[00:34] What is the LP poll date time format? YYYY-MM-DD HH:MM:SS didn't work [00:35] nhandler: That should work. What happens when you try it? [00:35] wgrant: "There is 1 error" "Invalid datetime data" [00:37] nhandler: Something like '2010-03-25 11:00:00' works fine for me. [00:38] wgrant: I had 2010-03-24 01:00::00 [00:38] With a double colon? [00:39] Ah, good catch. It also looks like I need to push it back to at least 12 hours from now [00:39] I think so. that restriction is quite ridiculous. === MTeck is now known as MTecknology [03:16] where's kiko been? [03:23] secret missions and the like [03:24] Why do you ask? Is there something you need help with or just curious? [03:24] :) [03:36] * wgrant has also been wondering that. [03:42] rotations - spread the knowledge around inside the company [03:49] Hello! I'm trying to add a team member and I get "Unexpected form data" "lp doesn't understand the form data submitted in this request" [03:49] Help? [03:49] both on edge & stable [03:50] xnox: Your browser doesn't send Referer information? [03:51] Hmmmm I'm using daily firefox 3.6 builds from ppa [03:51] I'll try with epiphany [03:51] It's more likely to be an extension. [04:18] poolie: I agree with you in your comment to bug 532055 [04:18] Launchpad bug 532055 in launchpadlib "Trusted credential-management apps are broken and doomed" [Low,Triaged] https://launchpad.net/bugs/532055 [04:18] thanks :) [04:19] poolie: One suggestion was to have an openid provider dedicated to desktop apps. [04:20] poolie: But openid consumers would have to accept more than one openid provider as valid for any given account. [04:37] wgrant, thanks. Dunno about extensions disabled some still didn't work. Worked in epiphany though [04:37] xnox: Odd. It probably was blocking Referer, though. [04:40] doctormo: so the idea is that there are very many different desktop apps that all need to talk to lp? [04:40] and they're not all trusted? [04:40] poolie: No, there all all these desktop apps that want to talk to launchpad and launchpad doesn't trust anyone. [04:40] The user could trust them quite fine [04:40] istm the simplest thing that could possibly work is to just say "there are credentials in ~/.launchpadlib/credentials" and that's it [04:41] and/or in the gnome keyring [04:41] How does U1 do it nowadays? [04:41] Still opens up a browser? [04:41] wgrant: I think so [04:42] But then again I don't think U1 uses launchpad credentials does it? [04:42] It uses USSO. [04:42] I thought it used ubuntu SSO [04:42] Which LP is about to. [04:42] right [04:42] near enough to the same doohicky [04:42] except wgrant can't audit our SSO code ;P [04:42] lifeless: Sadly, because it's probably full of even more pathetic holes. [04:43] poolie: If we could get it to the point of just pointing stuff at .cache/launchpad/credentials but the problem with that is you can't limit what each app can do and all apps will need all permissions. [04:43] If that's acceptable, then just do that. [04:44] lifeless: That's a horrible thing to say, taunting that it's not FOSS. ;-) [04:44] well [04:44] istm that if you don't trust this code, you really shouldn't be running it in your real account [04:45] tbh messing with your launchpad account is one of the least dangerous things it could do [04:45] Aye [04:45] if we are going to have a "do you trust $app to do $foo" then [04:45] The deb it installed with already had root access via preinst. [04:46] right [04:46] then firstly it would be nice to do that through the existing system mechanism [04:46] secondly i think it needs finer-grained permissions [04:46] hm [04:46] i would actually kinda like a role account for automatic stuff [04:46] doctormo: I'm not intending to taunt; wgrant has been very very helpful with lp and SSO in recent (and not so recent) times [04:47] like 'poolie-bot' [04:47] poolie: Have you seen the blueprint? [04:47] then it's a bit clearer that it can't change my actual account itself, its actions will be attributed, etc [04:47] maybe not? [04:47] * poolie checks [04:47] https://wiki.ubuntu.com/BluePrints/LaunchpadDesktop [04:48] actually yes [04:49] so istm it's kind of unnecessary complexity considering it doesn't actually add any security [04:49] sorry to be harsh [04:49] poolie: Correct [04:49] poolie: It's not meant to [04:49] this has to be done on the server side anyhow to accommodate non-ubuntu clients [04:49] Imagine the day when our desktop security model doesn't suck. Then it might be useful. [04:50] I think it's still useful, though, to have a common interface for authorizing all the applications. [04:50] makes it easier for app authors, and nicer for users. [04:50] so does this actually need to be different process [04:50] or could it be a standard library for getting tokens- [04:50] poolie: yup [04:50] that's a bit more advanced than "do the web browser thing then press enter" [04:51] poolie: If only the web browser could just quit and kick off a damn signal, that would be helpful for workflow. [04:51] doctormo: so to back up a bit, what is the actual benefit of this? [04:52] poolie: Desktop apps don't break trying to log onto launchpad by faking a web browser. [04:52] poolie: Desktop app makers are _not_ going to want to do OAuth via website, it's rediculous. [04:52] what do you mean? [04:52] Which? [04:53] doctormo: poolie: Desktop app makers are _not_ going to want to do OAuth via website, it's rediculous. [04:54] poolie: Would you want your lovely designed workflow to be springing up unknown applications that can execute any old random code, that can't singnal when they're done to do god knows what, interupting the user and breaking your app in two? [04:54] right, you're saying the current 'get me a token' is gross [04:54] and i agree it is [04:55] it manages to be gross for both command line and gui apps [04:55] OAuth was never meant for desktop apps and was further pulled back from doing authentication (which is why we use openid) [04:55] so if we look at the best/nicest web-api client apps, what do they do? [04:56] in my experience they mostly ask you for a username/password [04:56] and then some do an oauth dance, but asking you to click a button when you're done [04:56] (which is gross but not so bad; you could presumably avoid that by just polling a url to see if we're done.) [04:58] I could never understand why I couldn't get a token from launchpad by doing an ssh key challenge, forget passwords. [04:58] right [04:59] that would be pretty classy [04:59] at least for some subset of users [04:59] or gpg for that matter [05:00] it does seem a bit nicer that what eventually gets stored in a dotfile is a token not a password [05:02] poolie: I'd never recommend saving a password, we should always be encouraging apps to store tokens. [05:02] i did file a bug against this pointing out that reseting your password no longer locks out people who may have created tokens in your name [05:02] which is a bit of a drawback [05:02] but it's still better, yes [05:03] Changing passwords doesn't even revoke cookies :/ [05:03] The UI for token management in launchpad is a bit bad ATM I think, "Associated Applications" link, token deletion etc. [05:04] i guess the two channels that seem to make sense to me are [05:04] A- enter a username and password; they're sent to the server which gives back a token [05:04] this is nice for text mode apps or mobile apps [05:04] it does require the user actually have a password [05:05] and B- pop up a web browser window (probably in an already-authenticated session), click a button in it, and then have the app somehow detect that you've done this [05:06] Doesn't U1 redirect to a local HTTP server to indicate that it's finished? [05:06] i don't know, but that would be one way to implement it [05:06] another would be to just poll a url every so often [05:07] wgrant: yes [05:07] the U1 daemon provides a callback url [05:07] that the auth completion opens, and which is localhost [05:12] * doctormo is off to bed [05:12] talk later [05:18] im guessing this is the wrong place, but who's responsible for planet ubuntu? [05:21] pwnguin, read the docs about Ubuntu Membership on wiki.ubuntu.com [05:21] it has pointers to Ubuntu Planet, who gets syndicated etc [05:22] The lp branch were config is stored is owned by somebody probably you can get in touch with them [05:22] * xnox finished with random thoughts [09:20] hi there [09:20] Hello mat_ [09:22] I'm trying to figure out a problem we have, our trunk here https://code.launchpad.net/~openstreetmap/josm/trunk is failing due to http://launchpadlibrarian.net/41817977/openstreetmap-josm-trunk-log.txt [09:22] I'm wondering if it's a problem on our side of the repository, or on the launchpad servers [09:23] and if I should ask for it to be fixed or try to add the repository again :-) [09:24] mwhudson, Is it too late for you to be around? [09:25] mrevell: yes [09:25] :) Ok [09:25] mrevell: email? [09:25] mat_, Would you mind filing a request in Launchpad Answers? https://answers.launchpad.net/answers [09:26] mrevell, I don't mind, no :-) [09:26] mwhudson, mat_ above has an import issue but if you're not around, don't worry. [09:26] thanks mat_, we'll take a look later today. [09:26] mat_, I'd help but I'm not familiar with the code import system. [09:27] jelmer: ^^? [09:27] mat_: that does look rather like some odd server side problem [09:27] * mwhudson isn't here though [09:43] wgrant,mat_: That's a cscvs import, I'm not really familiar with it. [09:44] jelmer, cscvs ? it's a svn repository [09:44] mat_: Confusingly cscvs also supports svn repositories [09:44] ah, ok :-) [09:44] mat_: Older code imports are made using cscvs, newly created imports use bzr-svn [09:45] mat_: It might work better with bzr-svn, so we can try to remove the existing branch and re-add it using bzr-svn [09:45] jelmer, well, as it's been broken for five months, anything could be better than what it is now :-) [09:46] mat_: If you're in the openstreetmap team you should be able to do the delete + readd yourself [09:47] I'm not that familiar with launchpad yet, joined today, but I'll try to see to that today though, thanks :-) [09:48] mat_: if you need any help, feel free to ping me directly or just ask here [09:48] will do :-) [09:50] jelmer: Oh, I saw 'foreign branch' and presumed bzr-svn. Sorry. [09:51] wgrant: No worries, I'm familiar with cscvs as well, though not as much with its internals as bzr-* [09:56] mat_: fwiw the current branch appears to be a mirrorred bzr branch, not a code import [09:57] jelmer, yeah, mistake on my part [10:30] dpm: have you seen this? http://www.crowdin.net/ [10:39] <_Andrew> Is there a way to delete my comments on a bug? [10:42] _Andrew: no [10:44] bialix, someone pointed me to it some time ago, but I haven't looked at it since. It is proprietary, isn't it? [10:44] it seems so. I just like the idea of transaction memory [10:44] so maybe something could be used as inspiration to improve lp translations [10:52] bialix, sure, thanks a lot for the heads up. The problem, though, is not so much that there is a lack of cool ideas, but rather the resources to implement them :) I'd love to see translation memory support as well. I'm hoping to see more community contribution in LP Translations as well, and it is featured in the Google Summer of Code as well -> https://wiki.ubuntu.com/GoogleSoC2010/Ideas (from nr. 39 onwards) [10:57] dpm: yes, I understand. we have the similar problem in qbzr re lack of hands [10:57] dpm: the one thing I would like to see improved is the search for the strings [10:58] when there is several hundred strings it's very hard to find desired [10:58] because of presence of & to indicate hotkeys [10:59] bialix, you mean implementing string search on a global level through LP? Otherwise, what would you like to see improved in the current search functionality within translation templates? [11:00] search within specific template [11:00] maybe all templates of one project === mrevell is now known as mrevell-lunch === matsubara-afk is now known as matsubara [13:17] Hi all... [13:18] I like to help the community by building some common software as a contribution using ppa facility in lp. Can anyone help me to do this? === jamalta_ is now known as jamalta [13:22] Hello.. :-? [13:22] napster: Are you in the #ubuntu-app-devel channel? [13:23] napster: Such impatience the youth have these days. ;-) [13:23] doctormo, ? [13:23] oh... sorry [13:23] doctormo, Im not in ubuntu-app-devel [13:24] napster: That's where you can find help on making an application. [13:24] doctormo, Can you help me to use ppa facility in lp? [13:24] napster: As a user, uploader or back-end hacker? [13:25] doctormo, I'm not sure, I think I've nessessary build-essential + gcc. Can I build and package some apps and upload? === mrevell-lunch is now known as mrevell [13:28] morning. Does launchpad's ppa support PowerPC packaging? [13:29] hi! there is a huge queue again on builders - eg. 4 builders and 15h queue on amd64 - is someone attend on this? [13:30] pabelanger: I don't believe so, it's limited to what the builder is building. [13:37] Ya, I don't see any powerpc being build within the PPAs. Not sure if this is a limitation of launchpad or because there are not powerpc boxes attached. [13:40] pabelanger: Having no ppc machines is a bit of a barrier, but I believe it's policy to only have a limited number of targets. (IANALD) [13:46] wow... only 4-5 builders online for the PPA builds... thats got to be a record.. what seems to be the problem? [13:47] shadeslayer: it's usually not a problem, just that the extra PPA builders are borrowed from elsewhere... we get to use them when they're not needed (which is most of the time). [13:47] noodles775: ooh.. [13:47] shadeslayer: and you meant 4-5 i386 builders right? [13:47] noodles775: noth amd and i386 [13:47] As there are 12 ppa builders. [13:48] *both [13:48] Right. [13:48] amd64 4 === felix is now known as adevc [13:56] thanks for launchpad - problem resolved itself ... [13:56] bye all [14:59] hello all... [15:02] orbarron: hi [15:04] not sure if the is the right place to ask... but I have pulled down 9.10 through rootstock and I am wonder if rootstock locks it's release to 9.10 based on it's releases (e.g. rootstock-0.1.3 will always give me a base release for 9.10 as long as karmic is selected)... [15:06] When I try run dput a package to my team's ppa, it gives me this error: http://paste.ubuntu.com/400582/ [15:06] What should I do? [15:07] Flare183: can you paste your changes file too? [15:07] sure [15:07] I'm guessing you're missing the files section? [15:07] oh wow [15:07] the changes file it empty o.O [15:08] ;) [15:08] is* [15:09] * Flare183 can't seem to package this binary package at all [15:16] a bunch of packages in one of my ppa's failed to build because a package wasn't done building when they were uploaded. Strangely, they weren't put into Dependency Wait, but were instead marked as failed. Is there someway to mark all of them for rebuild enmass? There are alot of packages I need to have rebuilt. [15:20] ripps: you could use the API: https://edge.launchpad.net/+apidoc/1.0.html#build [15:20] But other than that, I'm not sure of an enmass method via the ui. [15:24] is the "upstream bug comments to show on LP bug" turned off? im not real sue what it is called [15:25] hello! I need to make tgz file availabe using the "Download" option in lp [15:25] I cannot find any doc on how to do this [15:25] any suggestions? [15:27] bac or sinzui, ^ ? [15:28] jenred: you add the file to a release. You can create a release from a series. [15:30] sinzui ok thx [15:33] sinzui is there anyway to generate a release off of a branch? without registering a series etc [15:33] jenred: no [15:35] thx === shadeslayer is now known as shadeslayer_ [15:40] So, I've a *brand new* unexpected login screen from launchpad. Did something change recently, or am I being spoofed? === salgado is now known as salgado-lunch [15:52] persia: No, that's as expected I think [15:52] persia: It's openid [15:52] doctormo: But I'm not being redirected to my OpenID provider. I'm aat a +decide page. [15:54] persia: screenshot? [15:55] sinzui we are actually trying to make a dev image available - and I wanted to keep it under version control [15:56] but it seems like there maybe some file size-limit issues? [15:57] jenred: large files that are about 60Mb+ tend to timeout [15:57] persia: I think I saw that too === flacoste is now known as flacoste_lunch [15:58] jenred: some projects set the releasefileglob on the series to point to a server where the releases are published. The servers is checked each day and new files that match the release file pattern are uploaded to launchpad [16:00] sinzui are the files actually uploaded or is it just a link to another download site? [16:00] jenred: the finder will create the milestone and from the version number in the file for you [16:00] * persia wants some confirmation that this is really LP before entering credentials. [16:00] they are uploaded to the launchpad librarian [16:00] persia is the screen green? [16:01] and it appears to be openid enabled now [16:02] Can someone take a look and see why my PPA does not seem to be purging superceded packages? [16:02] https://launchpad.net/~kde3-maintainers/+archive/trinity-svn-nightly/+packages?start=0&batch=75 [16:02] jenred: Yes. [16:02] sinzui sort an extra layer of complexity [16:02] persia you are fine [16:03] sinzui I really just want to take .tgz file - check it in and make it available for download [16:03] it's for GSoC students :> [16:03] jenred: launchapd downloads feature is badly named. It is intend as an automatic mechanism for projects to publish releases to users. it is not ftp storage. The feature in intrinsically flawed because most users that the location of the archive so that their OS installs everything for them === doko__ is now known as doko [16:08] sinzui thanks === matsubara is now known as matsubara-lunch === Ursinha is now known as Ursinha-lunch === leonardr is now known as leonardr-lunch === beuno is now known as beuno-lunch [16:41] barry: heh, http://lwn.net/Articles/380065/ === salgado-lunch is now known as salgado === flacoste_lunch is now known as flacoste [17:12] where can I find latest touched bugs in Ubuntu? [17:14] ari-tczew: come again. what exactly are you looking for [17:16] previously the page at https://bugs.launchpad.net/ubuntu/ shows latest touched bugs in Ubuntu, currect version doesn't work correctly as I see [17:18] ari-tczew: you can select the dropdown and say "Least recently changed" and click on search === matsubara-lunch is now known as matsubara [17:21] nigelb: "most recently changed" is better... btw. this is not what I need :( [17:21] why https://bugs.launchpad.net/ubuntu/ has been changed? [17:22] ari-tczew: I dont think it has been changed [17:22] are you looking for most recently reported bug? === beuno-lunch is now known as beuno [17:28] Concerning Bug #538236 now I'm not even able to submit replies to bugs.edge or answers.edge anymore. Any idea what I'm supposed to do? Thanks. [17:28] Launchpad bug 538236 in launchpad-foundations "UnexpectedFormData does not show helpful error messages" [High,Triaged] https://launchpad.net/bugs/538236 [17:53] how often does Launchpad pull in new packages from Debian? === deryck is now known as deryck[lunch] [17:57] for the LP Debian mirror? AFAIK twice daily [17:58] geser: okay, just wondering. I uploaded a package to sid earlier today, and want to merge in the changes via bzr in lp :) [17:59] (into lucid) [18:01] you might also want to check http://package-import.ubuntu.com/status/. If the package import into bzr failed in the past, it won't try for new uploads until the first error got resolved [18:01] geser: mk. it hasn't failed yet. there's no way to manually trigger it? (like how you can trigger mirrored branches) [18:04] lfaraone: what's the package name? [18:07] geser: autokey. === Ursinha-lunch is now known as Ursinha === deryck[lunch] is now known as deryck [18:27] lfaraone: the only trigger I know of for package imports is pinging james_w [18:28] but wait first till gina synced it to the LP Debian mirror and the automatic import attempt has be tried before pinging him [18:28] I can trigger them if LP has mirrored from Debian, but that should be automatic [18:51] james_w: Hi, a question for you on imports - I'm fixing a broken one, and the current situation is that testing == unstable for that package. I notice that the importer wants to import the current debian revision with lefthand ancestry from the previous version published in *testing*, with a *merge* of intermediate versions that were in unstable but never propagated to testing. This produces an unnecesarily obfuscated revision graph, IMO. Is t [18:51] his supposed to happen? [18:52] yes [18:52] Presumably if the import had never been broken, it would have been imported onto the sid branch when uploaded, and then pulled into squeeze when it propagated to testing - producing a different DAG [18:52] that's true [18:52] I'm not wedded to that implementation though [18:53] The package is subversion: what's your preference, should I just file a bug explaining how to fix the current import and let it happen with the odd merging? [18:54] Or do you view the inconsistency as something which should be fixed first? [18:56] maxb: hang on, it won't have pulled it it was working I don't think. If the uploads never made it to testing then they won't be in the left-hand history. [18:57] if testing has previously diverged from unstable that is [18:57] Oh, will it actually do merge commits for testing propagations? [18:58] I think it tends to do that yes [18:58] OK [18:58] if you think it should pull when possible then I'm open to changing it [18:59] The fix is to restore the missing debian branches by pulling the relevant revisions out of the ubuntu ones, and adding missing upstream-* tags - is it easier for you if I push branches and ask you to reassign or re-push them to ~ubuntu-branches, or if I provide a shell script to manufacture appropriate ones? [19:00] I can reassign quite easily [19:00] so I'm happy for either really [19:00] OK, I will push and file a bug on udd asking you to reassign [19:02] thanks === leonardr-lunch is now known as leonardr [19:29] james_w, geser , okay, just wondering. [19:50] james_w: I have extended LP 516464 with the reassign request. Should I subscribe you to it? [19:50] Launchpad bug 516464 in udd "Debian import branches are missing for 'subversion'" [Undecided,New] https://launchpad.net/bugs/516464 [19:51] maxb: I'm already subscribed [19:51] thanks, I'll do that for you soon [20:02] hey guys [21:00] Can someone take a look and see why my PPA does not seem to be purging superceded packages? [21:00] https://launchpad.net/~kde3-maintainers/+archive/trinity-svn-nightly/+packages?start=0&batch=75 [21:00] It should have over 1Gb free at the moment === salgado is now known as salgado-afk === matsubara is now known as matsubara-afk [22:15] i cannot remove pacakges from my ppa... i recevie an error [22:16] i remove succesfully [22:17] great === _thumper_ is now known as thumper