[00:54:29] importScriptURI is what I usually use. [00:54:35] Though it may be deprecated. [00:54:53] Nemo_bis: ^ [06:28:35] Elsie: yes, me too [06:57:11] lle1 [07:13:28] Ryan_Lane: is morebots ready for identi.ca migration? http://identi.ca/doc/pumpio https://github.com/e14n/pump.io/blob/master/API.md [07:13:58] It was supposed to happen on the 8th and I don't see any change yet, so I guess it will happen suddenly. [07:18:09] Nemo_bis: morebots talks to twitter [07:18:27] the identi.ca account federates [07:19:06] Ryan_Lane: what do you mean federates? [07:19:22] when you post something on twitter it posts it on identi.ca [07:19:30] twitter doesn't post to identi.ca [07:19:39] statuses say "from python-statusnet" [07:19:53] identi.ca will pull posts from twitter [07:20:26] HaeB told me just a week ago this is not possible, but let's check [07:20:33] unless I'm thinking backwards [07:20:48] it's been a while since I've looked at that code [07:21:03] yes, there is no such thing [07:21:14] morebots posts to identica, I think. [07:21:15] identi.ca posts automatically to Twitter if configured, but not the contrary [07:21:15] I am a logbot running on wikitech-static. [07:21:15] Messages are logged to wikitech.wikimedia.org/wiki/Server_Admin_Log. [07:21:15] To log a message, type !log . [07:21:23] bleh [07:21:28] seems it posts to identi.ca [07:21:28] I was just looking at that code earlier [07:21:41] I can fix it; I have an open patch anyway [07:21:42] which means identi.ca posts it to twitter [07:21:49] cool [07:22:01] why do we even bother with identi.ca anymore? [07:22:06] I pinged them at https://github.com/e14n/pump.io/issues/151 [07:22:23] because most of the peaople actually reading this stuff are there? :) [07:22:41] i'd be incredibly shocked if that was true [07:22:58] it's basically dead [07:23:00] it is, just look at the followers [07:23:22] yes, identi.ca folks usually just want to follow some topics easily, which is impossible on twitter [07:23:49] https://twitter.com/wikimediatech [07:23:53] http://identi.ca/wikimediatech [07:24:01] 85 on identi.ca [07:24:02] the number of followers is a poor approximation of the number of people actually watching it, Nemo_bis [07:24:08] 599 on twitter [07:24:47] what I'm saying is that we should just post directly to twitter [07:24:55] ori-l: in fact I'm not talking of number [07:25:58] it's been added into 3 groups [07:26:38] compared to: https://twitter.com/wikimediatech/lists/memberships [07:26:45] yep, with groups it's easy to reach the few hundreds followers of !Wikimedia and !Wikipedia [07:26:55] lists are useles [07:27:09] 57 lists [07:27:18] almost as many followers on identica [07:27:24] "The only form of authentication allowed for the activity outbox is OAuth; the user must authorize the client before new activities are created." [07:27:27] of course, those are random lists [07:27:27] * ori-l abandons ship. [07:27:40] ori-l: I'd say let's just post directly to twitter [07:28:11] is twitter's api simpler? i haven't used it in a few years [07:28:12] * ori-l checks [07:28:22] there should be python apis [07:28:26] and it should work with passwords [07:29:07] https://dev.twitter.com/docs/api tells you that the API is RESTful no fewer than 5 times [07:29:11] in case you were super, super worried [07:29:24] ori-l: is it xml-rpc? [07:29:48] I'm a little worried it may be soap [07:30:57] REST! REST! REST! [07:31:15] I have nothing against identi.ca, but I don't see much benefit in using it if we need to start from scratch and it's a pain in the ass [07:31:39] especially if it won't continue to post to twitter for us [07:33:05] hm, from what do you infer this? [07:33:12] it's oauth, too. [07:34:06] ori-l: is there a python library for the new identi.ca? [07:34:29] there's a number of twitter implementations already around. it should abstract most of the cruft [07:34:30] Ryan_Lane: i'm inclined to agree with you re: twitter [07:34:53] ubuntu precise has two libraries [07:35:00] python-twitter and python-twyt [07:36:03] identica's new implementation is pump.io? [07:36:19] it looks like it, though i haven't heard of pump.io before [07:36:38] I don't see any packaged libraries for it [07:36:40] i would have guessed it was a porn site [07:36:43] I already linked https://github.com/e14n/pump.io/issues/151 [07:36:46] :D [07:37:03] 'pypump' [07:37:22] https://github.com/xray7224/PyPump/blob/master/README.md [07:37:44] it doesn't say REST anywhere, so it must suck [07:37:47] heh. python 3, eh? :) [07:38:08] Requirements: My pyoauth3000 repo (drop it into the PyPump folder for now) [07:38:15] sounds legit [07:38:26] ori-l: curl | sudo bash [07:38:43] don't be scared, just run it. :) [07:40:11] Nemo_bis: what's your case for preserving identi.ca integration? [07:40:51] sorry, I don't mind about making my case, ask on wikitech or werdna; now I'll brush my teeth :) [07:41:08] i don't like ragging on an open platform, but the migration plan you linked to above is a bit scarce on details -- it says existing users will be migrated automatically, but doesn't specify when the API switchover happens [07:45:43] ok. I need to go to bed [07:45:45] * Ryan_Lane waves [08:09:40] MaxSem: hey, i have another question about GeoData if you have a few minutes [08:09:46] sure [08:13:26] ok so in the documentation for GeoData it uses the term "dim" which as I understand is basically how big the object is. is there a scale somewhere on how to convert this to degrees? I'm basically trying to convert the dim from the API into the precision system that Wikidata uses (https://www.wikidata.org/wiki/Q4115189#footer) [08:28:34] legoktm, basically dim is size in meters. therefore, its translation to degrees depends on latitude [08:29:30] ah. so this is going to require some math... [09:24:02] "Deploy Echo on international projects this summer", anyone knowing what "international projects" are? [09:24:58] Nemo_bis: a slightly obnoxious way of saying "wikis with a primary language other than english" most likely [09:25:33] Ah. I would have said the opposite. [09:25:51] very highly obnoxious? :P [09:27:10] https://www.mediawiki.org/wiki/User_talk:Fabrice_Florin_%28WMF%29#International_projects [09:27:33] i like '0.0 idea' [09:27:40] i may use that myself :P [09:28:38] Yes, I wanted some additional precision. [09:28:58] Especially now that TimStarling is depriving us of sufficient&necessary conditions. :P [16:53:55] Wikimedia Language Engineering team office hours in #wikimedia-office in 5 minutes. [18:53:42] greg-g: fyi, just checked my mail. there were in fact no replies to asaf asking about the calendar [18:53:57] * jeremyb runs away [18:54:41] :) [19:01:46] Is there any way that I could bribe anyone to review this change: https://gerrit.wikimedia.org/r/#/c/67367 ? [19:05:02] I have lots of nice things to offer: candy from Hamburg, a pack of "Career Values" cards, and several Forever stamps! [19:09:20] your immortal sould will do [19:09:34] what about my firstborn child? [19:09:52] well? [19:10:18] see, he's ready to go :) [19:10:37] http://www.youtube.com/watch?v=MtaxKNaEAns [19:11:45] Time to drown my sorrows in a bottle of lunch [19:11:51] kaldari: is there a way to... [19:12:18] opt-out of confirmation with js? [19:12:43] someone could write a user-script to do it pretty easily [19:12:50] this is technical_13 btw [19:13:00] but right now, it can only be opted-out per wiki [19:14:11] personally, I'm not a fan of the confirmation myself [19:14:18] if I could add the functionality to opt out of confirmation step to my NoThanks.js I would be happy with that. [19:14:22] even though I wrote it :P [19:14:55] also I'm not getting talk page noticey from your top script anymore [19:15:13] might be something in my .js though. [19:15:28] $( 'a.mw-thanks-thank-link' ).click( function( e ) { [19:15:28] var $thankLink = $( this ); [19:15:29] e.preventDefault(); [19:15:29] sendThanks( $thankLink ); [19:15:29] } ); [19:15:37] that should take care of it [19:16:20] and of course you can get ride of the var $thankLink part :) [19:16:26] I'm on my mobile. can you leave thst on my tslk? [19:16:53] sure [19:17:14] thanks [19:20:43] kaldari: about the "Thanks" dialog - have you guys considered something like i described in https://bugzilla.wikimedia.org/show_bug.cgi?id=47782#c5 about the rollback link? [19:22:05] MatmaRex: sounds like a good idea to me [19:22:37] although having the confirmation inline is problematic for i18n [19:22:51] would be better to use a dialog, IMO [19:23:07] dialogs are scary and disrupting [19:23:19] (and modal, brr.) [19:23:52] Yeah, I think we need a general solution that is something in between [19:24:07] like a tooltip that you can interact with [19:24:19] anyway, off to lunch... [19:24:28] kaldari: hm, how would it be problematic for i18n? should be simple enough [19:24:41] alright, lunch has higher priority :P [19:25:08] :) [20:59:19] gn8 folks [21:03:11] MaxSem: Thanks!!11 [21:03:25] :)