=== tasdomas_afk is now known as tasdomas | ||
=== tsimpson_ is now known as tsimpson | ||
=== tasdomas is now known as tasdomas_afk | ||
tsimpson | does anyone know how to authenticate a session to launchpad, via launchpadlib, for a website? the instructions on https://help.launchpad.net/API/launchpadlib#Authenticated_access_for_website_integration don't seem to work | 12:14 |
---|---|---|
wgrant | tsimpson: What's not working? The credential API may have changed a little since those docs were written. | 12:16 |
tsimpson | wgrant: when I try the line `launchpad = Launchpad(credentials, service_root="production")` I get a type error: __init__() takes at least 4 arguments (3 given) | 12:18 |
wgrant | def __init__(self, credentials, authorization_engine, | 12:19 |
wgrant | credential_store, service_root=uris.STAGING_SERVICE_ROOT, | 12:19 |
tsimpson | exactly | 12:19 |
tsimpson | so the documentation seem outdated, and I don't know what I should be putting for the other parameters | 12:20 |
wgrant | Have you checked how other callsites use it? | 12:20 |
tsimpson | not really, this is an old script, probably written around the last time that page was edited | 12:21 |
tsimpson | I just noticed that it's not working (because we never really used it too much) | 12:22 |
tsimpson | I'm hoping not to have to make a whole rewrite to pull a list of teams from Launchapd for authentication for teams | 12:22 |
wgrant | tsimpson: Why are you not using Launchpad.login_with? Do you have a custom credentials file? | 12:24 |
tsimpson | wgrant: because the code existed before .login_with did | 12:24 |
wgrant | Right, but can you just port it to login_with and be done? | 12:25 |
tsimpson | and because the guide in the wiki doesn't say one should | 12:25 |
wgrant | That's usually the easiest. | 12:25 |
wgrant | tsimpson: What about the section immediately before the one you linked? | 12:25 |
wgrant | You're not doing it for website integration, are you? | 12:25 |
wgrant | login_with is fine for most scripts | 12:26 |
tsimpson | all I need to do is pull a list of team memberships | 12:27 |
tsimpson | it's just a python CGI script | 12:27 |
wgrant | It's not so much a question of what the script does, more of how it's used. | 12:27 |
tsimpson | but it needs to know if someone is in certain lp teams | 12:27 |
wgrant | Hm, unless those teams are private it can probably do that anonymously. | 12:28 |
wgrant | Does it need auth at all? If not, login_anonymously is your friend. | 12:28 |
tsimpson | I'm not sure, I can't actually remember if this existed before anonymous login was supported | 12:29 |
tsimpson | team memberships should be public information, right? | 12:29 |
tsimpson | but I still need a way to know which person to look up team membership for, I need some kind of authentication from the launchpad side | 12:30 |
wgrant | Unless the team is private, yes. | 12:31 |
wgrant | Hm, most sites do that by using SSO for authentication, then asking Launchpad for the user corresponding to the logged in OpenID identifier. | 12:33 |
wgrant | https://lists.launchpad.net/launchpad-users/msg06239.html describes the changes to authentication | 12:33 |
wgrant | 4. Use login_with() instead of the Launchpad constructor. | 12:34 |
tsimpson | I was hoping not to need to rewrite the whole authentication stage | 12:34 |
wgrant | 5. If your script is not intended for interactive use, pass in a credentials_file argument to login_with() to avoid the keyring. | 12:34 |
wgrant | Are probably the relevant bits | 12:34 |
tsimpson | if that's the way to do it, I guess I have no choice but to just rewrite it :) | 12:34 |
wgrant | You certainly have to rework how your authentication interfaces with the launchpadlib API. | 12:35 |
wgrant | Because that interface has changed. | 12:35 |
wgrant | But the same workflow should be possible with a little work. | 12:35 |
tsimpson | yeah, it's just some work I hoped I wouldn't have to do | 12:35 |
=== cprov-afk is now known as cprov | ||
=== cprov is now known as cprov-brb | ||
ricotz | wgrant, hello, are you around by any chance? | 14:28 |
ricotz | wgrant, i am curious about the limitations of armhf ppa builds, it seems there are more and more daily-recipe building going on which obviously isnt going well with the amount of available armhf builders | 14:30 |
ricotz | just as an example for daily building with 9+ hours buildtime -- https://code.launchpad.net/~thopiekar/+recipe/cython-py2+3-daily -- https://launchpad.net/~cython-dev/+archive/master-ppa/+build/4972342 | 14:31 |
dobey | ricotz: yeah, and they are emulated builders, not real hardware. | 14:33 |
cjwatson | FWIW the number of builders isn't fixed, but we've had to keep it low for the past day or so because the x86 queues have been scary | 14:35 |
cjwatson | So most of the wanis have been on x86 duty | 14:36 |
ricotz | i see, i am asking because e.g. the elementaryOS daily ppa got the armhf support revoked for way less intrusive buildtimes | 14:38 |
=== cprov-brb is now known as cprov | ||
ricotz | but yeah, if there are more builders again the situation would be better | 14:39 |
ricotz | since currently this ppa blocks all available armhf builds for 9+ hours now | 14:40 |
ricotz | and it seems 14 hours are expected -- https://code.launchpad.net/~cython-dev/+archive/master-ppa/+build/4949216 | 14:41 |
cjwatson | I'll let wgrant reply as to the policy matter | 14:41 |
ricotz | thanks for taking notice | 14:41 |
cjwatson | I've thrown one extra builder over to armhf for now | 14:42 |
cjwatson | Really, though, people need to stop building everything for precise/quantal/raring/saucy ... | 14:43 |
ricotz | + lucid (in this case) | 14:44 |
=== cprov is now known as cprov-afk | ||
=== gkbhat is now known as bkg | ||
=== Ursinha is now known as Ursinha-afk | ||
=== cprov-afk is now known as cprov | ||
=== Ursinha-afk is now known as Ursinha | ||
me4oslav | Anyone in here good with reading buildlogs and founding the issues in them? | 18:21 |
=== BradCrittenden is now known as bac | ||
dobey | me4oslav: what log? | 18:26 |
me4oslav | dobey: https://launchpadlibrarian.net/150120363/buildlog.txt.gz | 18:34 |
me4oslav | I've had a similar error, but it was caused by cyrillic icon in the source, which is gone | 18:35 |
me4oslav | than one build was success | 18:35 |
me4oslav | and then this happened | 18:35 |
me4oslav | I checked the entire source for cyrillic | 18:36 |
me4oslav | not a trace | 18:36 |
dobey | me4oslav: looks like there's a zero width no-break space in a filename | 18:41 |
dobey | me4oslav: don't ask me how it got committed to bzr though | 18:41 |
me4oslav | dobey: so, what do I do? Here is the commit after which I got that error: http://bazaar.launchpad.net/~numix/numix-utouch-style/trunk/revision/51 | 18:45 |
dobey | me4oslav: i don't know. branch locally and try to find the offending file. then fix it? | 18:48 |
Soul-Sing | any staff here for a pm? | 18:55 |
me4oslav | dobey: right. So I downloaded the .diff and branched the branch. Should I look in the diff file for sth? | 19:14 |
dobey | me4oslav: compare the file names by their bytes, and not by the visible string | 19:17 |
me4oslav | dobey: right, did that and: http://pastebin.com/6CpzyBwZ (the file sizes are from Nautilus properties dialog) | 19:52 |
dobey | me4oslav: i don't mean bytes as in how much space they use on disk. i mean the actual bytes that make up the file name strings | 19:53 |
me4oslav | dobey: and you confused me. How to I do that? | 19:54 |
dobey | for instance "" <- that isn't an empty string, but a string that contains the \feff character (so it is 2 bytes instead of 0) | 19:54 |
dobey | me4oslav: find . -name "**" <- this should work the 0xfeff character is between the two * characters, there | 19:56 |
me4oslav | dobey: aha, here is the bugger: me4oslav@laptop:/media/Data/Main/Visuals/Icons$ find . -name "**" | 20:02 |
me4oslav | ./numix-utouch-style/Numix-uTouch/apps/48/org.gnome.Weather.Application.svg | 20:02 |
me4oslav | me4oslav@laptop:/media/Data/Main/Visuals/Icons$ | 20:02 |
dobey | me4oslav: great. rename it to the proper name without the non-breaking space and it should fix it then i guess | 20:04 |
dobey | it's probably failing on the builder because the builders are sadly still running 8.04 :( | 20:04 |
me4oslav | dobey: I deleted it, cos it didn't do any good. Thank you so much. | 20:06 |
MoonMaker | Hi! I have a question about my daily receipt: https://code.launchpad.net/~moonmaker/+recipe/twonkyserver-daily | 20:45 |
MoonMaker | I want to build 3 debs from one source. But it seems like that it isn't possible? | 20:46 |
MoonMaker | when I test it locally, then its working. But on launchpad I have no success. My control file is here: http://bazaar.launchpad.net/~moonmaker/+junk/twonkyserver/files/head:/debian/ | 20:46 |
MoonMaker | Does anyone has an idea? Thanks for help! | 20:47 |
MoonMaker | https://answers.launchpad.net/launchpad/+question/235686 <- I'm waiting for your hints ;-) | 20:55 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!