[00:08] rockstar: around for a chat? 5m? [00:08] thumper, dog's in the backyard. 5m is good. [00:13] thumper, call anytime. [00:15] thumper, plugged in, my headset was not. [00:15] rockstar: I'll call you when I've finished nom'ing [00:15] thumper, okay. [00:36] http://howsmycode.com/ [00:37] lifeless, awesome. I need to take a look at this. [01:05] mwhudson: Is the image happy now? [01:11] wgrant: seems so [01:11] wgrant: i guess i should boot it and check all the versions are right, can you tell me what versions i should look for? [01:14] mwhudson: dpkg-dev >= 1.15 [01:15] * wgrant checks the exact version. [01:15] dpkg-dev 1.15.4ubuntu2~launchpad1~bigjools1 is the important one. [01:17] ok [01:17] * mwhudson taps his fingers [01:19] Hm, so 3.1.12 is actually only two weeks long? [01:19] wgrant: ii dpkg-dev 1.15.4ubuntu2~launchpad1~bigjools Debian package development tools [01:19] looks ok [01:19] mwhudson: Thanks muchly. [01:20] Now... the buildbot AMIs need doing too. spm? [01:20] wgrant: once we get lp-deps updated, sure [01:21] but perhaps not the day before a release :-) [01:21] spm: Probably not, true. [01:23] tests appear to be running, should make this image public then i guess [01:24] mwhudson: If you want to test that it actually works, lp:~wgrant/launchpad/distroseries-source-format-selection is the branch that needs it. [01:24] wgrant: would running the tests on that branch be useful at this stage? [01:26] mwhudson: It's done and approved, but hasn't had a full test run yet, so probably. [01:26] wgrant: ok i'll give it a go [01:27] mwhudson: Thanks. [01:46] wgrant: you can follow along at http://ec2-75-101-192-167.compute-1.amazonaws.com/current_test.log [01:46] (assuming i did the security group thing right) [01:46] mwhudson: You did. Thanks. [02:32] Who is this Vikram character, and why is he giving bad advice on Launchpad questions? [02:37] example? [02:37] https://answers.edge.launchpad.net/launchpad/+question/92507 [02:38] oh him/her. right. sigh. [02:38] I've seen many similar cases :( [02:38] spm: is staging down? [02:38] mwhudson: yes [02:38] spm: rather, is staging expected to be down [02:38] spm: ok [02:39] spm: eta on back up-ness? [02:39] mwhudson: wrong db configs/versions/brokenness I was hoping it'd fix itself. bzzt. [02:39] mwhudson: not yet, no. [02:39] wgrant: boggle [02:39] spm: ok [02:40] While that solution will probably work, it's not really a solution, and is wrong. [02:40] wgrant: they popped up on #lp last week. "I am **the** LP answer contact". Asked if they could have admin privs because of. I was impressed at the ignorance; polite, but impressed. [02:40] spm: Yes, I saw... [02:40] I was also impressed at my ability to not lol in their face :-) [02:41] spm: but why didn't you give them admin privileges? You're so mean... [02:43] ajmitch: it goes with the territory of being a sysadmin. sign says "" on the door. sorry. ;-) [02:46] spm: and I suppose you have the BOFH calendar on your desk as well [02:46] the 128 column dotmatrix one? not anymore. ;-) [03:02] mwhudson: oh bugger. michael did a private branch landing on staging - and db update as well; hence the boomsville. crap. will need to disable staging updates, as they wiped the merge. [03:02] spm: yay [03:08] mwhudson: should be back-ish now. [03:15] class IPackageUpload(Interface): [03:15] """A Queue item for Lucille""" [03:15] this is not all that helpful [03:15] wgrant: what's 'lucille' ? [03:16] mwhudson: Looooong before my time, but I believe it was the original name for archivepublisher/archiveuploader/buildmaster [03:16] ie. Soyuz [03:16] oh [03:17] so that really isn't a very helpful docstring [03:17] back when everything was a proper name, like in debian [03:17] Not terribly helpful, no. [03:18] What's left? gina, zeca, poppy... [03:18] Must be something else. [03:19] There used to be a list somewhere of the names that Debian used, I don't know if there was something for ubuntu & launchpad [03:20] The Debian list (inside the dak source) includes a list of Canonical names. [03:25] wgrant: what does gina do? [03:25] mwhudson: Imports Debian archives into LP. [03:25] that's another thing i've never gotten straight in my head either [03:25] oh ok [03:25] Or, more often, breaks. [03:30] heh [03:32] does it break like the package branch importing is sometimes doing these days? [03:32] That relies on gina to notify it of new Debian packages. [03:33] wonderful [03:34] mwhudson: Solved Soyuz yet? [03:34] wgrant: haha [03:35] wgrant: I thought you were doing all the soyuz fixing :) [04:42] wgrant: lp:~wgrant/launchpad/distroseries-source-format-selection passed tests [04:42] mwhudson: Excellent. Thanks. [04:43] i could forward you the email i guess, but it's not very exciting [04:43] I would imagine not. [04:46] * thumper EODs [06:39] gosh, IBuild is a pretty fat interface [06:58] * mwhudson finds https://dev.launchpad.net/StormMigrationGuide and https://dev.launchpad.net/Database/StormMigrationGuide [07:03] * ajmitch wonders if they were meant to be 2 different URLs, or is one a redirect to the other? [09:01] good morning hackers [09:03] Hi bigjools! [09:03] :) [09:03] hey henninge [09:03] noodles775: Hi! [09:04] Hi henninge :-) [09:05] noodles775: Have you been using balsamiq? Do we have something like an "empty LP page" to start from? [09:05] henninge: no, I haven't, so not sure. [09:33] jml: you wouldn't happen to have a bzr plugin that queries launchpad and lists all my unmerged branches, would you? [09:33] BjornT_, sadly no. [09:34] BjornT_, I did, however, write this plugin yesterday, which isn't actually directly useful or related, but kind of cool: http://paste.ubuntu.com/332980/ [09:39] jml: i wonder, would it be hard to have a base 'lp bzr plugin' class, which would automatically log in to lp, and have switches for choosing whether to use edge or staging, and probably a few other useful things as well? [09:39] BjornT_, not hard at all! [09:39] BjornT_, I really want to get such a thing into bzr-core [09:39] yeah, that would be cool [09:39] BjornT_, but there are some dependency issues to navigate, so it should probably be in an external plugin for now. [09:40] BjornT_, btw, do you know about launchpadlib.uris? [09:40] BjornT_, this branch has my previous attempt to get that stuff into bzr: https://code.edge.launchpad.net/~jml/bzr/lp-login-oauth [09:41] BjornT_, I've also squatted 'https://launchpad.net/bzr-launchpad' just in case. [09:43] jml: i think i now a bit about launchpadlib.uris. i didn't add it, but i know where it comes from [09:44] BjornT_, I don't :) I only discovered it recently and thought "gosh, now I can delete code from a dozen little scripts" [09:44] (well, I can as soon as someone packages a version of launchpadlib that has the module) [09:45] jml: so if I'm looking at this API for blueprints, should I reassign the bug to me? [09:45] morning [09:46] ajmitch, with my blessing. [09:46] mrevell, hello [09:50] BjornT_: rocketfuel-status is supposed to do that [09:50] not as nice but it works [09:55] bigjools: not as nice == totally unreadable, and include all my abandoned branches. it also doesn't work for my lazr-js branches, and appears to take a really long time === salgado is now known as salgado-afk [09:57] BjornT_, oh, I should mention that I've written a plugin called 'bzr-removable' that looks at your branches on disk and tells you which can be removed (i.e. no changes, are merged) [09:57] it also has an inverse command, if you are into that sort of thing. [09:58] BjornT_, I haven't used it in anger since we got *four trunk branches* though. [10:25] jml: fwiw, http://tillenius.me/blog/2009/12/02/wanted-bzr-plugin-to-manage-branches-in-launchpad/ [10:28] BjornT_, I got a 500 when I tried to comment :) [10:28] jml: hmm... :) [10:29] BjornT_, the comment is there though, so no need to drop everything :) [10:30] jml: yeah, i noticed that now, after succesfully adding a test comment myself [10:30] BjornT_, Just Idented/Twittered your post :) [10:31] mrevell: thanks, i hope some will will actually implement it :) [10:35] BjornT_, well, I _do_ have a thirty hour journey this week [10:40] jml: that's nice to hear :) [10:44] james_w, hah! I am now tasting the bitter sting of 503 errors from launchpad APIs! [10:51] I wish I could take a train to Australia. It would be so much easier to get some coding done. [11:01] Morning, all. [11:03] mrevell, links to , which doesn't exist [11:03] thanks mpt, I'll fix that === matsubara-afk is now known as matsubara === salgado-afk is now known as salgado === henninge is now known as henninge-afk [11:58] matsubara: could you try editing a bug status and milestone on staging using chrome (for bug 490796)? i've confirmed it works in both firefox and opera [11:58] Bug #490796: "Target to milestone" takes you to +editstatus [11:58] BjornT_, sure [11:58] thanks [12:05] BjornT_, works fine on Chromium 4.0.257.0 (Ubuntu build 32926) [12:05] BjornT_, one small thing, when I click the green plus sign in the milestone control, I'm redirected to +editstatus [12:06] BjornT_, clicking on the "Target to milestone" link works as expected though (i.e. I get the overlay and can set the milestone accordingly) [12:16] matsubara: right, i see the same in opera and firefox :( [12:16] matsubara: it doesn't seem to be a regression, though. i see the same behaviour on edge and launchpad.net === henninge-afk is now known as henninge [12:18] BjornT_, ok, if it behaves the same in all the major browsers, it's fine. thanks for fixing it. :-) [12:18] o/` hug the mountain o/` [12:31] * jml -> lunch === henninge is now known as henninge-afk === henninge-afk is now known as henninge [14:56] reviewers, hackers, lurkers -> #launchpad-meeting in 5m === matsubara is now known as matsubara-lunch [15:00] reviewers, hackers, lurkers -> #launchpad-meeting [15:37] do we use the zope class registry anywhere? Is it safe to use within Launchpad? I am primarily interested in getClassesThatImplement(iface), please see http://docs.zope.org/zope3/Book/inspect/classregistry/show.html for details. [15:37] If not is there anything similar in Launchpad that I can use? [15:41] al-maisan, hmm. not that I know of. (Also, it sounds similarish to the bzrlib registry) [15:41] james_w, hello [15:42] hi jml [15:42] jml: thanks for the info! [15:43] james_w, I've got five things to talk with you about. Can we have a call sometime? [15:44] jml: sure thing [15:44] jml: this week? [15:44] james_w, yes please. I'll be in Australia next week, which will make voice comms more challenging [15:45] ok [15:45] (it's a long way to shout) [15:55] are there any plans for moving this wikipage to dev.lp.net ? https://launchpad.canonical.com/SearchingTranslations [15:56] it is linked from the Rosetta Search blueprint [15:56] adiroiban, Probably no plan but let me take a look at that page [15:56] danilos, are you happy for this page to move to the dev wiki? https://launchpad.canonical.com/SearchingTranslations [16:05] adiroiban: there are too many pages on that wiki, most are rubbish We have made every blueprint public that has been requested. In general, any spec over 18months old is probably worthless. === beuno is now known as beuno-lunch [16:06] sinzui: ok. no problem. I just wanted to know the plans for that blueprint [16:07] sinzui: but I will talk with Danilo about it === matsubara-lunch is now known as matsubara [16:22] jml: I am wondering if there is a way of namespacing this recipe stuff other than using the owner in the URL [16:22] bigjools, interesting. [16:23] bigjools, any specific ideas? [16:23] jml: nothing specific yet, I only just started to think about it [16:24] maybe tie them to projects, I dunno [16:26] bigjools, projects would be interesting. I think that would replace the source package, rather than the owner though [16:27] or we could just invent a new namespace [16:28] sure :) [16:28] I think I originally suggested lp.net/+recipe/ :) [16:31] of course, recipes should probably have names [16:31] :) [16:31] and once they need names, you need namespaces. [16:31] jml, is there a spec for recipes? [16:31] IDs don't look so bad after all [16:31] maybe allow duplicate names and prefer the ID [16:31] mars, probably not in the way that you are hoping. [16:32] mars: see the long email thread in progress now... [16:32] jml, perhaps the 'plan for generic build system jobs' thread? [16:32] mars, I'm probably going to hammer some out rsn. [16:32] bigjools, what are your concerns with having owner, sourcepackage and recipe name in the URL? [16:33] ok, I'm just trying to get the elevator pitch for it :) I'll check the thread. [16:33] mars, elevator pitch == "easy-to-make daily builds of your favourite upstream projects" [16:34] ah, so user-generated source package flavours [16:34] :) [16:34] jml, cool, thanks [16:34] I'm not sure what "flavour" means there [16:34] but if it works for you... [16:35] mars, check out bzr-builder for some more info on what recipes are in themselves [16:35] flavour means a slight customization of an existing construct [16:35] yeah, that's about it. [16:36] so that's why you are wondering about namespacing then. Because they are user-generated, but are also not portable between packages? [16:36] mars, yes. [16:37] mars, also, one day, recipes will be able to refer to other recipes [16:38] (he says, confidence and ignorance walking hand-in-hand across his soul) [16:39] * mars hears stub's distant screams of "Feature creep!" [16:39] :) [16:40] jml, interesting idea though. [16:41] jml, bigjools, you might find inspiration in the webservice API address scheme. [16:42] branch links come to mind as a similar concept: something that bridges between two pillars [16:47] I wish that porn spammer did not appear in in our oops reports. It seems like each day he tries to find an XSS hole in each application [16:58] jml, why would you use /+recipe/1 instead of /recipes/1 ? The latter is a bit easier for users to understand in the address bar, assuming they can browse that URL. [16:59] mars, because people should be allowed to have a product named "recipes" [17:00] ah === beuno-lunch is now known as beuno [17:54] hi sinzui [17:54] Hi bac [17:54] sinzui: do you have time today for a rescheduled 1:1 call? [17:56] yes, from 2-5 today [17:59] sinzui: wow, i don't think i can talk for three hours [18:00] You can listen to my ever growing fears about the software center [18:02] sinzui: how about a 2pm call? [18:03] timeboxed to 35 minutes [18:03] okay [18:47] night all! [19:10] good morning [19:11] i guess it's good to get the debate going and get the ideas out of my head, but i think al-maisan's mail is inaccurate in every point it makes :( [19:42] mwhudson: ! [20:14] jml: did you want to have a call? === salgado is now known as salgado-afk === matsubara is now known as matsubara-afk [20:23] mwhudson, abentley, stand up? I believe thumper is unavailable today. [20:24] rockstar: sure [20:24] abentley, you should host. I've been jumping between ISPs today. [20:24] rockstar: sure [20:25] rockstar: Okay, just waiting till you show as online. [20:26] abentley, one sec... [20:26] rockstar: can you hear us? [20:26] mwhudson, yes, through the speakers though, not through my headset. [20:26] abentley, try now. [20:32] rockstar: you went quiet [20:32] rockstar: yay skype [20:33] abentley, mwhudson, yes, I noticed after I finished my whore schpiel. [20:34] rockstar: We don't need to know about your whores. [20:34] s/whore/whole/ [20:34] abentley, call again? [20:35] twinkle? [20:36] mwhudson, my connection may just be terrible today. We have some sort of crew at the end of the street working on something, possibly damaged by the snow. [20:36] rockstar: was there more you wanted to add? [20:36] abentley, if there is, I'll reply to the notes. === jamalta_ is now known as jamalta [20:45] abentley: :) [20:45] oh, it's xx:45 [20:46] mwhudson: Skype ftw/wtf (your choice) [20:48] mwhudson: That issue stopped affecting my irc connection, so I thought it was gone :-( [20:54] abentley: :( [20:55] abentley: i guess we were more or less done [20:55] abentley: i was just trying to ask, is there stuff about dependent jobs in the database already? [20:55] mwhudson: Yeah, but I'd like to discuss testing my sftp change in a minute [20:55] mwhudson: No, job dependencies were vetoed. [20:56] i hadn't reallly thought about things in that light until now so I haven't looked [20:56] abentley: woo [20:59] Shall I file a bug that you can't request a vcs-import into a package branch, or is that deliberate? [21:01] https://dev.launchpad.net/Maintenance has not been updated for this planned maintenance [21:01] Perhaps it should be obsoleted in favour of one of the other communication methods [21:02] maxb: it's certainly not deliberate, there might already be a bug though [21:02] maxb: please have a look/file one [21:07] It's sort of related to bug 402915 [21:07] Bug #402915: Can no longer move a branch to another project [21:07] yes [21:09] mwhudson: chat? [21:09] abentley: ok [21:10] maxb: i'm surprised more people haven't given us grief over this bug [21:14] maxb: I think it was intentionally left unimplemented as a first cut [21:14] Ah, bug 369761, there's my bug. [21:14] Bug #369761: New code import page will not work for package branches

[21:14] mwhudson: http://paste.ubuntu.com/333394/ [21:18] sinzui: Around? I'd argue bug 400643 is not fixed - it now shows "Sid (98)" on edge - how is the number 98 meaningful to Debian sid? === mbarnett changed the topic of #launchpad-dev to: **Launchpad will be down/in read-only from 22:00 UTC until 23:30 UTC for a code update** This is Launchpad Development Channel | Week 4 of 3.1.11 | PQM is closed; for RC only. salgado is this cycle's RM | 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 lo [21:23] maxb: 98 is meaningful to some people the way that 9.04 is We decided that we cannot make the name-people and the number-people happy. We compromised and made the presentation of name (version) consistent though out launchpad [21:24] sinzui: No. 98 is picked out of nowhere to make Launchpad happy. [21:24] Same with experimental's 99. [21:25] sid/experimental do not have version numbers at all. [21:25] But Launchpad doesn't cope with that. [21:25] wgrant: I do not want to fight with people. Launchpad cannot give each distribution separate version/name rules [21:26] Yet clearly displaying made-up synthetic numbers in the UI is a bug [21:26] sinzui: The correct solution is to make the version optional. [21:26] If a distroseries should not have a version, then that is a separate bug. I do not see that bug reported [21:26] * wgrant checks how hard that is. [21:26] Hmm, sorting. /me gives up already. [21:27] I think soyuz really wants the distroseries to be a number [21:27] * maxb imagines a boolean "version is irrelevant" flag [21:27] maxb: I considered that. [21:27] that, or you need a separate numeric sortkey [21:28] I think letting version be None would suffice [21:28] We handle milestones with None for code_name. It would be the same... [21:28] sinzui: But there is an ordering relationship between squeeze, sid, and experimental which needs to be modelled [21:29] except that I do not know why a productseries and distroseries treat name and version differently [21:41] meep [21:41] I don't recall why either [21:41] where is kiko when you need him, or stub. [21:53] thumper: ping === mbarnett changed the topic of #launchpad-dev to: **Launchpad will be down/in read-only from 22:30 UTC until 23:30 UTC for a code update** This is Launchpad Development Channel | Week 4 of 3.1.11 | PQM is closed; for RC only. salgado is this cycle's RM | 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 lo [22:05] I see a mixed security warning on https://code.edge.launchpad.net/~jelmer/subversion/trunk. I do not see what is causing it [22:05] sinzui: A colleague noticed the same thing last week. We could not work it out. [22:06] it's not something loaded indirectly from JS? [22:07] I recall IE did stupid things like pull in the http feed (that is a meta) and declare that the page was unsecure [22:07] ajmitch: I suspect so [22:07] or firefox could be looking at the atom feed as well [22:08] sinzui: Other LP pages with HTTP feeds are fine, so that's not it. [22:08] firebug will probably tell you what it's fetching [22:08] Hm, some YUI stuff is HTTP. [22:09] http://yui.yahooapis.com/combo?3.0.0pr2/build/widget/assets/skins/sam/widget.css&3.0.0pr2/build/widget/assets/skins/sam/widget-stack.css&3.0.0pr2/build/overlay/assets/skins/sam/overlay.css& is pulled in by JS somehow. [22:09] Isn't all the YUI stuff meant to be embedded in one file? [22:09] that's the only http-only request I can see [22:19] I think code.branchstatus in the YUI.use() instruction is at fault. code.branchstatus is not listed in base-layout-macros.pt's load-javascript. Since the script is not compiled, YUI makes a call to retrieve it, and I know it always uses http. [22:22] sinzui: A team merge will transfer everything across, won't it? Subscriptions, memberships, etc.? [22:23] yes, if it does not fail [22:23] well [22:23] I have learned that I can keep merging until it succeeds. With barry, that takes 22 merges [22:24] sinzui: that's better than all the karma points in the world === mbarnett changed the topic of #launchpad-dev to: **Launchpad will be down/in read-only from 23:00 UTC until 23:45 UTC for a code update ** This is Launchpad Development Channel | Week 4 of 3.1.11 | PQM is closed; for RC only. salgado is this cycle's RM | 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 [23:12] sinzui: hooray for finding that js problem on the code page [23:12] mwhudson: I have not founded it [23:12] sinzui: oh [23:13] mwhudson: I added the suspect to the build script, but nothing was fixed [23:13] :( [23:13] :( [23:13] it's sooo hard to figure out how this is happening :( [23:13] can you get yui to drop into a debugger when it makes the request somehow? [23:14] indeed. I was going to remove pieces of the script to find the cause [23:14] mwhudson: I do not think so. [23:15] The yui loader was the cause of problems in the past. We fixed it by ensuring that all the scripts were loaded by the page or in our compressed script [23:19] \o/ [23:19] mwhudson: I enabled the net tab in firefox and loaded a branch page [23:20] sinzui: did you learn anything interesting? [23:20] mwhudson: when I mouseover the get combo line I see the browser got assets from Yahoo [23:21] sinzui: yes, that's what wgrant said earlier isn't it? [23:22] We seem to be missing widget.css, widget-stack.css, and overlay.css === mbarnett changed the topic of #launchpad-dev to: **Launchpad will be down/in read-only from 0:00 UTC until 01:30 UTC for a code update ** This is Launchpad Development Channel | Week 4 of 3.1.11 | PQM is closed; for RC only. salgado is this cycle's RM | 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 [23:25] mwhudson: Since we upgraded YUI on staging and I do not see the same error. I am not going to worry about it anymore [23:26] sinzui: oh that's good [23:26] yay for problems that evaporate [23:26] * wgrant wonders how everyone is going to know that the downtime is now two hours later than planned. [23:27] The downtime window is about to expire, and nobody has told identi.ca, or the appservers. [23:41] can someone tell me what I should read in order to understand the "request/lp:person/fmt:link:mainsite" part from [23:42] jml: where is that bug about testresources in lp [23:43] adiroiban: It's in canonical.launchpad.webapp.tales [23:43] adiroiban: lp:person would call RequestAPI.person, and fmt:link will call ObjectFormatterAPI.link [23:44] (although the second one has derivatives -- in particular you might want PersonFormatterAPI [23:45] wgrant: thanks. I'll look into it