=== kiko is now known as kiko-zzz === Ursinha-dinner is now known as Ursinha [02:55] hi [02:55] hello [02:55] i need to take over maintainership of a project whose owner is sick/uncontactable [02:55] is there a way to do this? [02:55] hi jjesse [02:55] i guess open a ticket? === barry-away is now known as barry [03:01] poolie: yes [03:01] hm [03:02] "help improve launchpad" is not so great [03:02] s//such a great label/ [03:02] what if i want help myself? [03:02] :) [03:02] poolie: what project [03:03] bzr-usertest [03:03] you need to take it over? its in the bazaar group already, you should be able to do most things to it as-is [03:04] hm [03:05] it is in the bazaar project group [03:05] that does not seem to give me any access though [03:05] like to change bug priority [03:05] hmm, [03:05] no bug supervisor set [03:05] I thought that that inherited [03:05] I've set the bug supervisor to bzr [03:06] ! [03:06] i will later want to change the trunk too [03:07] did you use superpowers to change that? [03:08] yes [03:09] out of curiousity through the db or the ui? [03:09] ui [03:10] doing stuff in the db -> not recommended [03:10] if you do need to take over the project, I can do that, but prefer not to - its better if you can get in contact with the current owner [03:10] I figure the bug supervisor thing isn't contentious [03:11] isn't likely to be, that is [03:46] Anyone know how to prevent a PPA build on lpia when the package architecture only specifies i386 amd64 ? [03:48] IntuitiveNipple: Why would you want to specifically block a build on lpia? [03:48] (and yes, it fails if the architecture only specifies i386 and amd64) [03:49] persia: kvm and libsmbios always fail on lpia but they're only for i386 amd64 ia64 [03:49] Right. Have you tried adding lpia to the list of architectures? [03:50] (admittedly most lpia HW doesn't support VMX, but I have seen at least one chip that did) [03:50] Hmmm... no... the idea is not to build for lpia at all :) [03:50] Right, that's the idea I don't understand. Why shouldn't libsmbios work on lpia? [03:50] * persia knows about kvm, and suspects that while it's mostly useless on most real lpia HW, it should build and install) [03:51] well, the original authors have specified just the three architectures I mentioned... I didn't want to mess with it :) [03:52] I bet it works. The differences between lpia and i386 are less than the differences between i386 and amd64. [03:52] I guess I could *try* it ... at least those horrible red crosses will disappear :) [03:53] Well yes... I just didn't want to alter the packages more than necessary. I'm mainly building up-to-the-minute package for hardy from upstream or intrepid. [03:53] IntuitiveNipple: Ah. Nice to fix that in intrepid then. I don't know what depends upon those, but they really ought to work for lpia. [03:54] I'll upload another version with lpia added and see what happens [03:55] lpia is less an architecture than a profile for i386 [03:56] There was a nasty bug in kvm-72 that I blamed on virt-manager :) It caused all grub installer actions to fail, and stopped XP rebooting [03:56] So I've put the fix up and wanted the builds to look all good with green ticks :p [04:01] The lpia build failure for kvm-72 is weird, anyone seen this before... [04:01] mv: cannot move `/build/buildd/kvm-72+dfsg/debian/tmp/usr/bin/' to a subdirectory of itself, `/build/buildd/kvm-72+dfsg/debian/tmp/usr/bin/kvm' [04:04] Looks like some serious work needed to make lpia builds of kvm work... the embedded qemu source creates build architecture-specific directories and code based on shell vars... in the case of lpia no shell var exists therefore the 'mv' source is the base-path with no extension [04:05] Strange. I just built qemu for lpia yesterday. [04:06] https://launchpad.net/ubuntu/intrepid/+source/qemu/0.9.1-5ubuntu2 [04:06] It looks to be because it doesn't create a qemu-system-lpia binary [04:07] * persia suspects another forum is more appropriate: perhaps #ubuntu-virt ? [04:07] indeed :p [07:08] <\sh> anyone who is able to enlighten me of the new license of the new LP logo? [07:14] \sh: it is specified in https://help.launchpad.net/Legal [07:14] \sh: which is linked from the LP footer [07:14] <\sh> kiko-zzz: when you are awake...I'm at my desk (somehow)from 06:00 UTC to 15:00 UTC (+/- one hour) [07:14] \sh: what do you want to know? There was an email to the launchpad-users list saying it that is licensed as CC-BY-ND, IIRC. [07:14] As thumper says, the legal page should have all the info. [07:15] <\sh> spiv: yes..that's why...the leonov logo was created with the new logo before it was licensed with CC-BY-ND...I need to know if we need a new one, or at least the info where can I ask for an exception so that leonov can use the base [07:16] \sh: I'd talk to kiko [07:16] \sh: (although if there is a problem, then it was already a problem when the logo wasn't licensed at all...) [07:17] <\sh> spiv: well, I want to be sure, that everything is correct...not to run into legal problems :) I'll try to catch kiko when he's awake ... [07:18] <\sh> now for some daily morning meetings and some coffee...laters [09:35] bigjools: I take it you found a reasonably easy fix for #159304, then? [09:36] wgrant: not easy but done nonethless [11:08] hi, does anyone know how I can delete milestones from a branch? [11:11] viciouslime, do you mean from a series? you need to request that through answers.launchpad.net/launchpad [11:12] kiko-zzz: thanks :) I'll do that [11:17] you're welcome === RainCT_ is now known as RainCT [11:31] bigjools: How is "Maintained" packages calculated on dogfood? The number seems high. [11:33] persia: it's where the user appears in the Mainainter tag on the source package [11:35] bigjools: Interesting. One of your examples seems to do things differently than many of us. [11:35] Also, since we're tracking uploads (which I like), maybe s/packages/uploads/ in the explanatory text? [11:35] persia: how do you mean? [11:36] bigjools: Most of use will use Ubuntu Core Developers or Ubuntu MOTU Developers for ubuntu-specific uploads. Your first example doesn't do that. [11:37] that's how it's always been on that page (even before it was broken) [11:37] So PPA packages aren't uploaded nor maintained? [11:38] in Ubuntu, no [11:38] Right, but it doesn't exactly say that. [11:38] But looking much better in general. [11:39] great [11:39] this is pretty much what I will release for now, barring any serious issues [11:40] I miss pre-Dapper package history, but that's not typically relevant for any of my non-vanity use cases. [11:40] if there are any more things you would like to see differently, please file bugs and we can do that separately [11:40] So it still doesn't show superseded SPRs? [11:40] Right. Nvaigation needs work, but mpt was already on about that. [11:40] wgrant: It shows Edgy, but not Breezy (at least for ~persia) [11:40] 'Displaying first 30 packages out of 726 total [11:41] That sounds a little strange. [11:41] Is there no text in a similar position elsewhere in Launchpad? [11:41] Because of the work "packages" or because of something else? [11:41] I'm not sure. [11:41] 'out of', maybe. [11:41] the lists only show items that are actually published [11:42] Right, that's probably a bad idea. [11:42] it's something that we could add extra filtering for on the batched listings [11:43] Why make the distinction about published/unpublished? [11:43] I am just wary of changing something that's clearly been that way for a long time [11:43] I don't think anybody actually likes it that way. [11:44] bigjools: Is it harder to get the data for the now obsolete releases? [11:45] no, the distroseries state does not enter the (current) query, it's purely done on whether the package is published [11:45] Doesn't that implicitly unpublish the SPPs, though? [11:46] no, the same package could be published in another series [11:46] if it's not, then the package is obsoleted, yes [11:46] Well, once all SPPs for a SPR are unpublished, the SPR will vanish from that page. That's what I meant. [11:46] correct [11:47] so would it be more useful to show *all* packages, regardless of state? [11:47] I think so. [11:47] Now it's batched there shouldn't be a problem. [11:47] I can tweak it on dogfood right now, just a sec [11:47] persia: What do you think? [11:47] let's see what happens [11:48] I'd be personally happy with *all* packages, regardless of state, but as I said, it's only for vanity use cases. [11:49] At the moment, if somebody only cares about N packages, but cares about them a lot, they only get N uploads shown. [11:50] ok, it's showing a lot more packages now [11:51] Indeed it is. [11:51] I must have uploaded an awful lot in Edgy. [11:52] Or multiple times in one series, I guess. [11:52] I had Edgy showing in my results before [11:52] Maintained packages is still broken, but that might make more sense like that. [11:52] persia: Probably because it has SPPs in other distroseries. [11:53] persia: It will show the series in which it was originally uploaded, not where it is currently published. [11:53] wgrant: Ah, quite possibly: it may have shipped also in Feisty. [11:56] Hmm. I'm still only seeing one entry per source per release. Am I missing something? [11:56] Rinchen: around ? [11:56] Ooh. [11:56] You import Lenny on dogfood? [11:57] I was looking for sid instead. [11:57] Hmm. Finally found 1 Breezy package. [11:57] persia: Must be published in Dapper. [11:58] I'm also certain that the numbers are off: e.g. dholbach had > 140 packages uploaded in Feisty, but it only shows 155 across all releases for him. [11:59] persia: 'Displaying first 30 packages out of 779 total [11:59] That's for dholbach. [12:01] Is it deliberate that I'm not shown as both uploading and maintaining a package? [12:02] wgrant: Hmm. I wonder how I got my result, but refreshing shows me yours. [12:03] Interesting indeed. [12:05] wgrant: yes that's deliberate [12:08] bigjools: Seems a but strange, but OK. [12:08] *bit [12:08] I don't know why that's done like that, I just looked at the code [12:09] Heh. [12:12] wgrant: ah it might be to filter syncs [12:14] Manual syncs should show up as an upload, and autosyncs are owned by katie. What do you mean? [12:14] where the maintainer is not the uploader [12:15] and we're interested in who uploaded it [12:15] Right. [12:16] But you filter things that I have uploaded but where I am the maintainer too. [12:16] really? got an example? [12:18] https://dogfood.launchpad.net/~wgrant/+uploaded-packages lacks soundconverter 1.2.0-0ubuntu1 [12:18] https://dogfood.launchpad.net/~wgrant/+maintained-packages has it. [12:18] And I'm in the Changed-By. [12:22] There are also inconsistencies in the text when somebody has uploaded nothing. [12:22] For the package pages, the username is used. +related-projects and everywhere else in the UI uses display name. [12:23] 'has no maintained packages' -> 'maintains no packages' [12:23] wgrant: right that's deliberate - the uploaded packages section only contains those where you're *only* the uploader [12:23] 'has no uploaded packages' -> 'has uploaded no packages' [12:23] bigjools: Seems a bit odd... [12:23] It would make sense in Debian. [12:24] maybe it should be renamed to "Sponsored uploads" [12:24] Except that's not it at all. [12:25] That's another section that we need, though. [12:25] Well, s/we need/would be nice/ [12:25] The intersection between that and anything else on the page would be {}, though. [12:26] Not necessarily: one could sponsor an upload to a package one maintains. [12:26] persia: GOod point. [12:27] Note that for the hypothetical "Sponsored Uploads", we're not interested in cases where Changed-By: and .changes signer are the same. [12:28] Actually, is there already a bug for "Sponsored Uploads"? I've a use case that might make it interesting (although not essential). [12:31] wgrant: thanks for pointing out those other inconsistencies [12:32] persia: I think there might be. [12:33] * persia will search later. [12:33] But it might be in a comment in one of the other bugs. [12:33] bigjools: It's one of the things I do best. [12:33] wgrant: I'll bear that in mind :) [12:34] Is there a timeframe for Soyuz to grow full rebuild support? [12:35] wgrant: define "full" [12:35] we do plan on doing rebuilds this year [12:35] Able to rebuild universe. [12:35] but not the "scorched earth" type [12:35] And give us results. [12:35] bigjools: Does that mean complete archive rebuild on import from the previous release series, or just button-push rebuild? [12:36] bigjools: Right, we've never expected more than that, I don't think. [12:36] the latter for now [12:36] but [12:37] ... [12:37] we're working on a spec that introduces derivative archive support so you can clone an archive (in part or full) and rebuild a selection of packages [12:37] Isn't that what a rebuild archive basically is, anyway? [12:37] yes - we've just made it generic [12:37] so the same code can be used for other purposes [12:37] Good to know. [12:37] bigjools: Any chance that it could be organised in such a way to allow something like a "test-rebuild-everything" feature? [12:38] Easy enough to do with appropriate ogreing. [12:38] My fear is that it would be abused to do archive-test-rebuilds on Ubuntu, and presumably you'd only want that once in a while. [12:39] Having an official "test-rebuilds" feature probably avoids that, although it may not be signfiicantly different in implementation. [12:39] persia: I guess if all packages are selected we'll be able to rebuild everything, but obviously there's issues around bootstrapping [12:40] bigjools: Understood. I'm just thinking that you'll want to have an official "Ubuntu rebuild test" section so you don't end up with twelve. [12:41] yeah this sort of thing is under discussion. The spec is here although it's a bit bare to non-Canonical people: https://blueprints.edge.launchpad.net/soyuz/+spec/archive-derivatives [12:42] * bigjools -> food [12:45] bigjools: I think we're used to that by now. [12:48] bigjools: Just out of curiosity, why would you want to disallow direct uploads? [12:48] I know at least the Ubuntu Desktop Japanese Remix and BlankOn would like to support that, rather than uploading to a PPA and copying the packages. [12:51] (or running DAK externally as both do now) [12:52] persia, direct meaning binary? [12:54] kiko-zzz: No, that would be bad. Source. [12:54] <\sh> Ah..kiko is awake ;) [12:54] From the blueprints page "The copy archive will never have direct uploads but will instead have packages copied to it from another archive (its parent)." [12:55] ah, ok [12:55] I just don't understand why if archives are generalised, it wouldn't be appropriate to allow uploads there (presumably with configurable permissions, etc.) [12:56] Also, for the rebuild-test archive-consistency use-case, would a new copy of the same source trigger a rebuild? [12:57] Alternately, if these copies include binary also, what happens when they are build with e.g. different C++ ABIs, and the resulting packages don't work together (because of, say, an experimental g++ in one of the PPAs from which it is sourced) [13:00] The world burns. [13:03] persia, I think there's some desire to maintain a level of simplicity and there's also a lot of corner-cases we avoid by not allowing direct uploads, but I'll talk it over a bit with julian [13:05] basically correct [13:06] bigjools, right. but how do you address persia's use case (or his concern)? :) [13:06] Understood, and I can imagine some of the corner cases. Is it at least a source-only copy? [13:06] * bigjools reads more scrollback [13:07] copies are only ever source only [13:08] Good. [13:08] But hmm. [13:08] Good. And does a fresh copy of the same source trigger a new rebuild? [13:08] Post-initialisation, you mean? [13:08] wgrant: yes [13:08] Copying only sources at initialisation seems strange unless you are rebuilding. [13:09] right, we intend that sources are copied and then you hit the Big Red Button [13:09] Right. [13:09] (it would be a bit easier if these specs were public, of course) === RainCT is now known as RainCT_ [13:09] persia: it could do, we've not discussed that in complete depth. The original intention is to make copy archives cheap as chips so you can just make a new one [13:10] There's enough information on the blueprints page to point at things, but yes, it would be. [13:10] More interestingly, if one subscribes to a spec early enough, one can get all the wiki edits by email, even though one cannot actually see the spec. [13:10] I can't imagine that creating a derivative archive could be cheap unless you publish in the same place. Which is bad unless it's a snapshot archive. [13:11] wgrant: who said they need to be published? :) [13:11] bigjools: Why do I want an unpublished non-rebuild archive? [13:12] bigjools: These aren't being published? [13:12] there will be an option to publish them or not [13:12] (except to confuse people who are looking at dogfood and see that lenny says it has no packages. That was confusing)( [13:12] dogfood is, erm, in a state of flux shall we say [13:13] I note that there is a nice unpublished Lenny copy, though. [13:13] yes, we're doing that to make syncs easier [13:13] Yep. [13:13] I thought gina might be dead for a while once s-i-s came along. [13:14] gina just won't die, I think kiko put some immortality code in there [13:15] Heh. [13:15] we don't currently use her for anything btw, but she's going to get a new frock for Debian importing [13:16] I thought she was used for importing from security dak. [13:17] wgrant: no, security uploads are re-uploaded to lp after going through dak (sic) [13:18] Ah. Hm. === thunderstruck is now known as gnomefreak [13:57] re: sponsoring earlier: bug #155956 (for any wishing to subscribe) [13:58] Launchpad bug 155956 in soyuz "+me/+packages should present different sections for sponsored uploads" [Low,Confirmed] https://launchpad.net/bugs/155956 === apachelogger_ is now known as apachelogger [15:41] hello [15:42] the Ubuntu wiki is broken for me [15:42] it seems to be linked to the kubuntu2 theme [15:42] http://pastebin.ca/1168083 [15:42] thi sis the trac [15:42] trace [15:49] * emgent need one launchpad devel [16:24] emgent: still need one? === ursula_ is now known as Ursinha-lunch [16:25] intellectronica: sure. [16:25] emgent: at your service, sir [16:25] intellectronica: nice :) [16:25] i need to know all people with Africa/Kigali timezone [16:25] (in launchpad) [16:25] I need it for Human project in Rwanda [16:26] intellectronica: can you grep this? [16:27] now some italian ubuntu people are in rwanda for tech ubuntu stuff (openoffice, firefox etc.) [16:27] but we should know if there is some ubuntu people in kigali for talk about possible collaboration [16:27] intellectronica: can you help me ? [16:28] emgent: i'm not sure what's our policy on doing something like that. the information is there (and you can google for it) but given that we don't provide a direct search function, i'm not sure it would be ok for me to run a search like that for a user [16:28] intellectronica: .... [16:29] emgent: i'll ask around and see what people think. in the meanwhile, would you mind filing a question? [16:29] i need to know only pubblic information.. [16:29] timezone is pubblic in launchpad, but i cant check this value in launchpad search filter. [16:30] emgent: yeah, that's my point. the information is there, so you're welcome to search for it, but since there isn't a way to search for it directly, i'm not sure if it's ok to run such a search for you [16:30] emgent: however, this discussion is quite theoretical, as https://edge.launchpad.net/+search?field.text=Africa%2FKigali should give you what you want [16:31] uhm i go to check [16:31] thanks [16:32] emgent: looks like it's only two users, and one of them writes in his profile that he's studying in england, so maybe the tz information for him is wrong :) [16:32] yeah saw that :( [16:33] ok, thanks === eMxyzptlk is now known as eMxyzptlk[away] === kiko-zzz is now known as kiko-phone [16:53] any lp admins around? [16:56] rosetta question... if I've added more strings to my .pot file, do I just upload the .pot and launchpad iwll run the msgmerge on the .po files it has? [16:58] I'm getting OOPS trying to moderate the ubuntu-bugcontrol mailing list [17:09] barry, ^^ [17:09] bdmurray, do you have an oops number for Ursinha-lunch to look up? [17:09] bdmurray: that's not good [17:10] OOPS-955G2074 [17:10] https://devpad.canonical.com/~jamesh/oops.cgi/955G2074 [17:11] * barry waits 10 minutes [17:11] I just generated that one though [17:12] TheEric, there always are, it's best to just ask your question and get the cool satisfaction of having it fixed! === salgado is now known as salgado-lunch [17:28] Rinchen, bdmurray wow. i've got the oops now. i'll submit a bug report on that [17:33] barry, is it a wow oops then? :) [17:34] kiko-phone: wow in the sense that it's triggering an assertion, so clearly unexpected [17:34] * barry guesses all bugs are unexpected tho ;) [17:34] heh === kiko-phone is now known as kiko-fud === barry is now known as barry-away === RainCT_ is now known as RainCT === fta_ is now known as fta === salgado-lunch is now known as salgado [18:29] barry-away, thanks for reporting that bug === barry-away is now known as barry === eMxyzptlk[away] is now known as eMxyzptlk === kiko-fud is now known as kiko-afk [19:54] uhm, how odd, i just got mails from bug 155947 and bug 51315 with the subject "Georgia: Ceasefire and Withdrawal Now" [19:54] Launchpad bug 155947 in libnss-ldap "libnss-ldap: calls to initgroups() causes boot to hang when using 'bind_policy hard'" [Undecided,Fix released] https://launchpad.net/bugs/155947 [19:54] Launchpad bug 51315 in libnss-ldap "udevd: nss_ldap: failed to bind to LDAP server" [Undecided,In progress] https://launchpad.net/bugs/51315 [20:53] anybody on translations? [20:53] I just tried to upload a .pot and now status shows "failed" [20:54] can't entirely say I understand how I'm supposed to update my .pot === kiko-afk is now known as kiko [21:26] Hi, I deleted a package from PPA and now I'm uploading a fixed version but getting that it was already accepted and I cannot upload it again although the delete was executed hours ago. [21:27] <\sh> Syntux: first, why are you deleting it, when you uploading a new one anyways... [21:28] \sh, I forgot to add 0ubuntu to the version so I had to reupload and when I tried that it said that I cannot upload older version [21:29] <\sh> kiko: is the archive package delete cron job still running so few times ? I thought you (the admins) thought about increasing the interval [21:30] \sh, it has already been deleted, in my PPA I have 0 packages, 0 MB binary and source. [21:32] \sh, I actually don't know the answer to that question! I need to check with the soyuz guys [21:34] <\sh> Syntux: ask the soyuz people :) [21:34] Who are they? [21:34] Are they from the dark side? [21:34] * Syntux hides === ursula_ is now known as Ursinha [21:57] hey folks, is it possible to rename a project completely? [22:05] sure it is [22:06] what project, Romario99? [22:10] gscrot [22:10] it is not needed immediately [22:11] Romario99, and what would it be renamed to, if it was? [22:11] gnapshot [22:11] what a better name indeed [22:11] should I do it now? [22:11] hehe, i know [22:11] but its a frontend for scrot [22:12] I know, but scrot is a really bad name! [22:12] yes, am refactoring the whole app to be a standalone application [22:12] and gscrot is obsolete ;-) [22:13] i will let you know when it should be done [22:13] thank you, kiko [22:13] Romario99, cool, no worries [22:26] Syntux: You can't ever upload an older version. [22:27] wgrant, yeah that should make sense but now I have deleted the package and want to upload it to fix that mistake [22:27] That is beside the point. [22:27] Uploading older versions can confuse things horribly. [22:27] It simply should not be done. [22:28] (and Soyuz enforces this) [22:29] wgrant, what do you recommend doing now? changing the package name? [22:29] :-) [22:29] increasing the version? [22:30] The version. [22:30] You would never change the name for something small like this... [22:30] As that does the same thing as reducing the version. [22:30] No upgrades. [22:30] And it's generally bad. [22:32] Ok [22:32] I admit it was a horrible mistake. [22:32] :-) [22:32] will increase the version now, thankfully I marked it as experimental [22:35] barry: hey, are mailman 'topics' case insensitive? [22:35] what is Soyuz ? [22:36] lifeless: the topic patterns are just regular expressions, and i believe they are case insensitive. [22:36] Soyuz is the part of Launchpad which manages the packages in distributions and PPAs. [22:36] we have a user saying they aren't working for her [22:36] this is the regex: [22:36] \[merge\] [22:36] \[codereview\] [22:36] note that that is two lines [22:36] does it need a | and branches ? [22:37] a mail got through with subject [MERGE] wordlist refactoring [22:38] so 'what are we doing wrong' [22:39] Syntux, the part of launchpad that knows how to manage packages. [22:39] oh I see. [22:39] Thank you [22:40] kiko: is it still the official name, or is it launchpad packages or something ? [22:40] lifeless, it's still the official name. [22:40] cool === Ursinha is now known as Ursinha-dinner === barry is now known as barry-away === salgado is now known as salgado-afk