[17:34:35] hi halfak. do you work today? :-) [17:34:43] o/ leila [17:34:51] yeah. Forgot it was a holiday and scheduled work :( [17:34:57] no worries. I work, too. [17:35:01] You can use it, later. [17:35:15] Can you put your abstract in https://www.mediawiki.org/wiki/Wikimedia_Research/Showcase#January_2016? [17:35:26] leila, Yeah! Will do. [17:35:33] I'd like to send out the announcement today, since Tuesday will be too short of a notice. Thanks, halfak. [17:52:29] leila, {{done}} [17:52:38] great, thanks, halfak. [18:40:26] Ironholds: yt? [21:35:16] https://meta.wikimedia.org/wiki/Research_talk:Measuring_edit_productivity/Work_log/2016-01-18 [21:35:19] ^ For the curious [21:35:56] Looks like bots and tool-assisted editing does not account for the widening disparity between anon editor and registered editor productivity on English Wikipedia [22:02:02] Just made some updates. The analysis now includes sample plots for individual users. [22:49:23] YuviPanda, halfak: any commentary on the good, bad and ugly of WikiChatter? Looking to use it to parse general talkpage noise; is it there yet? [22:49:48] IH|annotating: I'd say do it and file bugs. [22:50:03] haha, okay! [22:50:15] IH|annotating: it can parse all of TeaHouse except 12 questions [22:50:20] so that's 12 out of about 12.5k [22:50:22] which isn't bad [22:50:58] IH|annotating, FYI, ellery is planning to work on releasing a dataset of talk page discussions this quarter. [22:51:05] It seems like you might be working in the same direction. [22:51:32] halfak, oh cool. I haven't seen him mention it on any of the mailing lists [22:51:38] YuviPanda, yeah, that's pretty great [22:52:05] IH|annotating, yeah, I think that's yet to begin because we're still helping FR make the transition to "No more Ellery" [22:52:15] But there's already time set asside. [22:52:31] cool! [22:52:38] (followup; is it python3 only? ;p) [22:53:05] it better be! [22:53:07] :P [22:53:11] Yes. [22:53:21] Python3 is the present [22:53:28] It's also the near past! [22:53:35] And the forseeable future! [22:54:57] yes yes ;p [22:57:26] * IH|annotating submits a pull request [22:57:31] hey look, I contributed to a Python module [22:59:35] * halfak reverts Ironholds [22:59:41] Learn the rules or GTFO! [22:59:50] OR we could be nice. [22:59:52] Link to PR? [23:03:55] https://github.com/kjschiroo/WikiChatter/pull/20 [23:04:09] 'import X' instead of 'from X import Y' is kind of important [23:04:43] the former is confusing and leads to people going 'but that didn't work' and source-hunting ;p [23:07:10] halfak, did you write the 'get wikimarkup out of the API in Python' thingy? [23:09:09] IH|annotating, I didn't. [23:10:14] IH|annotating, wait... I don't see one of those in WikiChatter. [23:10:21] Are you talking generally about 'mwapi'? [23:10:35] http://pythonhosted.org/mwapi [23:10:57] o/ kjschiroo [23:11:05] IH|annotating is just picking up WikiChatter [23:11:09] \o [23:11:16] What is that? [23:11:26] :P [23:11:35] https://github.com/kjschiroo/WikiChatter [23:12:04] Oh! "IH|annotating" is Ironholds with some weird stuff in his name [23:12:12] Oh! [23:12:15] okay, i get it [23:12:28] I saw that pull request [23:12:48] I don't think that it is totally how it works right now, but I think it should work that way [23:13:05] kjschiroo, Also FYI, you may be interested in https://meta.wikimedia.org/wiki/Research_talk:Measuring_edit_productivity/Work_log/2016-01-18 [23:13:21] To get that behavior don't i need to `from .talkpageparser import parse` in __init__ [23:13:24] ? [23:14:11] "from wikichatter import talkpageparser as tpp" [23:14:20] That should work with the current setup [23:15:00] Yeah, that would also do it. [23:15:10] You could have "import wikichatter;wikichatter.parse(...)" if you imported parse inside __init__.py [23:15:30] Yeah, that is what I like about it. [23:17:43] halfak, ta! [23:18:26] That link is interesting. Seems like work is getting more professional [23:18:54] * halfak runs away. [23:19:01] Have a good one folks! [23:28:35] okay, best discovery of the day, courtesy of fhocutt; due to a unicode error the chi symbol in this paper was replaced by the printer with a penguin [23:28:39] the penguin2 test. [23:43:12] halfak, also, yes, I'm doing paper review at the mo [23:43:56] I thought I wouldn't work today [23:43:59] apparently I am