work on. [15:51] intellectronica: i was saying that on the assumption your code wouldn't break [15:52] leonardr: ok, maybe i need to understand how the new way of patching and retrieving a single field works [15:52] EdwinGrubbs: http://paste.ubuntu.com/272894/ [15:52] you already have the hack in place locally, and BjornT said the relevant part of my branch was already landed [15:53] EdwinGrubbs: So you could take over anything in lines 30-end. [15:53] leonardr, intellectronica: I am seeing doc/launchpadlib.txt errors in buildbot, would you guys know something about it? https://lpbuildbot.canonical.com/builders/lp/builds/141/steps/shell_7/logs/summary [15:53] intellectronica, take a look at line 32-35 of https://code.edge.launchpad.net/~leonardr/launchpad/remove-edwin-code/+merge/7253 [15:53] leonardr: if it doesn't break then that's fine. my only worry was that i'll have to take care of a broken result now, because i don't really have much time. if it doesn't break and just requires a cleanup later on then i'm happy to take care of that [15:53] deryck: could you look at http://paste.ubuntu.com/272894/ , lines 38-46? [15:53] (just because you are discussing launchpadlib, no other reason to think you might know more :) [15:53] deryck: those are pages in "launchpad" that Francis said are for bugs somehow [15:54] gary_poster: can I put that in a wiki page, so I can mark which ones I've started on? [15:54] intellectronica: i feel that if it was going to break, it would be already broken [15:54] EdwinGrubbs: absolutely, that would be much smarter [15:54] danilo-afk2: it looks like the server isn't accessible when that test runs? [15:54] gary_poster, looking [15:54] ty [15:54] leonardr: how did you test this? [15:55] intellectronica: i haven't tested it yet. i'm just trying to figure out where all the pieces of code go [15:55] has it really been merged? if yes and we don't have problems then i suppose you're right and it doesn't break [15:56]