[02:15] I was looking at fixing Bug #151113 . It looks pretty straight forward, Matthew did a great job of explaining exactly what needs to be fixed [02:15] Bug #151113: Notification footer has unnecessary leading spaces [02:38] nhandler, cool. [02:39] jml: The only thing I'm not completely sure about is what test I am meant to run. [02:44] I'm also not fully sure how to Demo it. AFAIK, rocketfuel does not setup a working mail server, so I would think that email notifications would not work [02:45] nhandler: It will use the local MTA and send emails to root. [02:45] Ok, thanks wgrant [02:45] nhandler: I just have postfix running with a local-only config, alias root to my user, and point Evolution at my mbox. [02:50] jelmer, https://bugs.edge.launchpad.net/launchpad-code/+bug/221370 [02:50] Bug #221370: Cannot checkout a Launchpad hosted project [02:52] wgrant: But would a MTA be setup on whatever machine they attempt to verify the patch on? I know that I currently don't have a local MTA setup, which causes some errors when it attempts to send the message [02:56] nhandler: The test suite uses a separate email-sending stub, so does not care about the local MTA. [04:41] sinzui, are you around? [04:54] jml: I am about [04:55] sinzui, I was going to ask you about which table links packages to productseries, but I've since learned that it's 'packaging' [04:55] I thought I had another question too, but I've lost it in my task switching [04:56] jml: yes. A surprising name [04:57] sinzui, there was also some confusion, given that the table has changed recently [04:57] I think it's quite a reasonable name. [04:57] wgrant, the two are not mutually exclusive :) [04:58] jml: I would have moved the objects to the registry months ago it was names SourcePackageLink [04:59] I think I would rename SourcePackage to SourcePackageRelease now that I understand it better [04:59] Huh? [04:59] You mean DistroSeriesSourcePackageRelease? [04:59] s/Release// [04:59] SourcePackageRelease is something else. [05:01] The sourcepacakge is a very mutable thing. it is just a meta object for for DistroSeriesSourcePackageRelease. So when I visit the url for a DistroSeriesSourcePackageRelease, I actually get the meta object that lists the real object [05:01] Currently a SourcePackage is a meta-object for many SourcePackageReleases. [05:01] Oh, right, I see. [05:02] TheseNamesAreTooLongAndConfusing. [05:02] sinzui, SourcePackage should be DistroSeriesSourcePackage [05:02] jml: yes, that might be better [05:02] wgrant, the model is too big and confusing [05:03] wgrant, which might be because what it models is too big and confusing [05:03] jml: I think it's more of a problem that it has names like IDistroArchSeriesBinaryPackageRelease in it. [05:04] I think a SourcePackage is wrongly a QuestionTarget because it's short name lead someone to think it was a long lived object. To make the implementation work we had to use a DistributionSourcePackage. So the interface and documentation is all messed up. The names are too long an still not clear [05:05] There's also the horrible two versions of DistributionSourcePackage which makes some stuff pretty confusing. [05:05] * sinzui needs a nice picture written in crayon to understand how this works [05:05] I only know one, and I don't know which one that is [05:05] (one is a DB object introduced for package-specific bug reporting guidelines. it is used for nothing else. the other is a separate object generated on the fly.) [05:19] how many packages do you think are linked to a productseries? [05:20] I'm doing some queries and trying to sanity check my results [05:28] jml: I would suggest a little over 3000, but the UI may well be lying. [05:28] wgrant, where are you seeing that in the UI? [05:28] https://edge.launchpad.net/ubuntu/karmic/+packaging [05:28] ta [05:29] wgrant, that doesn't look like 3000 to me. [05:30] it looks more like a few hundred. [05:32] hi mr. jml [05:33] jml: Over several series! [05:33] mwhudson, hi [05:33] wgrant, cool. that matches what I'm seeing with my queries. [05:33] thanks. [05:33] mwhudson, I was sorry to hear you were unwell. [05:34] jml: it's the same thing as i was under on friday [05:34] mwhudson, :( [05:34] i feel a bit better now, but as i've spent all day in bed up to now it's hard to separate cause and effect [05:34] heh [05:34] jml: must be late where you are? [05:34] mwhudson, it is about 11:30 [05:34] oh ok [05:35] mwhudson, but I had an urge to get this graph done [05:35] it turns out that we have many, many fewer upstream links registered in Launchpad than I had expected. [05:35] Well, the UI has always sucked. [05:35] Until sinzui fixed it last week. [05:35] i was vaguely aware that the count was a bit embarrassing [05:36] wgrant, you are correct, and sinzui is awesome [05:36] wgrant, and I was foolish for being optimistic :) [05:36] hey, what do you think a good filter for "working branch" would be? [05:36] If I was awesome I would have merged the packing linking UI like beuno ask me too two months ago [05:36] mirror_failures = 0 & last_scanned_id is not null? [05:36] As well as the UI sucking, there is little to no benefit. [05:37] Besides upstream bug linking becoming slightly easier. [05:37] But the UI is still scary, as it screams that you should almost never link trunk to a package, when that is in fact what you normally want to do. [05:37] hmm [05:37] * jml should actually look at the UI [05:38] (although that's perhaps a manifestation of another bug -- LP doesn't know much about all these placeholder projects) [05:38] I want to make sure that we have trunk imports for as many upstreams of packages as possible [05:38] Maybe once you have bzr-svn. [05:38] but perhaps I need to either rethink that or the way I am gathering my data about this. [05:38] wgrant, any moment now. mwhudson has promised it one hour after we have python 2.5 :) [05:39] also, jelmer says bzr-hg is finished. [05:39] I thought bzr-hg was very much not. [05:39] That's very good news. [05:39] jml: oh, awesome [05:39] Plus 2a doesn't suck. LP might be ignored less now. [05:39] (about bzr-hg) [05:40] I think we lost a lot of GNOME links because of the switch from main to master. We should be able to fix all those automatically [05:40] sinzui: How good is the p-r-f at creating appropriate series? For how many projects will it work? [05:41] Proper upstream linking would such less if one didn't have to find a project admin to create series first. [05:42] wgrant, all of these ideas are making me want a tag and a bunch of bugs [05:42] jml: Sounds like you need a BugBag™. [05:42] wgrant: It cannot create a series, in fact, the series must exist so that the PRF can be instructed how to find releases [05:42] sinzui: Oh, blah. [05:43] wgrant: now that the PRF is very good at finding files, users see that our hack to make milestones work with project-groups is does not work with projects that do not adhere to debversion names [05:44] sinzui: Does anyone use ProjectMilestones besides LP? [05:44] bazaar might [05:45] They work for GNOME's core apps/libs, which I think is about 10% of all GMOME projects [05:46] They do not work for Launchpad, we have lots of projects like lazr or oops tools that are versioned differently [05:47] They do work, in the most part, for LP's ridiculous use of products. [05:48] yeah, they work for a pathological case \o/ [05:53] prf doesn't work for pypi though, so I guess I should submit a bug on that. [05:57] <[reed]> where's kiko when you need him [06:00] Texas [06:01] stub: The log I have from two weeks ago shows the PRF reading pypi [06:01] I've sent some data & queries to the launchpad-dev list [06:02] stub: was this for pytz? Nothing was found for that [06:02] wgrant, sinzui: I'd really appreciate it if you could post your thoughts re linking so they don't languish here on IRC beyond thought and memory [06:02] jml: Those Lucid numbers are really encouraging. [06:03] wgrant, seriously? [06:03] jml: No. [06:03] jml: Also, your numbers for packages in main are wrong. [06:04] jml: I believe 1) creating a series from a parent should preserve links. 2) we can prepare a script to create the missing links to lucid [06:04] There are maybe 4000. Not 16000, although that will change in the next few months when the components are changed. [06:05] wgrant, what do you think I got wrong in the query? [06:05] jml: Ah, I didn't see the queries. [06:05] * wgrant looks. [06:05] jml: You're not doin any restriction over components there. [06:05] sinzui, cool. as I said, I'd appreciate email. [06:06] jml: I am preparing a report of sorts myself. I am looking at each gap between Ubuntu and an upstream to get a count of how many packaging links are missing, how many series are missing, how many series are missing branches how many missing projects. [06:06] jml: I hope to finish tomorrow and send it to launchpad-dev. [06:06] wgrant, of course. [06:07] ta [06:10] * jml pastes the IRC log into an easily findable file so that all the good ideas don't get lost. [06:11] jml: I'm not sure why you care about -updates in those queries. That will just duplicate packages unnecessarily. [06:11] Since except for very exceptional cases, -updates will not contain anything new, nor will it contain component changes. [06:11] wgrant, me neither. I was heavily advised :) [06:12] wgrant, I'm about to go to bed, so I won't do anything more on this tonight. [06:13] jml: Ah, you're at UDS? [06:13] wgrant, like the dickens. [06:13] wgrant, I looked for your name on the attendees list. [06:14] wgrant, and thus I assume you aren't here :) [06:14] jml: Correct. [06:15] wgrant, a shame. [06:16] jml: Well, I'm too frustrated with Ubuntu development to do anything useful there at the moment. [06:16] I'm off to bed, but before I go, I think you are mistaken jml. you are the one is Dallas Texas. There is a lot of same there. [06:17] sinzui, :D [06:17] wgrant, that sucks. [06:18] sinzui: Yes - pytz. I was assumng it doesn't find the links because of all the foo.tar.gz#md5sum=4545564564 links [06:19] sinzui: Are you aware of any projects successfully getting picked up by prf? [06:19] stub: I'll take a look at the issue during my lunch tomorrow [06:19] stub: yes I have heard from many projects that are very happy with it now. [06:20] stub: The last two fixes I landed were were diagnosed by users how read the code and identified why their files were missing. [06:21] sinzui: Know a product name? I might just need to copy their URLs. [06:23] stub: https://edge.launchpad.net/python-phynx/0.10 [06:24] stub: and this one too https://edge.launchpad.net/python-quantities/0.5 [06:24] g'nigh. [06:24] t [06:25] Night jml [06:26] I had http://pypi.python.org/pypi/pytz/pytz-*.tar.bz2 for my release URL. Others seem to be using http://pypi.python.org/packages/source/p/phynx/phynx-0.10.*.tar.gz. Didn't know about that second URL type, but it omits all the #md5 trailers from the URLs. [06:27] stub: I test the file globs in a launchpad.dev series then run the script in cronscripts [06:28] We should send owner emails when there are errors, and I think I know how to test the location to show the owner what files the PRF will see [08:36] good morning [08:48] elmo: Off topic - Are you still looking for a sys admin in the Asia/Pacific region? I have just read the posting, and I tick most of the boxes. from NZ, I am Debian dev, been using linux since 1993. [08:54] I'm a newbie... Just running rocketfuel-get and I am getting errors about missing branches from bzr... is this normal? [08:56] matgeek: If it's just shipit and canonical-identity-provider, that's normal. [08:56] matgeek: Those two components remain proprietary. [08:57] wgrant: Thanks, that's what they are, but then I get an exception form bzr about a.git not being there when trying to remove it -- seems a bit disconcerting? [08:58] matgeek: What exactly is that error message? [08:59] wgrant: 'OSError: [Errno 2] No such file or directory: '/home/grantma/launchpad/lp-sourcedeps/sourcecode/a.git'' [09:00] wgrant: along with a stack trace, which does not seem to be a tidy way of handling it! [09:00] matgeek: That's odd. I haven't seen it before. [09:01] matgeek: pastebin the full traceback, please. [09:01] wgrant: I'm dumb, where's the pastebin site please? [09:01] matgeek: http://paste.ubuntu.com/ [09:02] wgrant: OK [09:03] wgrant: up at http://paste.ubuntu.com/319863/ [09:05] wgrant: bazaar 2.0.2-1~bazaar1~karmic [09:07] matgeek: That's special. Rerun rocketfuel-get when it finishes, and see if it happens again. [09:13] wgrant: Already done that. That's a pastebin of a rerun, but I will try again. [09:16] matgeek: Does a.git exist where it says it does? [09:17] wgrant: The parent of it exists, which is full of other dirs, so I expect if it did it exist it would be a dir to? [09:18] wgrant: ie a.git directory does not exist [09:19] wgrant: would you like to ssh into my launchpad-dev virtual, and have a look first hand? === matsubara-afk is now known as matsubara === salgado is now known as salgado-sick [13:42] matgeek: have you fixed the a.git issue? [13:59] How's it going Launchpad people? [14:12] good morning [14:13] flacoste, yui3 didn't land on Friday. Apparently I am "not authorised to commit" to db-devel? [14:13] mars: you are [14:13] not according to PQM :) [14:13] mars: what --dry-run gives you? [14:14] mars: the commit list is shared between devel and db-devel [14:15] flacoste, --dry-run gives no errors. Same as Friday. It took a while, but both of my submissions came back with the "not authorized" error. [14:16] flacoste, here is the command I used: bzr pqm-submit --submit-branch=bzr+ssh://bazaar.launchpad.net/%7Elaunchpad-pqm/launchpad/db-devel/ -m "[rs=flacoste] Update launchpad to use YUI 3.0.0." [14:16] mars: first of all, the commit message isn't correct [14:16] oh? [14:16] you'll need something like [r=flacoste][ui=none] [14:16] rs is fine [14:16] but use r, since i did review it :-) [14:17] ah. Everyone else in log uses just [rs=...] [14:18] ok, my first submission was [r=flacoste], but since it didn't stick, I tried [rs=flacoste] instead [14:21] mars: that's a special case for the buildbot-poller [14:21] rs=buildbot-poller [14:21] flacoste, gary_poster, I think merging yui3 should wait until after Python2.5, given that the JS build system broke on staging when it was running the python2.5 branch. [14:22] mars: :-( [14:22] gary_poster, just a suspicion. BjornT_ noted that only staging was broken in a specific way, and that the difference was it's running of python2.5 [14:22] mars: you can work that out now with the pertinent branch if that helps; and I'm happy to help [14:24] gary_poster, I'd like to take a look at staging first [14:24] to isolate the problem BjornT_ pointed out [14:25] mars: ok. let me know if I can help. [14:32] bac: EdwinGrubbs: stand-up in 2 minutes === matsubara is now known as matsubara-lunch === leonardr is now known as leonardr-away === leonardr-away is now known as leonardr [16:05] flacoste: is the UI meeting canceled? [16:05] sinzui: yes, many of the attendees are at UDS [16:05] thanks [16:05] noodles, beuno, rockstar [16:05] barry === matsubara-lunch is now known as matsubara [16:06] and intellectronica [16:31] Chex do you know why staging is not updating? [16:33] sinzui: I am not sure, let me take a look [16:36] bac: I updated the the two breadcrumb bugs with my thoughts about what we need to do to fix them I think each fix is about 50 lines. [16:42] sinzui: staging updates were disabled for gary testing there for the python 2.5 upgrade [16:43] Chex: staging has not been updated in more than a week. I cannot QA items that landed on the first day of week one [16:43] that's pretty much done, but aiui we are also waiting for stub to fix something. Does that jibe with the notes, Chex? [16:44] Chex: gary_poster: Have we had staging disabled for that long because of py2.5? [16:44] Chex, no. Mine was from this past Friday, and was added to some kind of disabling because of stub, from what Tom told me. [16:44] oh, sorry, sinzui ^^^ [16:47] Chex, sinzui, you can remove the Python 2.5 testing from the reasons to disable staging. I think we are good. (Still check on the stub thing though, please.) [16:47] I think the 23rd is the latest I can risk with QA. I still have bitter memories of being locked out of staging and then spend release week with the entire team trying to get RCs [17:02] sinzui: gary_poster: ok, the hold for stubs landing has completed, so were all set with staging _restore and I have re-enabled it now. [17:03] Chex: \o/ [17:08] sinzui, ping [17:09] Hi mars [17:14] So, is staging currently Python 2.4 or 2.5 ? [17:18] maxb, staging probably is 2.5. Is it important to establish confidence? [17:19] Just curious, really. And wondering how automatic updates could be enabled if it was still 2.5, with 2.5 not on devel yet [17:33] the previous build is wiped away === abentley is now known as abentley-lunch [18:14] sinzui: "howdy!", did you submit wgrant's branch? [18:16] bigjools, I thought you guys were going to waverly. [18:16] bigjools: I did, but it failed. The branch has db changes so had to go to db-devel. There were conflicts [18:17] cody-somerville: we're in the room opposite [18:17] sinzui: argh, ok, thanks [18:17] he will have to fix those then [18:18] I showed him the conflicts [19:35] hi sinzui [19:35] hi bac [19:36] sinzui: i've changed the map portlet to not be shown at all (from the

Location

onwards) if no data is set and the viewer is not the user. [19:36] it seemed dumb to just show the tz when it is shown above [19:37] bac: Yes, I think that is a very sensible decision. [19:38] rockstar, ping: during the sprint, did you switch lazr-js over to use explicit references to the needed version of YUI? [19:48] morning [19:56] rockstar: ping [19:57] fuck, my keyboard and editor is suddenly typing in Cyrillic! [20:00] sinzui: Яeally? [20:00] Really it was, I had to paste that message together [20:05] after updating, i can't seem to run tests anymore. i get the following error: [20:05] Traceback (most recent call last): [20:05] File "bin/test", line 174, in ? [20:05] from subunit import TestProtocolClient [20:05] File "/home/tom/launchpad/lp-branches/bug-481356/lib/lp/scripts/utilities/importfascist.py", line 163, in import_fascist [20:05] module = original_import(name, globals, locals, fromlist) [20:05] ImportError: No module named subunit [20:05] anyone knows what's going on and what i need to do to fix this? [20:07] intellectronica: I wonder if there's a new dep in launchpad-dependencies that you need? [20:07] some python testing module that would be brought in, maybe? [20:10] good morning [20:10] mwhudson: how are you feeling today? [20:10] thumper: mostly better, we'll see how it goes [20:11] flacoste: it seems we have multiple large distruptive type changes landing, how are we co-ordinating python 2.5 and YUI 3 releases? [20:12] kfogel: i'm up-to-date, so if there's a new dep it must not have been added to our dependency package? [20:12] gary_poster: maybe you know? ^^^ [20:12] thumper: ask gary and mars [20:13] :) [20:13] thumper: atm, Python 2.5 wins, thanks to mars generosity. [20:13] thumper Python 2.5 is landing now, hopefully [20:13] cool [20:13] gary_poster: will you be emailing the dev list once it is done? [20:14] after installing subunit manually i'm good, it seems [20:14] so i guess it is indeed a missing dependency [20:15] thumper: I intend to mail everyone before then. Once it lands, there will need to be some clean up; if I get https://bugs.edge.launchpad.net/launchpad-foundations/+bug/481512 reviewed, then once it is landed it will just mean that you ned to do a make clean and make [20:15] Bug #481512: race condition when rotating logs [20:15] Hm, not that sorry [20:16] https://code.launchpad.net/~gary/launchpad/py25makeclean/+merge/14916 [20:16] mwhudson, abentley: lets try skype [20:16] intellectronica: I'm not familiar with that [20:16] intellectronica: sounds like something from jml/allenap. Let me see how you did that; maybe I can dupe [20:16] gary_poster: i suspect it might be related to the parallelization work allenap and jml have been doing [20:17] intellectronica: precisely [20:17] i thought subunit was in sourcecode or an egg or something [20:17] heh [20:17] yeah, i have it in sourcecode/ [20:17] subunit=lp:~launchpad-pqm/subunit/trunk/ [20:17] gary_poster: i did that by running bin/test [20:17] mwhudson: well, since installing the package works for me, why not just rely on that? [20:17] intellectronica: ^^ [20:17] gary_poster: approved [20:18] or is that version missing something i might need later? [20:18] abentley: we can hear you [20:18] intellectronica: um, why not install twisted or bzr from a package then? [20:18] abentley: let me try hosting [20:18] thumper: I'm getting that techno thing. [20:18] or zope or ... [20:18] thumper: thanks! [20:19] abentley: this sucks arse [20:19] abentley: do you have sip working? [20:19] mwhudson: donnow? maybe because the packaged version is too old? [20:19] intellectronica: I just tried ./bin/test -t test_all_scripts and it worked. can try another [20:19] thumper: No, I don't. Haven't tried. [20:19] intellectronica: in any case [20:19] abentley: kfogel, mrjazzcat and I did a three way sip call and it was perfect [20:19] intellectronica: i think the thing to worry about here is why you don't have sourcecode/subunit [20:20] we found that twinkle worked best [20:20] i don't appear to have that installed [20:20] I didn't have it installed either [20:20] gary_poster: to be precise i ran `bin/test -vv -t xx-milestone-add-and-edit.txt` but it can also be that you already have subunit installed [20:20] thumper: Installing that. What will I need as credentials? [20:20] thumper: I've since heard from elsewho that twinkle is Da Bomb too (Bradley Kuhn at SFLC uses it all the time). [20:21] abentley: you should have them in an email form IS a long time ago... [20:21] abentley: I asked on the #is channel to get them sent to me again [20:21] Okay, I'll dig that up. [20:21] mwhudson: yes, that oo [20:22] thumper: any idea what the subject: of the email was? [20:22] Your Canonical VOIP account details [20:23] intellectronica: FWIW, I ran that test command successfully, and do not have subunit installed. [20:23] hm, no hits [20:23] intellectronica: no idea in specific, but would vaguely try make clean and make if it were I [20:28] intellectronica: I am willing to help if you want. Also, if we can get that working and you are not at EoD, I'd be happy to have you be the guinea pig for migrating to the Python 2.5 branch, if you were willing. [20:29] gary_poster: sorry, if i'm unresponsive, i'm at UDS. anyway, it's working for me now, but my guess it's a slightly different different setup from yours since i have subunit installed from a package and not in sourcecode [20:30] gary_poster: and sure, i'm happy to try py2.5 later on [20:30] ...huh, weird. ok. Py 2.5 ok thanks. Ping me when you are available and we'll give it a whirl [20:33] gary_poster: sure thing. going into a couple of sessions in ~15m, but happy to give it a try after that [20:34] ok cool [20:34] thanks === matsubara is now known as matsubara-afk [21:37] al-maisan: Did the instructions otherwise work? [21:38] wgrant: I'll be able to tell you tomorrow -- I need to finish another bit of work before I can try them. [21:38] al-maisan: Ah. === Ursinha changed the topic of #launchpad-dev to: This is Launchpad Development Channel | Week 2 of 3.1.11 | PQM is open | I am Zero OOPS and So Can You! http://is.gd/4fkLl | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | This channel is logged: http://irclogs.ubuntu.com/ [22:36] gary_poster: do you know where I can find the bug filed last week about HTML support in named operation? [22:37] flacoste: I seem to recall it being in lazr.restful [22:37] gary_poster: nm, just found it: https://bugs.edge.launchpad.net/lazr.restful/+bug/480192 [22:37] :-) [22:37] exactly [22:40] um i guess "someone" should look at landing bzr 2.1b3 into rf [23:10] mwhudson: was that a dual volunteering I note ^^ ? [23:10] spm: no [23:10] * spm breathes a not too small, not too large, sigh of relief [23:52] I'm getting a "No module named boto" when trying to run utilities/ec2. What can I be doing wrong? [23:52] Ursinha: Have you python-boto installed? [23:53] wgrant, I guess so, unless something removed that [23:53] wgrant, and something did that :/ [23:53] Ursinha: :( [23:54] wgrant, I thought for a while this could be launchpad issue, because it was working after the last branch update I did [23:54] crazy [23:54] thanks wgrant [23:56] * mwhudson stops writing a crazy shell script