=== JanC is now known as Guest35472 === JanC_ is now known as JanC === ganesh is now known as Guest79962 === Guest79962 is now known as gkadam [06:38] hiya [06:38] somebody just asked "Anyone knows why launchpad is failing to upload 'my' snap to the Ubuntu Store? Well, I have only one login for all my Ubuntu stuff. The error is: Store upload failed: 403 Client Error: Forbidden" - I'm not quite sure what the issue might be. Does anyone have an idea? [06:40] dholbach: First thing to try is unchecking and rechecking the store upload checkbox on LP, to get new authentication tokens. [06:40] If that doesn't fix it then we'll have to dig deeper. [06:41] ok, I'll let them know [06:41] thanks === dpm is now known as dpm-afk === dpm-afk is now known as dpm === davmor2_ is now known as davmor2 [14:11] hello [14:13] I'm having some build issues with launchpad, and I'm not sure what to do. pbuilder needs an Internet connection in my case, how might I remedy this? I have some ubuntu package build deps and also pip dependencies. I am using dh_virtualenv to create a source package, which seems to be good. How might I accomplish a build with no internet on the ppa farm? [14:14] Would it be a better approach to build the packages myself and upload to the PPA? [14:16] you need to depend on the binary python packages [14:20] hmm [14:21] I'm not sure that I can do that, as not all of them exist as packages [14:21] if they're not available in the ubuntu archive already, you'll need to package them in your PPA [14:21] ahh [14:21] ok, crap lol [14:21] thanks [14:21] I guess I need to contact some maintainers. [14:25] i'm not sure what all snap builds can access, but you might want to look at packaging as a snap instead of debs. i think the snap builds are a little more lenient there [14:28] That's correct, but it depends what you need to achieve. [14:29] anyone here who can fix my launchpad login? :( [14:29] * apollo13 doesn't like to create new email addresses for services [14:29] roadmr: ^- I think you helped me last time with somebody who had a confused SSO state [14:30] https://oops.canonical.com/?oopsid=OOPS-b32e2f4e37f854e5a7614319da6b1398 is the oops [14:30] https://oops.canonical.com/?oopsid=OOPS-b32e2f4e37f854e5a7614319da6b1398 [14:30] cjwatson: yes, that time there were two SSO accounts and two LP ones, the openid was mismatched [14:31] I can generate new OOPS if needed :D [14:31] :) [14:32] apollo13: Do you know your LP username? [14:32] cjwatson: it might have been apollo13 or f.apolloner [14:32] (Error ID: OOPS-98fee25e8cb199977fa9a83f2afea2cf) <-- new oops if it helps [14:32] https://oops.canonical.com/?oopsid=OOPS-98fee25e8cb199977fa9a83f2afea2cf [14:33] ~apollo13 exists and has the OpenID from that OOPS [14:33] yeah, apollo13 should be relatively old (or rather recreated) [14:34] I think I have/had one in 2007 or so [14:34] hmm [14:34] I'm actually building snaps and debs [14:34] what is that snaps thingy? something new canonical is trying to push? :D [14:35] * apollo13 has the feeling that we are ending up with more package formats than we need [14:35] ~apollo13 is recorded as having been created in July [14:35] cjwatson: yes, since I deleted the account last time the ubuntu forums got hacked or so :þ [14:35] and then recreated in july [14:36] I can upload my own debs correct? What would be the difference between that an packaging the python deps? [14:37] There is another problem, some of the pip requirements need to use wheel to compile headers for that specific arch [14:38] aatchison: not sure why you'd need wheel if you need binary stuff, just fetch the sdist and run bdist_ext or whatnot [14:38] So, I would have to make dsc for all of the pip requirements, and build each one for each arch? [14:39] Well, I know that I need to compile python bindings I mean [14:39] for at least one package [14:39] yes, but how does that have anything to do with wheel? [14:39] either that or update the upstream package (pocketsphinx) [14:39] roadmr: Looks like there's an old ~apollo13-deactivatedaccount hanging around that this is conflicting with [14:39] Created in 2005, deactivated in February [14:40] Wheel compiles locally during pip installation, unless I'm mistaken [14:40] Mismatching OpenID with this one of course [14:40] cjwatson: Bug #1607242 [14:40] bug 1607242 in Launchpad itself "Logging into a placeholder person OOPSes if email address already on another person" [Critical,Triaged] https://launchpad.net/bugs/1607242 [14:40] cjwatson: oh wow, didn't know that I already was writing ubuntuusers.de at that time -- good old times :d [14:40] wgrant: Ah yes, now I remember you filing that [14:40] * cjwatson reads [14:42] wgrant: In this case A is ~apollo13-deactivatedaccount and B is ~apollo13, right? [14:42] cjwatson: Correct. [14:45] wgrant: Do I then read it correctly that you suggest an admin merge of ~apollo13-deactivatedaccount to ~lp-void in this case? I'm not sure I see how that clears out the email address [14:46] cjwatson: Quite, the process was originally very dodgy and is now only quite dodgy: https://wiki.canonical.com/InformationInfrastructure/OSA/LPHowTo/DeleteAccounts?action=diff&rev2=3&rev1=2 [14:47] Grief [14:49] Do we need to check for multiple SSO accounts as well, then? [14:50] cjwatson: Hm, how is that relevant? [14:50] I don't know, but https://wiki.canonical.com/InformationInfrastructure/WebOps/LPHowTo/SSOandLPAccountWoe is linked from that page and suggests it [14:51] cjwatson: That is only a problem when there are multiple OpenID identifiers on a single LP account; both SSO accounts will advertise the same username. [14:52] So there are two main options here: fully delete the old LP account, or reactivate it and merge the new one into it. [14:53] apollo13: Do you just want the old account back, or are you intentionally aiming for a fresh start? [14:57] aatchison: you can only upload source packages to launchpad [14:58] aatchison: so you'd make appropriate source packages and just specify arch: any or arch: all on the binary packages as appropriate, and launchpad builders will build the binary packages on the appropriate archs in the ppa [15:22] cjwatson: I am fine either way -- whatever is easier [15:32] wgrant: I think we should probably merge the two, then. Procedure check: ~apollo13-deactivatedaccount/+reviewaccount -> Active, +adminpeoplemerge ~apollo13 into ~apollo13-deactivatedaccount, then once the merge completes rename ~apollo13-deactivatedaccount to ~apollo13? [15:49] @dobey thanks [15:50] This approach should work? https://github.com/spotify/dh-virtualenv/issues/112 [15:52] That way I could keep my virtualenvs and include the pip requirements in the same source package I think... === JanC is now known as Guest6029 === JanC_ is now known as JanC === mwhudson is now known as Guest72628 === nesthib is now known as Guest68734 === Guest4608 is now known as pcw === pcw is now known as PaulW2U === smoser` is now known as smoser === tsimonq2alt is now known as tsimonq2 === Guest605 is now known as caraka === Guest72628 is now known as mwhudson === acheronuk_ is now known as acheron === acheron is now known as Guest1658 [22:31] 'ello! anyone around? [22:31] getting some oops [22:36] jose: be more specific? [22:38] Launchpad oops [22:38] what pages? [22:38] any page that isn't being mean? [22:43] when filing a bug, just tried with someone else and it's just oopsing [22:43] someone over at IS should be able to take a look at the oops report === hloeung is now known as Guest86088 === Guest86088 is now known as hloeung [22:46] jose: you need to tell us the oops id [22:46] jose: we can see them but not telepathically [22:46] cjwatson: OOPS-717ad70cf1704bd8a887098f992bd093 is the first one [22:46] https://oops.canonical.com/?oopsid=OOPS-717ad70cf1704bd8a887098f992bd093 [22:47] IllegalTarget: Package rlec not published in Juju Charms Collection [22:47] you can't file bugs against non-official branches [22:47] let me try again, I had another one but closed the window [22:47] just one quick sec [22:48] hmm, its not timing out anymore. looks like it's fixed - thanks though! [22:48] I changed it and tried to file it against the general project and it would oops as well [22:48] bug timeouts are usually transient [22:48] good :) [22:49] there's a known problem with triggers that ~always goes away in about ten minutes === signed8b_ is now known as signed8bit_Zzz