[09:57] <czajkowski> Aloha
[09:58] <czajkowski> mrevell: might be an idea to post http://twitter.com/#!/planetubuntu/status/75484018019278848 to loco contacts
[10:09] <mrevell> czajkowski, /me looks
[10:12] <czajkowski> mrevell: morning btw :)
[11:11] <poolie> o/ mrevell
[11:14] <mrevell> Hey poolie
[12:32] <Philip5> anyone else who have problems with built packages on launchpad ppas to get published? have been staling in Pending publication mode for two hours here
[12:36] <soren> Philip5: Same here.
[12:39] <soren> halp! :)
[12:40] <soren> bigjools: Is this something you can look at or is there someone else I can poke?
[12:40] <bigjools> I can take a quick look
[12:40] <bigjools> ah, adeuring can take a quick look :)
[12:41] <soren> Awesome timing there :)
[12:41] <bigjools> I suspect he'll need my help :)
[12:41] <adeuring> bigjools: I'm afraid that IO need too help in this case ;)
[12:41] <adeuring> bigjools: so... could you have a llok?
[12:41] <bigjools> yes I was referring to you :)
[12:44] <ondrej> Hi, is dinstall on PPAs broken?
[12:45] <bigjools> adeuring: can you manage this as an incident please?
[12:45] <ondrej> I have several packages in pending state for some hours.
[12:45] <bigjools> ondrej: yeah we just found out that it's broken
[12:45] <adeuring> bigjools: ok
[12:47] <nigelb> james_w: you about?
[12:50] <nigelb> james_w: When I suggested using the LP API, instead of screen scraping for getting information for summit, you said LP will not be able to handle the load.  Would that happen even if they were splint into individual requests?
[13:06] <bigjools> PPA publishing should be ok again
[13:11] <soren> bigjools: Is it still catching up, or does it constitute an actual problem if I still have pending packages?
[13:11] <bigjools> soren: it will have a lot of catching up
[13:12] <soren> Ah, there it goes.
[13:12] <soren> \o/
[13:12] <soren> bigjools: Thanks for the quick fix!
[13:12] <bigjools> it was a deployment issue, ho hum
[13:26] <james_w> nigelb, I said that it would be very slow, not that LP wouldn't handle the load
[13:27] <nigelb> james_w: Ah, apologies. My memory was vague on what you said :)
[13:27] <james_w> that's ok
[13:28] <james_w> instead of ~1 roundtrip to get most of the info, it will likely do several round trips
[13:28] <james_w> maybe a*n where n is the number of blueprints accepted for UDS
[14:11] <nigelb> james_w: well, we already do a*n+1 + m where n is number of BPs and m is number of participants
[14:12] <james_w> nigelb, well, it may be even more
[14:12] <james_w> it's hard to say given that the API can't give us that information currently
[14:12] <nigelb> james_w: hang on, let me try and explain what I'm saying :)
[14:15] <nigelb> james_w: So, what I'm talking about is this http://bazaar.launchpad.net/~summit-hackers/summit/trunk/view/head:/summit/schedule/models/meetingmodel.py#L188
[14:15] <nigelb> Here we screen scape the data of each blueprint.  Isn't it more elegant to use the API? Because a change in LP UI might break our data import and we most probably won't know about it until it fails, aka the few weeks before UDS.
[14:16] <james_w> nigelb, you mean replacing line 210 etc?
[14:16] <james_w> I have no problem with that
[14:16] <james_w> I thought you meant replacing the XML file with the API, which is what I think will be slower
[14:16] <james_w> replacing what you point to should be roughly the same performance
[14:17] <nigelb> james_w: yup. that bit.  No XML file isn't screen-scraping, that's an export although LP folks don't really support it.
[14:17] <nigelb> *No, XML
[14:17] <nigelb> james_w: aha, I'll create a bug and start working on that.
[14:18] <james_w> great
[14:25] <nigelb> hrm, why doesn't launchpad give me a proper url to access the specification API?
[14:25] <nigelb> https://launchpad.net/+apidoc/1.0.html#specification tells me URL: https://api.launchpad.net/1.0
[14:27] <nigelb> adeuring: Hi, around?
[14:27] <adeuring> nigelb: sure
[14:27] <nigelb> Could you help me get data out of the specification end-point of the API?
[14:27] <adeuring> nigelb: I'll try. What is our problemß
[14:28] <adeuring> ...problem?
[14:28] <nigelb> heh
[14:28] <nigelb> I have a link to a specification, I'd like to figure out how to access the data in json
[14:28] <nigelb> Currently, summit screen-scrapes LP for this, I want to use the API to pull this data instead
[14:29] <adeuring> ah, now I see it... The URLs shows in the docs are simply broken...
[14:29] <adeuring> nigelb: the quick workaround would be: use launchpadlib
[14:30] <nigelb> adeuring: heh, that was my backup plan :)
[14:30] <adeuring> ;)
[14:30] <nigelb> adeuring: Is the doc thing something I can help fix/
[14:31] <adeuring> nigelb: that would be great -- problem is that I have no real clue how the API docs are generated...
[14:31] <wgrant> adeuring, nigelb: Nothing uses those URLs, and they are defined separately for each class in the XSLT.
[14:32] <nigelb> wgrant: how does the ajax stuff work if its not using the API?
[14:32] <wgrant> nigelb: It uses the API.
[14:32] <wgrant> nigelb: You wouldn't normally construct those URLs manually.
[14:32] <nigelb> wgrant: so, at uds flacoste showed me a magic url, I just didn't save it then :)
[14:32] <wgrant> nigelb: API paths are the same as webapp paths.
[14:33] <adeuring> nigelb: but regarding your "backup plan": you can write a small test script with launchpadlib, and if you have a statement like "print specification", you'll see the URL
[14:33] <wgrant> nigelb: But normally you'd get a spec using a method on the object that owns it.
[14:33] <wgrant> The AJAX Web UI just uses the current path.
[14:33] <nigelb> adeuring: w00t, that'll work
[14:33] <nigelb> wgrant: I can 'technically' construct the url for the API if I want to, right?
[14:34] <wgrant> nigelb: Yes, but it's not guaranteed to be stable.
[14:34] <wgrant> /ubuntu/+spec/whatever
[14:34] <nigelb> darn
[14:34] <wgrant> Same as the web UI.
[14:34] <wgrant> Note that this URL hasn't changed in probably 5 years, so it's not much of a risk :)
[14:34] <nigelb> heh
[14:35] <nigelb> https://blueprints.launchpad.net/summit/+spec/community-o-summit is my BP, so what would be the api endpoint for that spec?
[14:35] <wgrant> $APIROOT/summit/+spec/community-o-summit
[14:36] <wgrant> eg. https://api.launchpad.net/devel/summit/+spec/community-o-summit
[14:36] <nigelb> ah
[14:41] <nigelb> that's strange
[14:41] <nigelb> I don't see all the information I wished I could see
[14:42] <wgrant> eg?
[14:42] <nigelb> ah, needed to use devel instead of 1.0
[14:42] <wgrant> Those objects aren't exposed under 1.0, since we supposedly need to support that until 15.04
[14:43] <nigelb> ah.
[14:43] <nigelb> hrm, what format is that.  That isn't JSON.
[14:44] <nigelb> ....
[14:44] <nigelb> I get XML :\
[14:44] <wgrant> Add ?ws.accept=application/json. It will serve HTML to things that say they can accept it.
[14:45] <wgrant> (it defaults to checking the Accepts header, but ws.accept in the query string overrides that detection)
[14:45] <nigelb> ah, this is beautiful
[14:47] <nigelb> Thanks wgrant, adeuring :)
[15:49] <deryck> adeuring, I should be IRC now.  Sorry I just remembered.
[15:50] <adeuring> deryck: no problem
[15:50] <pmatulis> mneptok: o/
[17:46] <ckozak> Is there any way to get top contributors to a project using the launchpadlib api?
[18:00] <ckozak> is there a seperate channel for the launchpadlib api?
[18:07] <ckozak> Is there any way to get top contributors to a project using the launchpadlib api?
[18:20] <abentley> ckozak: This is a good channel to ask.  There is also launchpad-dev.
[18:21] <ckozak> abently: thanks!
[18:21] <abentley> ckozak: I don't see an obvious way to get that information over the API.
[18:22] <ckozak> abentley: hmm nor do I.  I guess I'll have to figure it out some other way
[18:24] <abentley> ckozak: It seems the method we use is not currently exported over the web service: getTopContributors
[18:26] <abentley> deryck: I relieve you.
[18:27] <deryck> abentley, thank you.  I can swap you out time during your shift, if you like.
[18:27] <abentley> No problem.
[18:27] <ckozak> abentley: is there any way to access "getTopContributors" without scraping the webpage?
[18:28] <ckozak> I guess that wouldnt be the end of the world though
[18:28] <abentley> ckozak: No, but if you really want it, you can always send in a patch.
[18:29] <ckozak> abentley: will do, thanks man!
[23:03] <james_w> lazr.restfulclient.errors.HTTPError: HTTP Error 301: Moved Permanently