[15:31:26] o/ [15:31:43] good mornin' o/ [17:28:52] harej: http://librarybase.wmflabs.org/ is pretty nifty. [17:29:33] :> [17:29:43] I want it to be a big collaborative project! Unfortunately I'm at work right now. [17:34:59] It's really cool and I'm glad you have a proof of concept already. [18:04:40] halfak: you joining meeting today? [18:04:44] devops researcher? [18:07:55] ottomata, can't [18:08:00] debugging an outage. [18:08:05] Not much to sync on anyway [18:14:30] hey halfak, do you know who to contact if I want to join the Wikimedia organization on GitHub? [18:15:29] Oh! I can add you [18:15:31] Username? [18:16:08] jtmorgan [18:16:24] thanks, halfak [18:17:22] {{done}} [18:39:45] halfak: np [19:16:12] halfak, so your geeking out over the intertimes with search.... ;p [19:26:45] Ironholds, yeah? [19:26:53] :D [19:27:21] log( \___ ) = /\_/\_ [19:27:22] halfak, and you know how we have that task for yinz to collaborate with us on digging into this data...? ;p [19:27:37] mathemojics, I like it [19:28:06] Indeed :D [19:28:18] We should consider aiming for a CHI submission. [19:28:23] CHI would love the session stuff. [19:28:34] Sept. 29th deadline [19:30:01] Mathemojics. I <3 the concept and the name. [19:30:39] halfak, just that we saw it somewhere else? [19:31:12] Na. I think we can do better. E.g. thinking about searching behavior. [19:31:41] halfak, agreed, I'm not sure if we can do better than by 29th though [19:31:49] I gotta get this dataset available for my paper with J-Mo! [19:32:20] I was more thinking on a short-term collaboration within the org to validate this dataset and identify a breakpoint properly - which can feed into a paper-based project, of course, but I won't be free in 2015 :( [19:32:40] No worries. [19:32:45] CHI would be a rush anyway. [19:32:47] what paper, Ironholds? [19:32:51] I've already got two in the pipe. [19:33:01] J-Mo, talkcicity? [19:33:08] halfak, ooh get you :D [19:33:23] ah yes. the one that I kicking down the road… :/ [19:33:29] J-Mo, Frances and I (mostly Frances) hath debugged the script but I ran into one weird-ass error I need to debug this evening [19:33:38] it's okay, there's nothing for you ivory tower academics to do until the code works [19:34:01] you leave it to us code-miners, toiling away all day, *more moaning I can't think of right now* :P [19:34:05] so much research [19:34:49] * J-Mo looks back fondly on his days as an ivory tower academic… until he remembers grad student wages & hours [19:34:57] Ironholds, I didn't know the project was moving forward. [19:35:09] The talk page conversation extraction and toxicity labeling? [19:35:29] halfak, yeah, Amanda (bless her cotten socks. Or whatever she makes her socks out of, I'm not a sock elitist) pushed us forward on it [19:35:35] J-Mo, ahh, but the health insurance! [19:36:49] Oh well. Godspeed then [19:36:58] halfak, we'll see? This code is... [19:37:08] fully academia-compliant [19:37:42] but on the plus side, worst-case scenario we write a paper on the toxicity and unpleasantness of having to deal with a codebase full of lines opening "# TODO:" [19:38:11] Ironholds, switched to someone elses code? [19:38:16] who's working on talk page toxicity? [19:38:26] We were working on that talk page parser. [19:38:35] I was hoping to use wikilabels to label the dataset. [19:38:50] halfak, we were but I didn't know Python well enough to do the work (read; any python, honestly) [19:39:06] wikilabels? [19:39:16] The hand coding system that runs in wikipedia [19:39:27] https://en.wikipedia.org/wiki/Wikipedia:Labels [19:39:28] ahhh! [19:39:56] I read that as "the hand-coding system that runs Wikipedia" :p [19:40:23] I mean Wikipedia is basically a mass of subjectivity in the hopes that objectivity comes out of it with enough coders, so ;p [19:46:48] ^^ whoa… so deep, Ironholds [19:46:58] hah [19:51:35] https://meta.wikimedia.org/w/index.php?title=Meta:Requests_for_comment/Enable_flow_in_the_Research_talk_(203)_namespace&curid=8549460&diff=13563874&oldid=13560835 [19:51:37] sign [19:51:40] sigh* [19:51:56] Why can't people make rational, good-faith arguments. [19:54:43] guillom, ask a bad-faith question, get a grumpy answer [19:55:17] oi. no trust anymore. [19:59:03] now the RfC is even more partisan: mostly Metapedians who don't participate in the Research namespace vs. WMF staff who do. Very little overlap, other than Nemo_bis. We really need to get more external researchers to participate, otherwise I don't see how the results of this RfC can be seen as valid by anyone. [19:59:33] Or we could let Nemo_bis decide :D [19:59:41] (No I don't believe that) [19:59:47] * halfak shakes sticks at Nemo [19:59:59] Anyway I think you're right J-Mo [20:00:12] It's very frustrating that the staff/community divide is playing out here. [20:02:01] I really would like some more specificity in the concerns raised by the opposed [20:02:15] It would be great if we could turn this RFC into a to-do list for the Flow team. [20:02:52] Before you deprioritize, we need 1,2,3,... in order to use Flow. [20:02:57] Something like that. [20:04:03] WSC's comments are "it might be too slow for someone somewhere" [20:04:13] MariaDB [metawiki_p]> select count(distinct(rev_user)) from revision join page on rev_page = page_id and page_namespace = 203; >>> 637 [20:04:20] Plenty of opportunity for participation. ;) [20:04:41] J-Mo1, got a proposal for how we pull in more people? [20:05:19] follow-up email to the research-l list? [20:06:16] I'm not sure getting more people to participate will help (either way). It might just be bad timing because feelings are raw after the recent (not-very-clear and widely-misunderstood) announcement. It might be best to just withdraw and try again in a few months. [20:06:40] Whatever the outcome, it feels like it'll be a lose-lose situation. [20:09:17] unrelated... [20:09:43] halfak: J-Mo1 guillom https://phabricator.wikimedia.org/T109715 asks for ElasticSearch (Aka full text) replicas of current page contents onto labs. If any of you know of good use cases that that will enable, please comment [20:10:12] Current revision only? [20:11:20] halfak: yes [20:11:29] hmm [20:11:32] Ironholds: Nemo_bis ^ as well [20:11:39] leila: legoktm ^ too :) [20:11:51] it's elasticsearch so you can query it fairly nicely [20:12:31] For my uses cases I either need all revisions or none; just the last one doesn't really help. [20:13:11] YuviPanda, ping shilad in #wikimedia-ai [20:13:18] He does a lot of stuff with the current rev,. [20:13:21] YuviPanda: https://phabricator.wikimedia.org/T109715#1559111 ? [20:17:29] YuviPanda: mwgrep for everyone! [22:09:18] * guillom removes "Meta:Requests for comment/Enable flow in the Research talk (203) namespace" from his watchlist. [22:09:36] Too much dramah. [22:10:35] New drama? [22:10:49] Theo came. [22:10:53] So, yes. [22:41:51] guillom, I'm sorry if I set him off [22:42:12] Ironholds: I'm not sure you can take credit for this one :p [22:42:51] heh [22:43:19] I often wonder if our project is 'worth it' [22:43:33] in the sense of, there is so much toxicity, is it really worth me spending my time on it? [22:43:37] * YuviPanda shall continue to wonder [22:43:39] Worth the abuse and the bad faith and the trolling? [22:43:51] yes [22:44:50] That's humanity. [22:45:35] Our projects aren't better or worse. We have beauty and selflessess as well as ugliness and pettiness. [22:45:39] indeed, and I sometimes wonder if humanity is worth it at all [22:46:37] * YuviPanda goes back to instead looking at more important things, like the behavior of sysctl and abac in kubernetes [22:46:51] I think if we manage not to self-destruct in the next few centuries it'll be worth it, yes. [22:48:23] * guillom --> bus --> home. [22:48:48] Today has been a productive day. [22:54:34] YuviPanda, I was having that discussion with myself today [22:54:52] like, the last couple of days have reminded me very much why I have had so much fun burying myself in work and books and the R community [22:55:13] it's not the toxicity I hate, actually. I mean, I hate it, I hate the smarmy arrogant dissonant angry let-me-shit-on-this-work-and-cry-freedom-of-speech bullshit [22:55:33] it's that the more time I spend having poop thrown at me the easier it gets to justify throwing poop back. [22:56:06] I like not editing because I prefer the person I am when I don't have to deal with...all of that. [23:04:51] hey halfak [23:05:12] when can I quickly brief you on that thread with Lila/Trevor? I have time now or I can maybe block 30 minutes tomorrow? [23:05:46] Tomorrow would be best. [23:05:52] I still need to head home and it's kinda late. [23:06:35] sounds good [23:18:45] * halfak gets on bike [23:18:49] have a good evening!