[18:24:54] Who's ready for some research? [18:25:16] is there a youtube linky yet? [18:25:30] Not yet, I believe [18:25:41] alternatively, do the speakers have slides someplace so we can follow along that way? cause... it starts in 5, right? [18:30:33] can someone put the link in topic please [18:31:10] Yes please, links are good [18:31:58] is youtube link for the meeting available? [18:32:01] Woohoo, showcase party! [18:32:11] http://youtu.be/-We4GZbH3Iw [18:32:21] Hey everyone! [18:32:22] Thanks EdSaperia__ [18:32:25] EdSaperia__: thanks! [18:33:19] EdSaperia__, thanks! [18:34:28] EdSaperia__: youtube says "coming soon", did the meeting start? [18:34:52] No idea! [18:34:58] I just got sent ye link! [18:35:33] it hasn't started yet [18:35:43] once it starts, the Youtube link should show the actual streaming [18:35:43] leila: thanks [18:35:52] I'll let y'all know so you know when to expect it [18:35:57] np, zeljkof [18:36:42] Totally stoked [18:37:38] okay, it's live now [18:38:25] yay [18:38:38] Yay it's working! [18:38:38] * heatherw waves [18:38:42] All my favourite people :D [18:39:33] you will hear shortly Dario's speech. Here is the link he is referring to: https://boards.greenhouse.io/wikimedia/jobs/32299 [18:42:18] there is some echo, I guess from the office [18:42:32] maybe the office should mute while somebody else is speaking [18:42:46] better now [18:42:46] zeljkof: we muted now [18:42:48] I just did [18:42:55] link to full paper: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0104880 [18:43:06] DarTar, leila: thanks; i think the sound is better now [18:43:13] DarTar, is the setting on 150? cause that should help with with sound quality [18:44:12] DarTar, I know you may not want to mess with it now. ;-) [18:46:16] what's the API call? [18:47:23] are the dimensions assumed to be orthogonal? [18:48:31] notconfusing, if you specify a specific sldie/topic of presentation, I can ask it at the end. [18:48:49] There is a bit delay in the streaming, so I'm not sure which slide you're referring to. :-\ [18:49:05] There are many 2nd language speakers on wp:en with more limited idiom. Does that influence results, e.g. ANEW scores? [18:49:22] i'll ask David ezachte [18:49:28] slide 17 - do any of the measures account for sarcasm or unexpected use of language? [18:50:09] Amanda_: I made a note of your question [18:51:07] thanks +leila! also, it would be nice to see some more nuanced treatment of sex (e.g., male and female) vs. gender (i.e., socially constructed gender identities that aren't available when one creates a user account) [18:51:38] I'll make the comment [18:51:56] Amanda_, what do you think that might bring to the analysis? [18:52:29] sorry guys, what's the presenter's name? [18:52:37] David [18:52:41] thanks [18:52:53] FYI: https://www.mediawiki.org/wiki/User:Sdivad [18:53:03] cool [18:53:45] spooky, I was reading his reader behaviour paper yesterday. [18:54:00] heh. Academia is a small world. [18:54:11] https://en.wikipedia.org/wiki/Small-world_network\ [18:54:21] Bah. -\ [18:54:22] https://en.wikipedia.org/wiki/Small-world_network [18:54:56] +halfak, i'm finding in my own work that addressing all women on wikipedia as "female" and not allowing for some of the nuances of gender can actually further alienate editors. semantically, there is, in fact, a difference between the meaning of sex and gender. [18:55:51] Amanda_, +1 I agree that the UI might consider a more nuanced and *real* version of gender, but I'm wondering if this is a critique of this analysis or not. [18:56:00] The word 'terrific' can be interpreted as both positive and negative. Are such cases taken into account, anyone? [18:56:09] leila, ^ [18:56:27] (making sure she sees the question) [18:56:43] I'll ask bmansuro_ [18:56:44] +halfak, i'm still waiting to hear if this analysis is about gender of sex. the title says "gender". i admit: i haven't had time to read the paper! [18:56:44] leila: ^ (double making sure) [18:56:49] thanks [18:56:54] :D [18:57:37] gosh, I have a ton of questions I’d like to ask [18:57:47] Post em so we can think about 'em too. [18:57:54] ok, [18:57:59] Amanda_: I'll try to capture your question but you may want to follow up with David [18:58:11] there is also a cultural difference apart from depth of idiom, for example US use 'exciting and thrilled' all the time where Europeans would be feel similarly positive and say 'nice and glad' [18:58:12] since having one degree of separation can make it hard to communicate [18:58:37] interesting [18:58:43] ezachte, made a note of it. [18:58:48] ezachte, and a midwesterner would say "not totally awful" [18:58:56] 1. the example he mentioned before (with words such as “abuse” and “protest”) was illuminating: these were discussions about the *topic*, I wonder how we can tell apart topic discussions vs meta discussions using this method [18:59:03] halfak :-) [18:59:33] we need a follow up -- this is fascinating [19:00:00] I missed out how he identified "males" and "females". [19:00:07] Can anyone tell me? [19:00:30] self-identified, halfak [19:00:37] With the user preference? [19:00:39] halfak 1. self-identification on user pages and 2. crowd-sourced [19:00:45] Or template? [19:00:49] i.e., editors specify their gender. the data is from there [19:00:56] :P [19:00:58] I suppose that this is actually about gender and not sex, because it's about the person's declared gender identity s opposed to their biology ( Amanda_ ) [19:01:09] I missed the crowdsourced part, fafloe. [19:01:26] 2. there are emotional values associated with the use technical jargon (WIkipedia policies, acronyms etc), I wonder if the analysis could be finetuned by using a handcoded corpus for words that are not plain English, but that are prevalent on talk pages [19:01:32] user preference, halfak in the first instance (got 2k cases that way) [19:01:47] Indeed. How was gender "crowdsouced"? [19:01:53] and by lookking at the user page for references/info that point to the gender [19:01:58] they gave the users talk page comments to the crowd as far as I understood halfak, leila [19:02:13] but this isn't gender; it's sex [19:02:22] 3. why a 2010 dump? [19:02:50] Amanda_: none of us know what gentials ny of these people have, we only know the gender they declare as [19:02:56] DarTar: @q3: probably because the first one is a 2012 paper [19:03:14] fafloe: yes, I guess that question would count as trolling [19:03:20] :) [19:03:50] I think the admin/editor comparisons are more interesting than sex/gender, especially when we think about community [19:04:06] ^ agree because of assumptions [19:04:07] Wikipedia doesn't allow for people to declare gender in user accounts--only via user pages using text or userboxes [19:04:07] but I am curious about the intuition that sentiment may (or may not) have changed in 4 years (which is almost half the life of the project) [19:04:22] Amanda_, the user preference asks "How do you prefer to be described?" (I prefer not to say) vs. She edits wiki pages vs. He edits wiki pages [19:04:24] That's gender [19:04:44] tnegrin, along the same line, I recently read a study about editor politeness that has compared politeness before and after becoming an admin http://www.mpi-sws.org/~cristian/Politeness.html [19:04:55] DarTar: +1 , a rerun of the analysis would be nice [19:04:57] that's a WMF POV [19:05:04] It was fascinating to see how editor behavior changes around an event [19:05:43] So.. If we're going to end on time, we have 25 minutes left. [19:06:03] 4. given that this is public data, all of these scores could potentially be generated for individual usernames (allowing arbitrary aggregations), would that be creepy? [19:06:06] +halfak, i respectfully disagree. :) [19:06:12] Amanda_, sfine [19:06:18] :) [19:06:18] halfak: ok on time check [19:06:54] we can also potentially stretch the showcase since we started late [19:06:59] (assuming people can stick around) [19:07:19] DarTar, IBM did tag cloud by user (topics most edited) they decided not to publish the code as too creepy (2005) [19:07:46] ezachte: right [19:08:02] DarTar, I am OK with stretch. Not sure about other participants. Going over is a common bit of negative feedback that we get. [19:08:15] -.o.- [19:08:18] ezachte: at the same time there are detailed stats on all of my posts to wikimedia mailing lists associated with my actual name, not username ;) [19:08:32] DarTar, might want to schedule stretch in the future. [19:08:44] halfak: ok, I’ll give david a sign [19:08:45] halfak, data doesn't support it. ;-) [19:09:01] Seems like an extra 15 devoted to questions and discussion would be a good call. [19:09:18] +1 [19:09:20] leila, I like data :) [19:09:27] DarTar (frequency not content), also happy to add an alias to your username [19:09:36] yeah, in the survey, folks didn't complain about start/end times. [19:09:48] that's the only data point I have. [19:09:54] Good point. [19:10:20] ezachte: I’d like to think that we should spend some effort on socializing the implications of public data [19:10:38] DarTar, let's skip intro. I'll jump right in. [19:11:12] Could we have david jump in here? [19:11:31] DarTar yep good topic, implications [19:12:34] hairy wikidata, ick [19:12:45] “halfak is not jimbo” [19:14:10] Citation needed [19:14:25] He's Jimbo's Tyler Durden [19:16:19] https://en.wikipedia.org/wiki/Zooniverse_%28citizen_science_project%29 [19:19:58] (woo!) [19:20:00] related: https://en.wikipedia.org/wiki/Systems_theory [19:21:12] key question people ask in the viability / resilience field (for example when modeling an ecosystem) is what counts as “viable” [19:22:13] Hi, here I am :) I've just gone through the whole conversation [19:22:20] relatively straightfoward for a cell or an organism, less so for a complex system like a community driven by a goal and a set of norms [19:22:50] SuggestBot [19:22:57] ^--- Nettrom [19:23:47] Sorry for using words males, females etc, thi is confusing in the paper we haven't, I think ;) [19:24:00] we just refer to gender [19:29:07] huge fan of Beschastnikh’s work [19:29:27] -> http://www.cs.kent.edu/~javed/class-P2P12F/papers-2012/PAPER2012-wikipedia-governance-Bechastnikh.et.al.ICWSM08.pdf [19:30:04] * DarTar waves at Giovanni_Luca [19:30:12] Hello [19:36:45] *snaps* [19:36:51] Link to quarry: quarry.wmflabs.org [19:36:51] thanks halfak [19:36:55] :) [19:37:00] * halfak fast talkin' [19:37:04] thank you! [19:37:15] thanks Aaron really interesting! [19:37:28] HI Giovanni_Luca! Would love to chat more :) [19:37:45] good talk halfak [19:37:49] I wnted to know about this (In the talk description): " the Wikimedia Foundation's analytics strategy currently focuses on outcomes related to a relatively narrow aspect of system health and all but completely ignores productivity." [19:37:57] o/ subbu (BTW, I'm in MN this week) [19:37:58] I guess time constraints didn't permit it [19:38:26] apergos, indeed. That's something I'd like to devote a whole presentation to. [19:38:33] halfak or anyone: Do you know of a popular way of measuring community complexity? [19:38:34] Any chance you want to collaborate on it? [19:38:38] :-D [19:38:48] halfak, ok. [19:38:49] how could I help? I'mnot a research analyst [19:38:55] apergos: I think halfak is referring to https://meta.wikimedia.org/wiki/Research:Metrics_standardization for some context [19:39:04] EdSaperia__, I know *of* Bar-Yam's work, but I haven't looked into the details yet. [19:39:16] I don't know if that's relevant (perhaps it is?) [19:39:17] ezachte, your questions are answered [19:39:17] apergos, just conversation is great. :) [19:39:21] sure [19:39:30] EdSaperia__, not sure that it is. I'll have to admit. [19:39:38] I was thinking, it would be interesting to use community complexity measures as success metrics for community managers [19:39:39] bmansuro_, I didn't ask your question since david____ responded to it while answering ezachte's question [19:39:43] (in wikipedia and/or elsewhere) [19:39:50] EdSaperia__, +1 [19:39:55] basically, they are not dealing with the complexity that a word like treffic may mean positive or negative [19:40:03] Did we loose Amanda_? [19:40:03] I have some related complexity ideas I'd like to run past you at some point. [19:40:14] All over it. [19:40:14] leila: sure [19:40:24] leila, we did, but she can watch the video later. [19:40:40] okay. her question was answered to. just wanted to let her know [19:40:43] DarTar: thanks, bookmrked [19:41:49] ah ha :-D (good correction on language use there) [19:42:00] uh... ants? [19:42:44] Aaron, I missed the first part of the talk but I noticed that your were doing an analogy with biology that I find really intriguing. Do you think we can apply concrete ideas from biology to WP, and how would they help in shaping the overall strategy? [19:43:03] biology can always be applied everywhere ^^ [19:43:11] ehehe, indeed [19:43:32] only slightly less than thermodynamics [19:44:53] but for example, I think Aaron was making a comparison between automatic vs semi-automatic anti-vandalism tools and cells mechanisms (if I understood correctly...), can you explain a bit more? [19:45:13] Ironholds: re: successful login, there's https://meta.wikimedia.org/wiki/Schema:LoginUserAgent [19:45:25] AWESOME [19:45:27] thanks YuviPanda! [19:45:36] I'll use that for that number, then :) [19:45:53] and then rely on my crazy-ass parallelised R for the rest. [19:47:15] Ironholds: :D ok! [19:49:03] I have been thinking, another way to think of the problem with (semi-)automatic tools is that it stops user action being local. So someone can apply a rule globally, which stops experimentation happening everywhere; local divergence (which admittedly can take you away from local maxima) is prevented. In fact even Watchlists can have this effect. [19:49:30] I realise that this is another way of stating the same thing, but feels slightly more general to me. [19:50:01] how does WMF's attempts to social engineering relate to Wikipedia's anarchistic roots (self organizing) [19:51:16] great talks :-) [19:51:18] So if we consider the WMF being the tool for centralisation, why is it necessary? Handles legal, handles fundraising (i.e. money), handles centralised software development, anything else it's required for? [19:51:18] ezachte, I wish I could answer that in voice. [19:51:19] thanks guys, it was interesting [19:51:34] how required is required? [19:52:05] ezachte, my strategy -- the only one I think has a chance to actually work -- is not to try to push the community/system in a direction we think is good. [19:52:06] folks, thanks for the lively discussion – I gotta run to grab lunch and to another meeting but feel free to stick around, this channel is up 24/7 [19:52:14] handles the hardware end too don't forget, and keeping the site up [19:52:28] I'd much rather find ways to enable the underlying, distributed, anarchistic mechanisms that are already in place. [19:52:36] And there are alternate ways of approaching this; e.g. If you had a truly decentralised fundraising model, you would have a community curated queue of projects that would compete for time on the central fundraising banner [19:53:07] halfak, sounds good, but WMF sometimes seems to have a tendency to want to steer the crowd from afar [19:53:17] Bye! [19:53:21] Rather than centrally fundraise and disburse via a centralised grant mechanism [19:53:21] ezachte, I think that a good strategy for doing this is to (1) figure out how/why leader-less organization works well and (2) identify opportunities to make it more powerful. [19:53:35] ezachte, agreed [19:53:36] I'm also going.. for dinner, here it is 10pm ;) [19:53:47] See you, david____ [19:53:48] o/ [19:53:48] thank you! [19:54:13] in a setup with admins and crats going leaderless is going to be tough [19:54:31] tougher the longer the power structures have been around [19:54:34] halfak let's get back to this sometime, again I thought your presentation was unusual and the more interesting [19:54:50] ezachte, sounds good and thanks :) [19:55:51] for now have a nice day:-) [19:58:16] Has anyone read this? http://falkvinge.net/2013/02/14/swarmwise-the-tactical-manual-to-changing-the-world-chapter-one/ [19:58:25] Saw what you like about falkvinge but it's a great book [19:58:42] halfak: do you have a paper r a pointer to th slides for the talk? [19:58:42] Lots of very practical tips for setting up social machines [19:59:06] apergos, I'll post the slides in a couple minutes. I'll link here when I do. [19:59:13] great! [20:25:39] halfak: thanks for the zooniverse reference [20:26:01] i vaguely rememberd the lunar crater counting, but all those other projects in it nowadays,, wow [20:26:09] new bookmark now [20:26:36] * mutante used to be SETI@home power user [20:26:47] :) [21:00:12] Ironholds: btw, I saw MaxSem's message on IE6 not having JS events, shouldn't affect this schema, since it is done in PHP (patch ref: https://gerrit.wikimedia.org/r/#/c/93526/) [21:01:32] YuviPanda, cool! [21:01:48] YuviPanda, Status: Abandoned [21:02:06] shit [21:03:40] Ironholds: it says, however, 'Logged by default' [21:03:44] so I think it's logged *somewhere* [21:04:38] I guess the easiest solution here is "pull up some logs from this week and see if IE6 is in there" [21:05:11] on that schema? [21:10:09] YuviPanda, it's not in the log table on analytics-store. [21:16:29] hmm [21:16:37] I guess it's being logged somewhere else... [21:16:41] which is why that patch was abandoned [21:16:44] poke ori, he might know [21:49:44] Hey folks. anyone still have the youtube link for the showcase? [21:50:42] Ironholds, ^ [21:50:47] leila, ^ [21:51:44] halfak, we used this: [21:52:00] I don't :( [21:52:16] leila, was there supposed to be something after the colon? [21:52:26] https://www.youtube.com/watch?v=-We4GZbH3Iw&feature=youtu.be [21:52:41] http://youtu.be/-We4GZbH3Iw [21:52:47] they should be the same [21:53:34] Thanks! [21:53:37] np! [21:53:43] * halfak adds it to the showcases page :) [21:53:54] heehee [21:54:11] halfak, how long do you think it would take to grab 30 days of sampled logs, parse all the UAs and calculate what percentage are IE6? [21:54:38] answer: 50 minutes. I love this library. [21:54:53] :P Was going to ask how big N is. [21:55:07] But 50 minutes is still pretty good. [21:55:22] * Ironholds counts [21:55:31] 240m lines, ish [21:55:32] FYI: slides from my talk -- https://www.mediawiki.org/wiki/File:Wikipedia_is_socio-technical_--_Research_Showcase_(October,_2014).pdf [21:55:34] it varies from day to day [21:55:37] I finally got them up. [21:55:41] yay! [21:55:47] I had meetings back-to-back and was doing upload wizard inbetween.