=== cody-somerville_ is now known as somerville32 === kiko is now known as kiko-zzz [01:15] I have a question: how may I change the e-mail address asociated to a Launchpad account? I see that I can add another but not replace the existing one [01:15] pdenapo: Once you add another, you should be able to remove the old one. [01:16] doesn't seem to be the case... [01:16] there is no option for removing the old one... [01:16] You might have to set the new one as primary first. [01:19] Now, I've succeeded. Thanks [01:19] it seems that it was a matter of first replying an e-mail confirmation [01:32] cprov-out: squawk if you happen to be around, eh? [01:47] Yay, I like the new ordering of the buildd listing, though it seems the archs are ordered by some internal ID. [01:51] Fujitsu: yeah.. I was wondering what the sort key was. [01:51] since one would think that amd64 should land above ppc... or some scuh [01:52] * lamont debates whether to hug Hobbsee, or aim the water cannon [01:53] lamont: what have i done? [01:53] lamont: you could just be nice and kill me. [01:53] so make sure it's got something else apart from water [01:53] lamont: I presume it's their serial primary key, what with hppa then lpia being last. [01:55] Hobbsee: who says I need a reason to drench you? [01:55] * Fujitsu throws waterbombs at lamont. [01:55] lamont: ah right [01:56] * Fujitsu is glad to see that the assigning stuff should be fixed by next edge rollout. [01:58] * Fujitsu doesn't like kiko's latest email to launchpad-users... LP shouldn't be filing bugs upstream itself. That will infuriate them. Particularly Debian. [02:00] Fujitsu: if they were actually to do it sanely.... [02:00] Fujitsu: having said that, i'm thinking of the current mess with assignees and such. [02:01] We'll see how they've fixed that tomorrow. [02:14] Fujitsu: it depends on the package, actually. [02:14] for example, I would have no issue with LP filing bugs against my debian packages [02:14] lamont: Of course. [02:14] with luck, hppa should be making better time now [02:15] What has changed? [02:15] 50% increase in the number of buildds building hardy [02:15] nice! [02:15] You got more hardware, or somehow got the security buildd to do it too? [02:15] * Fujitsu checks. [02:15] I, uh, talked the security buildd into sharing [02:16] and I owe elmo a patch. [02:16] Yay, castilla is back. [02:16] I suppose it doesn't get too much load. [02:16] With only two supported releases. [02:16] castilla has always been the security buildd, never been launchpad buildd until just a few min ago [02:16] ANd no livefses. [02:17] yeah - I need to fix livefs on both ia64 and hppa [02:17] ia64 has livefs, but the boot block believes that all 3 cds are alternate. [02:17] makes desktop/server kinda boring [02:17] and hppa needs some real love (options file, anyone? > 120 characters of commandline?) before it'll do non-alternate [02:19] I'm not sure there'd be much of a market for live hppa, really. [02:19] building: 13:23:40 7.97% [02:19] install : 00:00:00 0.00% [02:19] removing: 00:00:00 0.00% [02:19] idle : 148:21:10 88.30% [02:19] total : 168:00:02 [02:19] so, yeah. mostly kinda idle. [02:19] wow. 2 leap seconds last week. :-) [02:19] a base-livecd has some real potential. desktop live, not so much [02:20] that is, at some point, we start making server-live for all architectures, right? [02:20] * Fujitsu didn't know of any leap seconds last week, let alone a double one. [02:20] server-live sounds a while off. [02:20] cronjob ran 2 seconds later this week than last... :-) [02:20] Ah. [02:20] so there were 168:00:02 hours in the week... must be leap seconds... FTW!!! :) [02:21] Ohh, I see. Yes. [02:21] What was it doing the other 5%? [02:23] being lots. [02:23] lost, even [02:23] hrm.. it's been up 42 days, so it wasn't crashed. [02:31] Are those stats from buildd? [02:31] DAK buildd, yes [02:32] That's what I thought. [02:35] lamont: How much longer do you expect hppa to take to catch up? [02:36] that depends entirely on how frequently kde, gcc*, python, and glibc get uploaded. [02:36] kde builds are _SLOW_ [02:36] i'm sure we can help that along a bit [02:36] something to do with having huge monolithic packages [02:36] * Hobbsee uploads more of it, then [02:37] lamont: yeah, just the ones we *odnt* want to be uploaded repeatedly [02:37] I was horrified to see that they merged dolphin into kdebase recently. I *really* don't see the benefit. [02:37] Fujitsu: because then it all builds together. [02:38] apparnetly 4 isn't as bad as 3 [02:38] they merged it in for the same reason that they merged everything else in [02:38] lamont: And takes hours and hours and hours and all has to rebuilt every time there is a tiny bug in kdepim. [02:38] Or whatever else there are regularly tiny bugs in. [02:38] Fujitsu: sadly, such arguments have historically fallen on deaf ears. [02:39] * Hobbsee had taht argument with them too [02:39] Fujitsu: it's -base, mainly. [02:39] Hobbsee: Are you still on the right side? [02:39] depending on what you define the right side, yes [02:40] the stuff doesn't get test built, as it's so big, and doesn't get tested, so it needs multiple uploads [02:40] It has more than 35 binaries, damnit. [02:40] I'm not _saying_ that was a factor in ubuntu choosing to go with gnome... [02:40] Haha. [02:40] this was one of my reasons for stepping down from it [02:40] lamont: sure sure [02:41] Oops. [02:41] * Fujitsu notes that one should check which audio output one has selected when using PulseAudio. [02:46] mmm...spaghetti makes good breakfast [02:47] mmm [02:47] Breakfast at 2pm? Sounds rather odd. [02:47] (as does the spaghettiness) [02:48] lunch at 4.30 or so, dinner at 11. [02:48] now that sounds odd :) [02:48] It does. [02:51] if breakfast is at 2PM, then lunch is 8PM. [02:51] and dinner is after work, so after 11:30. :-) [02:51] and welcome to swing shift [02:51] hello Hobbsee [02:51] Evening, ajmitch. [02:52] evening? it's getting close, but not quite [02:52] hi ajmitch [02:52] * ajmitch is still at work, having fun with zope & plone [02:52] Zope 2, then? [02:52] lamont: lunch would be later, except htat i can't take a risk of being off the floor for that long, that late. [02:52] of course, with some zope 3 stuff mixed in [02:52] oh, right. [02:53] retail makes for strange schedules [02:53] ajmitch: Fun, fun. [02:53] * Fujitsu likes his position, with no customer interaction. [02:53] lucky you [02:53] although, if they've given me someoen with a clue tonight, i might be able to have a later dinner [02:53] Although there is a bit too much PHP interaction. [02:53] * ajmitch is in a small enough company that customer interaction is required [03:21] there is a part of me that would really love to be able to specify regexps that affected default build-priority for packages. [03:21] then I could just demote all of kde for the moment. :) [03:21] just on hppa, of coures. [03:21] * Hobbsee could just demote lamont [03:22] lamont: You should be able to mark them as failed and eventually give them back soon, no? [03:22] I can mark them as failed? [03:22] how:? [03:22] `soon' [03:22] oh, ok [03:22] Because dropping the priority only works for so long. [03:22] and I want one that I can feed a regexp to... :-) [03:23] nah - I don't mind them building once universe is done... :) [03:23] Convince one of the DBAs that you need access to the DB. That shouldn't be hard... much... no, of course not. [03:24] he was a canonical employee. he should be able to get it [03:24] no. [03:24] he doesn't _WANT_ it. [03:25] * Fujitsu turns lamont into a duck. [03:25] heh [03:26] * Hobbsee turns Fujitsu into a pigeon [03:26] * Fujitsu doesn't know of an pidgeon-emblemed teams. [03:27] Er, pigeon [03:27] then you don't get a team [03:27] * Fujitsu recalls the days when launchpad-beta-testers was the upside-down duck. [03:32] Hobbsee: I just want to let universe catch up.... I promise to build kde before they upload it more than 5 or 6 times [03:32] lamont: *g* [03:33] lamont: fair enoough. i doubt anyone on hppa actually uses kde anyway [03:33] * lamont figures a couple weeks tops, if kde/toolchain/python/glibc don't upload too often [03:50] New bug: #173812 in malone "Remerge the enable bug expiration per project branch changes" [High,Fix committed] https://launchpad.net/bugs/173812 === stu1 is now known as stub [05:23] dear launchpad. it'd be nice if the dep-wait release code knew about ogre-model. [05:23] (see hardy curl for an example. and tuxtype and... [05:23] ) [05:24] on the bright side, it tells us when queue-builder finishes... :-) [05:25] and gnome-games [06:06] thumper, jml, spiv, BjornT: reviewer meeting [06:06] * jml is here [06:06] hi [06:06] == Agenda == [06:06] * Roll call [06:06] * Next meeting [06:06] * Action items [06:06] * Queue status [06:07] * Graduations (and recruiting?) [06:07] * Mentoring update [06:07] * Review process changes [06:07] * On-call reviewer [06:07] * Cover letter [06:07] * Death to [trivial] [06:07] * Tool update [06:07] Is the same time next week okay for everyone? [06:08] Sure. [06:08] yeah [06:08] yes [06:08] did we have any action items from last week? [06:08] No. [06:08] * Queue status [06:09] the queue is pretty short, with only two branches past the due date [06:09] BjornT: how is your one going? [06:09] been quite busy, sorry, but i'll get it done today. [06:10] If you are too busy, you can reject it [06:11] yeah, i know [06:11] three of the remaining branches are stub's [06:11] I'm not sure what sort of response time we're expecting on those [06:12] I don't think we have any mentored reviewers at this meeting, so I'll skip that [06:12] * Review process changes * On-call reviewer [06:13] I've signed up to be on call on Fridays. [06:13] I haven't signed up for on call review yet [06:13] and I guess BjornT and thumper shouldn't [06:13] It went well, and it seems the system is helping. [06:13] it's been suggested that team leads don't [06:13] so I removed myself [06:13] however [06:13] Neither have I, but then I'm another special case... [06:14] I've found it helpful to me [06:14] for the asia-pacific timeslot, we don't have many reviewers [06:14] mwhudson will be moving into this slot [06:14] in jan [06:15] jamesh: but we have a high percentage of reviewers :) [06:15] I wonder if it'd be worth removing the "don't work on any of your own code" bit and roster people on more frequently? [06:15] jamesh: tbh, I think it would diminish the value of being on call [06:16] jml: I think it is reasonable to work on your own code as long as reviews take priority [06:16] if you are on call [06:16] especially for Monday morning (a slot now available) [06:16] thumper: I guess what I mean is, the sort of work I'll be doing while on call would be pretty light. [06:16] otherwise too many interruptions. [06:16] jml: right [06:16] The trick is making sure you don't get lost in your own code for 40 minutes while a dev is waiting for you to respond to a review request. [06:16] simple bug fix et al [06:17] thumper: I guess I could take on Monday morning -- it's a quiet time anyway. [06:17] jml: don't you have a slot? [06:17] jml: well, the value of an on-call reviewer is pretty low if there is no on-call reviewer [06:17] I do. [06:17] jamesh: good point. [06:18] and rostering one person on multiple days will reduce their effectiveness as a developer if they can't work on their own stuff [06:18] jamesh: right [06:18] there are 15 slots [06:18] how many available reviewers are there? [06:18] jamesh: It will reduce their effectiveness full stop. [06:18] I count 10 [06:19] so I'd suggest we try to get 5 more [06:20] thumper: one point is that we've got a fairly high reviewer:developer ratio for this time zone [06:20] so the fact that we can't fill all the slots indicates that an on-call reviewer probably won't be doing a full day's reviewing [06:20] (in the expected case) [06:21] for the asia-pac slots we only really have 3 people and 5 slots [06:21] (3 when mwhudson gets here) [06:21] and how many people would they be doing reviews for usually? [06:21] jamesh: Last Friday I did 5 reviews for 3 or 4 people. None of them asiapac [06:22] jml: I take it you got the americans at the end of their day [06:22] my guess is that there will be less interruptions for review than in other slots [06:22] thumper: pretty much [06:23] I suggest we force some europeans to shift to NZ/Aus [06:24] It's nicer here anyway [06:24] should we move on? [06:24] jamesh: yes [06:24] please [06:24] * Cover letter [06:24] thumper: relocate everyone to a pacific island :) [06:24] I think these have been quite useful [06:24] jamesh: definitely. [06:24] better than the 1-line descriptions we had previously [06:24] even if it's just 2 paragraphs on the wiki page rather than one. [06:25] +1 [06:25] * Death to [trivial] [06:25] +n [06:25] (where n > 1) [06:25] So SteveA wants [trivial] turned off [06:26] * thumper agrees [06:26] this pretty much requires attentive reviewers for small branches [06:26] * jml too [06:26] jamesh: right. on call should balance it out. [06:26] for trivial changes it should be easy for a reviewer to OK it [06:26] yes [06:26] jamesh: I'd suggest the normal - who can look at a trivial change [06:26] even if I'm not on call, I'm happy to look at those [06:27] yep. [06:27] if it is quick and easy to get a second set of eyes to look at small changes, then there is less excuse not to [06:28] right [06:28] * Tool update [06:28] Is anyone here actively involved in this? [06:28] or is it more of an AMEU item? [06:29] only indirectly. [06:29] move on :) [06:29] * Other business [06:29] pressups? [06:29] :) [06:30] thumper: slow but steady progress [06:30] jamesh: none from me. [06:30] okay. Meeting ends [06:30] jamesh: thanks! [06:30] thanks everyone. [08:18] Yay, bugwatch flood. [08:24] oh, the patch to make bug expiration optional per project landed yesterday [08:24] (will probably only really matter on the next release though) === \sh_away is now known as \sh [08:45] For my SSH keys, I'm asked to insert the contents of a id_dsa.pub or id_rsa.pub which is obviously from the example on a linux system. However I am on windows, do I need to install cygwin for this or do I have another option? [08:46] distatica: Which ssh client are you using? [08:47] I haven't installed one yet, I've used putty usually on windows, but I wasn't sure if I would need to install cygwin so I haven't bothered to download it. [08:47] Is there some way to close a project? [08:48] https://edge.launchpad.net/compizsettings/ is dead, I want to get rid of it [08:48] or at least make launchpad stop trying to file compiz bugs against it [08:48] Amaranth: https://answers.launchpad.net/launchpad [08:49] distatica: That really depends on what you want to do with your ssh key. putty supports key authentication. [08:49] ah, goody, maybe now launchpad still stop messing up when i try to upstream bugs [08:49] distatica: does this help? http://bazaar-vcs.org/Bzr_and_SSH [08:50] jamesh: reading now, bzr is the reason I need it.. [08:50] I guessed :) [08:50] the SSH keys in Launchpad aren't used for anything else at present [08:50] oh, hehe [08:51] yes, this will work fine, I'll follow the windows instructions and check back if I encounter problems, thank you. [08:51] alrighty then, question aksed: https://answers.edge.launchpad.net/launchpad/+question/19312 [08:52] Amaranth: in the mean time, you can disable filing of new bugs by saying that the project does not use Launchpad for bug tracking ... [08:53] awesome, now when i try to upstream compiz bugs it still says it's for the compiz-settings project but lets me put in a url [08:54] Amaranth: ah. The problem seems to be the invalid packaging links [08:55] Amaranth: "compizsettings" says it is packaged as "compiz" in Ubuntu [08:55] I take it that this is not the case [08:57] jamesh: nope [08:57] no? [08:58] Who is allowed to do that? [08:58] Amaranth: I don't think it is easy to do that through the UI at present [08:58] You can link things but you can't unlink them [08:58] But who is allowed to link them? [08:59] Amaranth: perhaps update the ticket to ask for the packaging links to be removed rather than deactivating the project [08:59] And why is a random project allowed to break package->upstream connections for ubuntu? [08:59] No, the project is dead anyway [09:00] So if I create a project, set it to use launchpad for bugs, and say it's packaged as gnome-screensaver in ubuntu then ubuntu people won't be able to link to gnome bugzilla bugs anymore? [09:00] Amaranth: No, you can say the bug belongs to another project. [09:01] I see no way to do that in the UI [09:01] Otherwise I wouldn't be here [09:01] Amaranth: When you say that it affects a project in the first place, there is an option to say that it's not the one linked to the package. [09:01] But you have to choose one that is in launchpad [09:01] Once the task is created, you're stuffed, due to a bug which means you can't reassign tasks for projects that use LP. [09:02] Right, but that doesn't mean it uses Launchpad for bug tracking. [09:02] jamesh, or anyone really. Ok, I got all the putty stuff, I generated private and public keys, I opened my private key here, with my password, it's running in pageant, I copied and pasted the contents from my public key file to the launchpad edit ssh keys page, and it says invalid key. I even tried removing all line breaks to make sure it was all one line. Any thoughts? [09:02] So I have to find a project that doesn't use launchpad for bugs and reassign to that just so I can then link to a completely unrelated bugzilla from another project? [09:02] I tried with and without the BEGIN SSH2 PUBLIC KEY part. [09:03] distatica: the format it expects is the one-line form that openssh uses [09:03] Amaranth: No, you should have a project representing the real upstream project. [09:03] jamesh: hmmmm.. that's what I tried to do.. [09:03] distatica: something like "ssh-rsa $lots-of-chars $comment" [09:03] Oh, right, this is "everything is in launchpad" [09:03] Amaranth: It's the only way to model it properly. [09:04] Or you could not allow projects to screw with distros [09:04] (you don't see me defending Launchpad often, but this is somewhere I will) [09:04] distatica: iirc Launchpad does not accept a key without a comment at the end. You can put whatever you want there though [09:04] The connection should be the other way, someone managing the package in Ubuntu should say "this package comes from this project" [09:05] You can do it from either end at the moment, and cannot currently delete links. I believe the eletion feature has been deferred a number of times, and is still coming RSN. [09:05] Amaranth: for a long time we've talked about automating the linking process [09:05] just haven't gotten round to implementing it [09:05] jamesh: hmm.. my comment comes before, maybe that's it. [09:06] nope [09:06] Amaranth: if we have the release tarball for the project and the pristine source tarball for the package we can infer a link if they are identical [09:06] jamesh: may I quick PM you? [09:06] distatica: sure [09:06] Fujitsu: I actually have written some stuff for the deletion feature last week. [09:07] ddaa: Ah, very good. [09:07] So it's finally happenign? [09:07] it needs a bit more polish to be release-worthy, and even then it will be a transitional hack [09:07] until we re-do the whole upstream association ui [09:07] but it will address the critical "cannot delete link" issue in the short term [09:07] This is probably the 100th time I've hit this [09:07] And why I don't upstream bugs much, which upstream doesn't like [09:08] we realise it's a nuisance [09:08] It is ridiculous that they weren't either restricted or deletable in the first place, but it is being resolved. [09:08] but this is the sort of thing that are tricky to get done, for organisational reasons [09:09] What reasons would those be? [09:09] it is not within the core responsibility for any team [09:09] or didn't use to be, until we invented foundations. :) [09:10] launchpad is all about linking things toghether, but that means that there's a lot of things which lies in the cracks between dimensions, to borrow from Dan Simmons. [09:10] ddaa: Isn't it Soyuz? [09:10] I would have thought it would come under that, really. [09:10] I thought it was malone [09:10] But I guess it is sort of more general and Foundations-ish. [09:10] other people would think it's Translations [09:11] Strictly speaking, it's part of "registry". [09:11] I guess that bits of Soyuz are Registryish... but the distinction looks very blurry. [09:12] Anyway, good to see it finally being sorted out! [09:12] Fujitsu: I think you starting to put your finger on why it's organisationally tricky :) [09:13] ddaa: I always thought some things were blurry, and wondered how they were handled. [09:13] "not as well as they should" [09:13] kiko-zzz: Thanks for the dgetting ability! I'll be watching for the edge rollout tomorrow. [09:14] Though we're getting better at it. [09:14] Wait, kiko-zzz, what are you doing up? [09:14] ddaa: Good to hear. [09:27] heh, could have fixed my problem anyway [09:27] found the place to change a link in the package side [09:27] and there is a compiz project in launchpad already [09:35] New bug: #173853 in launchpad "Validation problem while choosing package" [Undecided,New] https://launchpad.net/bugs/173853 [09:40] there is a server for sparc ppa ? [09:41] no, PPA only supports i386 and amd64 due to it's reliance on secure virtualization [09:42] elmo: thanks for the info :) [09:42] i've sent a package to ppa with architecture: sparc [09:42] and it's accepted :( [09:42] and published [09:42] Kmos: the Architecture in the .dsc is 'sparc' only? [09:56] elmo: yes [09:56] elmo: it's afbinit package [09:57] Kmos: what is the URL of your PPA? [09:57] Kmos: please file a bug on soyuz - it should reject such packages, I think [09:57] elmo: I think so [09:57] jamesh: 2 sec [09:57] jamesh: https://edge.launchpad.net/~gothicx/+archive [09:58] Kmos: I only see a source package published [09:59] i've 8 packages in ppa [09:59] I mean for the afbinit package [09:59] jamesh: but it shouldn't be rejected? [10:00] and even not published [10:00] well, there isn't anything inherently bad about publishing the source package [10:00] although rejecting it would stop you from expecting it to be built :) [10:00] exactly [10:00] i think it would be built, because it was accepted [10:01] jamesh: you report the bug, or I do it ? [10:02] Kmos: you probably have more information on what you did and what you expected to happen [10:02] jamesh: ok =) i'll do it [10:11] done [10:20] New bug: #173866 in soyuz "When specific arch is not available at PPA, it should reject" [Undecided,New] https://launchpad.net/bugs/173866 === bigjools-afk is now known as bigjools === cprov-out is now known as cprov [11:40] hiya Hobbsee [11:40] heya mpt! [11:43] hi BjornT [11:45] hi Hobbsee === kiko-zzz is now known as kiko [12:04] Fujitsu, I was up but then I went to sleep! this patch ate my dinner [12:05] kiko: A...ha. [12:06] mmm...hungry patches. [12:23] Is there a way to upload a file directly to "Download project files" on LP without using the web interface? [12:25] While I'm uploading my PPA via script, it would be great if I just updated the upstream tar.gz at the same time. I use the "Download project file" page for hosting. [12:47] frenchy, well, not right now, but it's very easy to script a web upload using something like zope's testbrowser [12:50] kiko: Thank you. I'll have a read about that. [12:50] frenchy, I actually have sort of an example script I can show you when you've done some reading of it [12:51] it won't work for you because it requires some horrible hacks in testbrowser itself to make it simpler (I think) [12:51] but the code is a good example [12:53] kiko: Sure, that sounds great. At the moment it seems like any solution is a good solution. === kiko is now known as kiko-afk === kiko-afk is now known as kiko [13:35] New bug: #173899 in malone "E-mail interface isn't advertised" [Undecided,New] https://launchpad.net/bugs/173899 [13:37] bug 173670 [13:37] Launchpad bug 173670 in rosetta "IndexError on pluralforms during import" [Critical,In progress] https://launchpad.net/bugs/173670 - Assigned to Jeroen T. Vermeulen (jtv) [13:46] hi there :) [13:47] i have some problems with my PPA (related to missing .mo files in generated .deb) :) who could help me please ? :) [13:47] my PPA is located at : https://edge.launchpad.net/~kervala/+archive [13:49] the .deb was build successfully but all .mo were removed :( [13:50] in the log, i can see .mo were put in wxmtpchat_0.12-1_lpia_translations.tar.gz [13:50] bu i can't find this file anywhere [13:55] kervala: Hi there, are you saying that when you build the deb yourself that the mo files are in there? [13:56] New bug: #173902 in soyuz "PPA "Activate" button is available even if I haven't accepted the terms of service" [Undecided,New] https://launchpad.net/bugs/173902 [13:56] frenchy: yes :) [13:57] all is correct when i build it locally [13:57] .mo files are installed in /usr/share/locale/... [13:57]