[00:01:22] o/ Helder [00:02:07] oi! [00:02:10] hey, I just noticed that nosetests are not passing on master of revscoring [00:02:16] * halfak checks [00:02:41] it was failing before I merged that patch [00:02:41] They all pass for me. [00:02:54] It could be that I have a weird version issue in my env. [00:02:58] Can you show me the error? [00:03:27] http://dpaste.com/0KPMPHK [00:03:51] Oh. Install myspell-fr :P [00:04:06] and myspell-he [00:04:19] And aspell-id [00:04:28] and myspell-fa [00:04:32] and myspell-es [00:04:38] I don't want to do that again actually... last time it created a lot of entries in the context menu of firefox when I was trying to check the spellings of what I write [00:04:41] and hunspell-vi [00:04:55] Oh. Then those tests will fail. [00:05:10] also, didn't something related to languages were made optional? [00:05:16] Recent changes made installing them optional for operating the rest of revscoring though :) [00:05:21] Yup [00:05:26] Test will fail if you don't have them [00:05:34] hmm [00:05:35] But you can just use pt, for example. [00:05:42] Or pt and en [00:06:05] Oh wait! I see some assertion errors in here. [00:06:17] Aha! upgrade your version of deltas. [00:06:27] pip install deltas --upgrade [00:06:32] what was the command to update everythin on pip ? [00:07:13] pip install -r requirements.txt --upgrade [00:07:32] * Helder does that [00:09:55] halfak, http://dpaste.com/0XSJQ83 [00:10:06] AttributeError: _ep_map [00:10:15] any ideas what that means? [00:10:41] No idea. Try running it again [00:13:27] ok, now it is processing scipy... [00:13:34] which usually takes longer [00:17:56] Yeah... [00:17:59] sadpy [00:19:01] and it goes on... [01:31:43] And there he fell. Installing scipy took too long. [01:52:32] halfak: hey, if you're around. I'm trying to learn advanced machine learning for a while and I just got to clustering and choosing best number of clusters, I found some exciting methods and I'm trying to learn them or examine them on our system to get best results [01:52:37] it will be done very soon [02:37:33] Amir1, a new clustering strategy? [02:58:01] halfak: no, clustering strategy won't change, it's the best we've got. It is about choosing number of clusters to have best results [03:01:02] Gotcha. Looking forward to it. :) [03:01:14] How's the work on features? [03:06:36] I'm getting feedbacks, collecting them and turning them to features [03:06:49] it'll be done by Monday [03:08:17] ! That'll be fast. [03:08:31] I'll be around over the weekend. [03:08:45] Gotta catch up on weekend work due to last week's vacation :S [03:10:10] awesome [03:10:12] :) [16:00:36] Hey folks. Probably skype today [16:02:29] * halfak waves harder [16:02:44] I'll hop in the hangout just in case. [16:03:03] Nope. no one there either. [16:03:08] o/////// [16:09:02] Amir1, ToAruShiroiNeko, awight? [16:11:20] * halfak goes back to work. [16:11:35] See my proposed agenda here: https://etherpad.wikimedia.org/p/revscoring [16:22:11] YuviPanda, I've been thinking about what mwdb should *be*. You celebrated the idea the other day. I was wondering what you had imagined when I brought it up. [16:23:28] I've been considering setting up SQLAlchemy for mediawiki's DB structure and just providing that as a library. [16:35:03] * halfak considers getting grumpy about the non-attendance of meetings. [19:18:25] halfak: Hey, I just got to my pc [19:18:43] My struggle to find a place to stay has started [19:18:51] from today [19:19:17] Bummer. A place to call home or a place to connect from? [19:20:02] a place to sleep [19:20:12] :/ Anything I can do to help? [19:20:13] until I get this month slaary [19:20:28] not so much :) [19:20:40] I appreciate your offer [19:20:41] :) [19:20:43] thanks [19:21:07] Any time you just happen to be in Minnesota, you can stay in my spare bedroom. [19:21:10] :) [19:22:04] oh, thanks. Same happens if you come to Iran (and I got my home) [19:22:32] Anyway, you weren't the only one to miss today. [19:22:48] ToAruShiroiNeko, had connection troubles and aetilley forgot what day it was. [19:23:36] we can have another one soon [19:23:51] I considering the value in getting upset about this, but instead, I think it's better to insist on make-up sessions. So I'm planning to work with aetilley before our hack session tomorrow. [19:23:54] and the thing is, we had one two days ago [19:24:11] Yeah, so it's not too critical, but it was on the schedule :/ [19:24:19] yeah [19:24:21] sorry [19:52:31] halfak: were you expecting me? I always presume I won't make it... [19:53:11] Oh no. [19:53:35] The only reason I pinged awight was because he responded "yes" [19:54:19] Ah OK :) [20:20:59] YuviPanda, if you're still around, I have a Q for you though. [20:21:19] Re. mwdb read the scrollback. [20:21:22] I'll copy paste [20:21:54] YuviPanda, I've been thinking about what mwdb should *be*. You celebrated the idea the other day. I was wondering what you had imagined when I brought it up. I've been considering setting up SQLAlchemy for mediawiki's DB structure and just providing that as a library. [20:45:26] Hmm... Looks like we might have some downtime. [20:45:31] for ORES. [20:49:20] woah... looks like the system is running super slow. [20:49:23] But it's up. [20:56:49] !log restarted ores-worker-01,02,03 because of bad performance. It seems the system is doing better now. [20:57:13] wm-bot4, are you logging my things? [21:07:28] somebody better be logging my things [21:08:15] Looks like nope [21:08:16] https://wikitech.wikimedia.org/wiki/Nova_Resource:Revscoring [22:19:55] halfak: you need to do !log in -labs [22:20:02] halfak: need any help?