[00:00] any lp sys admins around? [00:01] dendrobates: heyo, what's up? [00:18] what is the command to add a tag to bzr branch? [00:22] ari-tczew: bzr tag TAG_NAME [00:22] spiv: thanks [00:23] ari-tczew: (you can optionally pass -r REVISION and -d BRANCH, etc, see "bzr help tag") [00:23] does bzr tag prevent branch to be overwrited by souyz ? [00:23] I don't know much about soyuz, but I doubt it. [00:24] spiv: so uploading patch by bzr makes no sense, if then will be overwrited by soyuz importer? [00:29] ari-tczew: It's not part of Soyuz that does that; it's the UDD importer. See https://wiki.ubuntu.com/DistributedDevelopment/Documentation/UploadingAPackage for instructions on uploading. [00:33] wgrant: so reading from above site, command 'bzr mark-uploaded [00:33] is enough instead adding a tags? [00:33] It adds the tags automatically. [00:34] ok I'll test [00:52] little problem here, I changed my id and now I'm getting "No such Launchpad account: " when trying to push [00:52] wgrant: what should do I first? dput or bzr pull? [00:53] tombaugh: bzr launchpad-login [00:54] bzr is kinda of a very launchpad-centric version control system, isn't it? [00:54] MTecknology: no [00:54] launchpad is a very bzr-centric website :-) [00:54] oh [00:54] so it's just the ubuntu packaged version that has the extras that make it feel like that? [00:54] ari-tczew: It doesn't really matter, just do them roughly at the same time [00:55] maxb: so I must be fast [00:56] MTecknology: the 'launchpad' plugin is in upstream bzr, (bzrlib/plugins/launchpad), along with a small handful of others. [00:56] MTecknology: no. there's no ubuntu customizations. bzr has some small amount of launchpad integration in the core bzr distribution. But, that's a pretty tiny amount of code compared to the whole thing [00:56] It's not a large plugin, but it is pretty useful for bzr's own devs :) [00:56] ok- that makes sense [00:56] I just randomly got curious fr4om the "bzr mark-uploaded" above [00:56] ari-tczew: well, it would be advisable to push within half an hour of uploading, I think, which isn't very demanding [00:57] (Also, having a few plugins in the core distribution provides some useful examples for would-be plugin authors) [00:57] MTecknology: 'bzr mark-uploaded' is provided by bzr-builddeb. It's Debian-packaging-related rather than Launchpad-related [00:57] MTecknology: mark-uploaded is from the bzr-builddeb plugin, that is a separate package. [00:57] oh [00:58] ("bzr help COMMAND" will tell you which plugin the command is from, FWIW) [00:59] I'll have to look around at the plugins instead of assuming it's all built into one thing [01:00] I knew there were plugins - I just never looked to see what parts are from plugins and what plugins they're from [01:01] There's also "bzr plugins" if you want a quick overview. [01:03] maxb: thanks, but now I'm getting "Permission denied (publickey)." I did create a new key and uploaded it [01:06] tombaugh: ok, first things, check your https://launchpad.net/~your-id/+sshkeys page. Is the public key corresponding to the private one your trying to log in with listed there? [01:07] maxb: yes, that's what I copied from ~/.ssh/id_rsa.pub [01:09] how can I build source package from bzr signed to someone else? [01:09] I want to sponsor one thing. [01:10] bzr bd -S <- what I misseD? [01:10] tombaugh: try "ssh -v -l bazaar.launchpad.net", and look at the output to see which key file(s) your ssh client is trying [01:14] tombaugh: thanks for that paste [01:14] "Agent admitted failure to sign using the key." is probably the issue here [01:15] The client is trying to use the right key, but it presumably has a passphrase. [01:16] ari-tczew: -k [01:16] StevenK: bzr: ERROR: no such option: -k [01:16] And you have an SSH "agent" running locally, possibly gnome-keyring, that is responsible for unlocking it, but I guess it doesn't have the passphrase. [01:17] ari-tczew: bzr bd --builder="debuild -kkeyid" [01:17] spiv: problem solved using ssh-add (via google) [01:17] tombaugh: cool. [01:17] I didn't enter a passphrase by the way [01:18] Ah well, so long as it's working for you it doesn't matter if I'm a bit wrong :) [01:19] maxb: so, debuild -S -sa -kMYMAIL would work same as bzr bd -S' ? [01:19] spiv: thanks anyway, you pushed me in the right direction [01:20] tombaugh: you're welcome! [01:30] is launchpad's openid broken? [01:33] CaioAlonso: It's working fine for me. What's the trouble? [01:34] wgrant: I'm having some problems connecting it to askubuntu's login [01:34] wgrant: http://pastebin.com/NJeyfGxA [01:38] CaioAlonso: It sounds like there's something wrong with your account. Could you open a support case at https://forms.canonical.com/lp-login-support/? [01:38] Ah, https://forms.canonical.com/sso-support/ is probably better, actually. [01:41] oh, sure [01:41] thanks! [01:53] wgrant: well, as 90% of errors, it was my mistake, I forgot that there was a hifen in my launchpad url [01:54] CaioAlonso: Which username did you try? [01:54] CaioAlonso: There's still something wrong. [01:55] wgrant: I was using https://launchpad.net/~caioalonso [01:55] but the correct one was https://launchpad.net/~caio-alonso [01:55] everything worked here [01:56] Odd. === almaisan-away is now known as al-maisan === Lcawte|Away is now known as Lcawte === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === mrevell is now known as mrevell-lunch === mrevell-lunch is now known as mrevell === fjlacoste is now known as flacoste [13:48] Hi all [13:49] larz JS has this awesome testing runner that is able to run yui3 tests on jstestdriver. I'd like to build a module from it so it could be released on the yui-gallery which means it would be on the yui CDN. Now the problem with that is that all code that is distrubuted there needs to be under BSD license. Do you think it would be possible to relicense this part of the code? [13:51] I am talking about this file: http://bazaar.launchpad.net/%7Elaunchpad-pqm/lazr-js/toolchain/annotate/head%3A/src-js/lazrjs/testing/testing.js [13:51] rockstar, sidnei told me you might be the guy to ask? :) === matsubara is now known as matsubara-lunch [13:58] so, I'm trying to use the summit software in conjunction with lp, and some of my bps are not showing up in https://edge.launchpad.net/sprints/ods-b/+temp-meeting-export for example: https://blueprints.edge.launchpad.net/nova/+spec/service-provider-readiness [13:58] can anyone tell me what I am doing wrong? === LinuxJedi|away is now known as LinuxJedi === al-maisan is now known as almaisan-away === zyga is now known as zyga-coffee === zyga-coffee is now known as zyga === matsubara-lunch is now known as matsubara [15:14] dendrobates, hi [15:15] dendrobates, looks like the priority is making that specific bps to be filtered out. The view only shows bps with a priority set higher than undefined [15:16] s/higher/greater/ [15:16] matsubara: thanks [15:19] np === deryck_ is now known as deryck === deryck_ is now known as deryck === Ursinha is now known as Ursinha-lunch === Guest63625 is now known as jelmer === beuno is now known as beuno-lunch [17:20] hey LP-ers :) -- What's the LP stance with creating meta-projects? === almaisan-away is now known as al-maisan === zyga is now known as zyga-nc10 [17:31] paultag, hi [17:31] matsubara, Hello :) [17:31] paultag, what do you mean? a project to group projects? [17:32] matsubara, No, I know how to do that -- I mean a project to file Community issues. I was considering using a LP project to file LoCo bugs ( with the LoCo Council ( even more so if we could set them to private ) ) [17:33] matsubara, so if we have a community member who has issue with a loco team ( on whatever ) they could file a bug against that project privately and all that jazz. [17:34] paultag, I see. I don't thing there's anything in the ToS disallowing you to do that. you'd basically be using the bug tracker as a task/issue tracker rather than software bugs, right? [17:34] but it would not be a F/OSS project ( and all that other stuff ) [17:34] matsubara, I had a project deleted once on me because it was not a F/OSS project, so I just want to tripple check :) [17:34] matsubara, yeah, exactly [17:35] mrevell, jml: do you know the answer for paultag's question ^ ? is this usage of the bug tracker allowed? [17:35] * jml reads [17:36] paultag, matsubara: yeah, that's allowed. the thing we don't allow is hosting proprietary code without paying us. === beuno-lunch is now known as beuno [17:36] jml, could I still have the bugs set as private to start off? I'd like to keep reports such as this private ( unless they'd like to make it public ) [17:36] jml, this is also for Ubuntu it's self [17:37] oh right. hmm. don't know about that. [17:37] this is not entirely dissimilar to a problem the Launchpad project itself faces with feedback email. [17:38] jml, mmmhum :) [17:40] I'd really like to ditch the bugzilla instance we're using ( and allow other people to file issues for us ) :) [17:41] paultag: policy-wise, there wouldn't be a problem [17:41] paultag: I'm just less certain about how to actually do it. [17:42] jml, well, if we set the loco council as the owner ( and everything else ) the only people with admin / owner access would be the LoCo Council and the Community Council. We've all signed the CoC -- if we swear to not use the code part of it, could it be turned into a proprietary project? ( or disable branches for it or something ) ? [17:42] I guess in the worst case, you could file bugs as "security issues". [17:42] jml, hummm. [17:46] jml, well, would you mind if I ping'ed you in a day or so? I'd like to think it out before I start spouting out hacks around LP. Let me see if I can come up with anything ( and if you would not mind doing the same ) :) [17:46] paultag: sure. maybe you could send an email to launchpad-users? [17:46] paultag: there are lots of people there who know more about this part of Launchpad than I do, and who aren't awake right now. [17:46] jml, Sure, let me email the rest of the LoCo Council and make sure this is something we want to go after -- and then I'll send off to lp users :) [17:46] jml, np, I'll sit and idle, then follow up [17:46] paultag: sounds like a plan :) [17:46] jml, thanks so much :) [17:47] cheers, thanks matsubara as well [17:47] paultag: my pleasure. [17:47] paultag, np === Ursinha-lunch is now known as Ursinha === LinuxJedi is now known as LinuxJedi|away === yofel_ is now known as yofel [18:22] paultag, Would https://launchpad.net/ubuntu-community meet your needs, or do you need something more specific? [18:37] sinzui, hello! During UDS, I believe you mentioned that the ability to upload GPG & SSH keys had just landed. Is that reflected in the API docs yet? (I see ssh_key and gpg_key in current API, but not clear how to upload) [18:38] mterry, We landed the code that permits the user to give temporary desktop access to do the task === brianchidester is now known as brianchid [18:40] mterry, We can move the views that allow a user to register keys to the model so that we can access them over the api. SSH is pretty simple [18:40] sinzui, right... And it would last for, say, 30 min. === al-maisan is now known as almaisan-away [18:42] mterry, the gpg key will still require a hop through email. Maybe we can find a trusted way to avoid email. [18:43] mterry, Signing the CoC is still a challenge for users. Maybe we could create a stand-alone CoC signer app that does the Lp registrations. Quickly can employ it [18:43] sinzui, maybe because I wasn't there for the last UDS's round of discussion on this, I'm a little unsure of the technical details. quickly will call some LP API to enable special privileges for the next 30 mins (is this call documented?). Then, there isn't yet an API for the SSH push? And the GPG you say won't have an API at all? I had come away from this UDS thinking gpg/ssh were solved problems [18:44] * sinzui want to reduce the steps and confusion about creating, publishing, and signing (with the correct key) [18:44] I realize CoC is a separate also hairy issue [18:44] mterry, they are solvable in a week [18:45] sinzui, that's fine. I just had thought that I could start coding my end of it. Didn't mean to rush you. Just thought the ball was in my court and wasn't sure how to hit the ball [18:45] * persia worries about special permissions windows and wonders if there isn't a way to do it that isn't time-based [18:45] mterry, sshkeys really are simple...we should never have put the registration logic into the view. the core object was and is good enough to do it all [18:47] mterry, I'll look into getting both of these done in the next week weeks. [18:47] sinzui, that was a jargon-heavy sentence. I'm not hip to views and core objects. you mean I can do ssh pushes today with the ssh_key object? [18:48] mterry. API works with the objects. The pages are views on an object (or an aggregation of objects). There is a lot of logic in the wrong place when we work with older objects [18:49] mterry, you cannot setup anything today. [18:50] sinzui, understood. Let me know when there is enough for quickly to do a thing [18:51] mterry, We are using the "quickly" tag to track issues that must be done: https://bugs.launchpad.net/launchpad-registry/+bugs?field.tag=quickly [19:03] Hey guys, can anyone tell me how to get access to branch something out of devpad? [19:04] I've got access to chipstrap, but when I try to branch I get denied === menesis1 is now known as menesis === LinuxJedi|away is now known as LinuxJedi === matsubara is now known as matsubara-afk [21:02] persia, I'd like the bugs to be filed private ( a lot of the issues we hit on are personal, and it would really be kinda lame of us to make that public without talking privately ) [21:03] paultag, I think I like the idea of using "security" for that, and have the security contact be a narrower team that can be exposed, but perhaps another project works as well. [21:04] persia, I think I might just add it to my budding loco admin tool, and have it ask ( and default ) to setting the bug as a security issue [21:04] persia, let me ask the rest of the council before I done goofed :) [21:05] paultag, Sure. I think we want the same class of solution for all the delegate councils, although I'm not convinced it needs to be the same specific solution. [21:05] persia, aye :) [21:23] deryck: were we going to catch up? [21:23] thumper, oh, I hadn't thought about that since Francis came back. I'm about to leave actually [21:23] deryck: geez man [21:23] deryck: how about next week? [21:24] thumper, definitely. Do you want to keep it as a standing weekly or bi-weekly thing? [21:24] deryck: I'd be happy bi-weekly [21:24] but yeah, some form of standing chat would be good [21:24] hi everyone [21:24] thumper, ok, I'll update the calendar then. bi-weekly, starting next week. [21:24] deryck: awesome [21:24] hi Puck` [21:25] i have a problem logging in to the ubuntu wiki pages. Whenever i login with my launchpad login it always asks me for a new username, doesn't take the last one. I did reset my password, but the same problem occurs [21:25] i got sent here by paultag, so blame him if i'm in the wrong place :P [21:25] Puck`, :P [21:27] The trick is that LP is no longer really the place that authentication stuff happens, but other places aren't really extant. [21:27] so where should i look at? [21:27] or what should i change? [21:29] persia, I thought SSO was using LP as the provider [21:31] also another question that came up, is there a way to get inactive "name's ? [21:37] paultag, Other way about [21:37] persia, ahha [21:41] I can't figure out how to use lplib's iPerson.inTeam -- It's always throwing an error -- I have two team objects ( I can check with is_team, and name ) -- and it will not find inTeam or in_team or anything. It's really lame. Anyone remember the syntax? [21:41] I really don't want to loop through memberships [21:52] wgrant: 'NoneType' object has no attribute 'processor' <-- WTF> [21:52] ? [21:53] * lamont decides he doesn't care that much [21:54] paultag: iirc, inTeam is not exposed in the api for reasons i don't understand [21:55] wgrant: actually, it turns out that I do care, since it threw the same exception. you around? [21:59] lamont: iridium is still being hogged by a runaway prism build, is my only hope to delete the build-dep for it? [21:59] mwhudson, hurmmm. Well looks like I'm going to have to iterate through and hammer the server? [21:59] Smart! [22:01] micahg: let me go stab it in the face [22:01] lamont: ok, if you can kick the build from the queue, that would be great (it will always fail) [22:02] micahg: ah, so it needs to be made to not start, yes? [22:02] lamont: yes, PIE on natty with xulrunner/firefox is broke ATM [22:02] in case you were wondering the specifics [22:02] ok [22:08] micahg: all dead now. [22:08] and corpsified, even [22:09] lamont: thank you, I'll try to make sure that xul192 is fixed before natty Mozilla daily builds are reenabled [22:09] thanks [22:09] there really isn't a pretty way to kill those today [22:09] * lamont goes to see if he already filed a bug about that [22:18] wgrant: do you know if the 'break lp-buildd out into its own tree' has a bug associated with it? [22:19] meh. it does [22:23] btw can someone take a look at build https://launchpad.net/~pali/+archive/kopete/+build/1973460 ? I doubt it's still uploading for over a week now [22:41] lamont: Hi [22:41] hi === Ursinha is now known as Ursinha-afk [22:42] lamont: What's the issue with processor? [22:43] https://edge.launchpad.net/builders/adare <-- wgrant: can you see that page? [22:44] I can. [22:44] That is rather special. [22:44] What's the traceback? [22:46] is that buildd side, or build-manager side? [22:46] manager. [22:46] buildd doesn't know about procesors. [22:47] http://paste.ubuntu.com/526017/ [22:47] It sort of sounds like we have a BPJ without a BPB. [22:47] But that's impossible. [22:47] ... [22:47] congratulations [22:48] got any sql queries you want me to do to validate that assertion? [22:48] No, no, that's not the issue. === LinuxJedi is now known as LinuxJedi|away [22:48] it was doing a build when it crashed, iirc [22:48] Hrmph. [22:48] And it's reproducible? [22:48] rebooted about 8 hours ago [22:49] Just that one builder? [22:49] I cleaned up the build-$mumble tree, and restored the builder to OK status, and now we're where we are. let me stick it back in OK again just for good measure [22:49] yep [22:49] said builder falls over a lot, to be fair [22:49] it hates its kernel [22:50] Do you know if cesium has had its code updated lately? [22:50] And are there any adare-related errors just before that one? [22:50] dunno, and looking [22:50] there was a rollout of something earlier today, dunno what or where [22:50] mbarnett: ^^ [22:51] 2010-11-04 14:32:09+0000 [QueryProtocol,client] adare was made unavailable, resetting attached job [22:51] 2010-11-04 18:39:08+0000 [QueryProtocol,client] Builder adare failed a probe, count: 1 [22:51] and a traceback at 18:39 of course [22:52] Did the same exception occur? [22:52] yep [22:52] and the closest prior adare mention before that exception was 4 hours earlier [22:52] quite possibly about the time that it fell over in mid-build [22:53] wgrant: and sadly, I'm about to EOD [22:56] if you happen to smack bigjools with that in his morning, one of y'all might get things straightened out before I come online in the morning [22:56] lamont: bigjools isn't around this week, AIUI. [22:56] meh [22:56] either way, gotta run. [22:56] But I have some idea of what's going on. [22:56] k [23:04] OK, so I have to check if a team is a direct member of another team. Iterating through takes a minute(+) and slaughters the server. What's the best way to do this? [23:04] Can I compare two membership collections and find the common ones? [23:05] I can't string index the membership collections, either [23:05] ( and doing a .name call really starts killing speed ) [23:07] ( also with lplib, but I figure ya'll knew that ) [23:09] paultag: You could do a direct request for the membership, and then check its status. [23:09] paultag: There's no method to do this, but you can construct the URL to the team_membership manually (/~TEAM/+member/PERSON) [23:09] wgrant, Ah, great. I think you helped me do this for another bit, that sounds great [23:09] A direct inTeam method isn't exposed due to concerns around revealing private team memberships. [23:09] Ahhha [23:10] wgrant, thanks a ton, again [23:10] I'm not sure if that's still a valid concern, with recent changes to the data model. [23:10] Meh :) [23:10] if there's a way to do it, there's a way to do it :) [23:10] I need to start documenting all these little bits somewhere [23:12] man, I'm pushing lplib hard :) [23:12] sorry :) [23:20] wgrant, I have it "working" ( python is not throwing errors ) -- but it seems to be returning unspecified behavior when I request ~team/+member/team_not_on_the_team -- It's returning it's status in the 0th team it's in ( I had it output "status", and it's returning "Administrator" even though they're not in ~team ) [23:20] wgrant, any ideas? [23:20] paultag: What's the code? [23:20] let me pastebin it [23:21] wgrant, http://pastebin.com/WDGffiC5 [23:22] paultag: What if you use 'team.name' when constructing the URL, not just 'team'? [23:22] wgrant, AttributeError -- No attr "name" [23:23] just tried team.team.name [23:23] same failure [23:23] Oh wait! [23:23] I got a 404, this is good [23:23] Hm, team.team.name should work. [23:23] wgrant, Yup, I think that'll do it. You rock :) [23:24] Great. [23:24] I need to see if we can expose inTeam now. [23:24] wgrant, :) [23:26] wgrant, looks great & it's working, cheers! [23:26] wgrant, I owe you a few beers :)