[17:56:13] * James_F waves. [17:57:25] I'll be your friendly IRC person today. [17:57:32] Friendly? [17:57:37] * Jamesofur will hold you to that ;) [17:57:43] Sorry I was away at the Hackathon last month. I trust that Jamesofur was adequate in my stead. [17:58:32] We'll be getting started in a couple of minutes. [17:59:06] As always, please do shout out comments and questions as we go so (a) I have a stash ready to ask at the end, and (b) people can answer them then-and-there for you rather than have you wait. [18:00:23] o/ [18:00:32] And we're live. [18:00:48] If you have A/V issues please shout. The always-awesome brendan_campbell will try to fix things. [18:00:55] yes please! [18:01:58] "Half" "acre" for future reference [18:02:08] Like half of an acre of land [18:02:42] I usually hear "Hal-fakker" [18:03:12] James_F, only four years? [18:03:26] foks: As staff, yes. [18:03:31] Huh! [18:04:14] I'd also settle for "half-ak" [18:04:27] or "hal-fak" [18:04:30] I suppose it's about 10 years since I became Chief Research Officer. But that wasn't a paid position. :-) [18:05:01] Heh. [18:05:12] when ORES makes bad decisions, it's HAL-fuck [18:05:25] ! [18:05:25] There are multiple keys, refine your input: bugzilla, logs, clippings, [18:05:32] James_F has been Wikimedia's Eminence Grise for over 10 years, but we don't talk about that. [18:05:36] ori, lol given the hilariousness of this, I'll also find it acceptable. [18:05:36] * Jamesofur likes "Touristic Interest" as a definition [18:05:37] foks: Don't trigger the bots, please. [18:05:49] Soz! [18:06:18] halfak: 'hæfeɪkər or hæ'feɪkər? [18:06:20] * James_F blushes at guillom. [18:06:52] guillom: don't feed his ego, it's big enough and we need to keep it in check so that he keeps working for us :P [18:06:56] <3 James_F ;) [18:07:25] Psh. How could I give up IRC duty? [18:07:40] It begs the question though; what color is a blushing eminence grise? [18:07:55] guillom: Cardinal red would be traditional. [18:08:02] Also, svwiki reached their 3 millionth article. [18:08:07] 'text units'? [18:08:09] Emufarmers, I don't speak phonetics regretfully [18:08:11] Hackathon video: https://commons.wikimedia.org/wiki/File:The_2016_Wikimedia_Hackathon_in_Jerusalem.webm [18:08:12] \o/ [18:08:16] blog post soon [18:08:38] halfak: darn, I was hoping to trick you into learning IPA :< [18:08:42] lol [18:08:43] Is it stressed on the first syllable or the second? [18:08:51] Thanks for the video link! [18:08:52] first! [18:09:11] James_F: heheh true Praveena would hurt you [18:09:28] and Ramallah : https://commons.wikimedia.org/wiki/File:WikiArabia_tech_meetup_in_Ramallah_2016.webm [18:09:42] Jamesofur: Let's not suggest we live in a climate of fear of physical violence. Some of us are friendly. ;-P [18:09:50] victorgrigas: Awesome videos, thank you. [18:12:53] James_F: question for the end of the meeting. Can we get an update about the CTO search? [18:13:01] Assuming that Katherine doesn't give us that update right now. [18:13:09] heh [18:13:47] Here it is now! :3 [18:13:48] Pine: Was that update just then enough for you? [18:14:01] My video is behind so I'll know in a minute or two. [18:15:23] James_F: I'll withdraw my question for now, but may bring it up again next month. Thanks. [18:16:00] Sure. [18:20:12] Yeah! Surprise transparency presents! [18:22:07] Katherine is pretty impressive. :) [18:22:52] It is a lot. Very glad to hear. Glad it's being presented in a transparent way [18:23:42] Yeah! Awesome pronunciation. Props to neilpquinn [18:24:15] Hah. I like the monster truck visual. [18:25:02] krmaher <3 <3 <3 [18:25:39] With no drop in productivity! [18:25:47] Just less reverting necessary [18:25:49] \o/ [18:26:58] neilpquinn gets props for putting the caveats front and center [18:27:02] Science is messy [18:27:05] +1 [18:27:34] oh-res [18:27:56] tough to measure the creation of "knowledge" [18:28:01] Top tip: Don't have an MN accent if you want people to pronounce things the way you think of them. ;-) [18:28:12] (Can everyone hear Aaron?) [18:28:13] echo! [18:28:14] Yes! [18:28:18] Echo is gone [18:28:22] \o/ [18:28:24] Excellent. [18:28:31] bedankt! [18:28:35] foks, I'm sorry to lose it kinda. It was unusually charming. [18:28:42] (halfak's MN accent is basically nonexistent) [18:28:46] Heh. :D [18:28:50] James_F: we seem to be be doing remarkably well on time. [18:29:00] Pine: All credit to the speakers. [18:29:11] We're a couple of minutes behind but not vastly so. [18:29:45] I'd need to re-listen to Katherine's presentation at 0.75x playback speed so that I can absorb it all. [18:29:46] OT, but wow, they're jailing people for not participating in the nuclear program. That's a real damned-if-you-do, damned-if-you-don't. [18:30:54] So the implicit thesis of this approach would be knowledge = text that is added to Wikipedia and left present for a long time or through many revisions? [18:30:55] "word-like token" <3 [18:31:09] What did he say - git what? [18:31:12] git blame [18:31:14] git blame [18:31:16] ah. thx [18:31:17] Part of git version control system [18:31:19] hey harej :) [18:31:25] hello, director! [18:31:27] foks: "Aka wikitext sucks for algorithms for human-written text". [18:31:30] got it. thx [18:31:33] Heh. :D [18:31:36] halfak, it doesn't have to be eventually removed to be a persistent word, right? [18:31:58] krmaher: will your slides be on Commons? I'd like to look at them at a contemplative pace. [18:32:13] Pine: The slides are always published afterwards, yes. [18:32:17] :) Thanks [18:32:23] Pine yes, Metrics slides are always published [18:33:07] I think they will show up here - https://meta.wikimedia.org/wiki/WMF_Metrics_and_activities_meetings/2016-04 [18:33:47] So we're getting more efficient, but spending less time? [18:34:05] matt_flaschen: Yes, I think that's the thesis. [18:34:16] krmaher: also, I +1 the others' comments about transparency. Very much appreciated on that front. [18:34:20] Very interesting. Imagine if we could spend more person-hours (with diverse persons) and keep getting more efficient... [18:34:29] THIS ANALYSIS IS AMAZING [18:34:30] +1 [18:35:01] Ah, dogs. [18:35:07] Dog!!! [18:35:10] Hello, Aaron's dog! [18:35:10] :D [18:35:22] Out of sequence, that was an inexplicable comment, James_F. :) [18:35:30] Kids and dogs at metrics. <3 [18:35:31] don't get the hypothesis. what if the quality level of contributions is constant over time but people simply stop improving content? [18:35:33] Just a passing fancy.... [18:35:37] victorgrigas: +1 [18:35:40] Maggie_Dennis: Just think of me as prescient :-) [18:35:44] LOL [18:35:58] pajz, I don't understand what you mean? [18:36:08] pajz, the flat line shows a constant amount added, not a constant amount in existence. So people are still adding a lot of stuff, just not adding an increasing amount of stuff. [18:36:14] Though I am trying to read and think with Aaron's dulcet tones in my ears [18:36:54] I wonder what that massive bot spike in 2011 was [18:37:55] Interesting that Twinkle isn't there [18:38:05] James_F, halfak, does this use the hashtag edit summary support (#proveit, #refill) for on-wiki tools? [18:38:06] foks: It's about reverting, not adding content, so… [18:38:19] matt_flaschen: In the stash. [18:38:22] True, true [18:38:27] Thanks. [18:39:28] matt_flaschen, hmm, I get that, but if you have, say, 500 persistent contributions in 2006 and 490 persistent contributions in 2016, and the number of contributions dropped much more sharply, that could be an indication that people stop improving content. changes are more persistent because what would have been fixed in 2006 is no longer fixed in 2016. [18:39:41] Am I missing something? [18:40:57] pajz, oh, so you're saying the amount of new persistent content would have fallen (more) if stuff that should have been reverted/tweaked was. But there aren't enough reverters/tweakers? Yeah, that's a less optimistic but very possible interpretation. [18:41:08] James_F: question for you or anyone else about VE. What do you think would need to happen to increase VE "productivity" significantly above the baseline? [18:41:43] matt_flaschen, yeah, that's about what I mean. [18:41:58] Pine: I don't have an answer for that. There are some ideas we have, but we'll need to see whether they work out. (See the Annual Narrative for some words about that.) [18:42:41] VE has a button that does what reFill does, does that make us a "force multiplier" tool? [18:43:04] Does VE do it en masse in the same way? [18:43:18] (because that would be superb) [18:43:19] foks: No, prompting users to do things is one of the ideas. [18:43:57] wow, this tool has come a very long way since I checked it last 2 months ago. [18:44:02] edsanders, is there an expand one, or just Citoid for new ones? [18:44:03] very well done, y'all. :) [18:44:05] yeah - a toolbar to let you go through all instances of a page would be (a) neat and (b) not that hard to build [18:44:12] matt_flaschen: expand [18:44:23] Cool, didn't know that. [18:44:31] click on a ref that consists of one external link in VE [18:44:33] Oh wow [18:44:41] The stream is breaking up for me. :( [18:44:43] I didn't know you could change the charts! [18:44:55] foks: They're neat, aren't they? [18:44:58] So many pings :) [18:45:09] "Donut, which is basically the same except you have a hole in the middle" [18:45:11] touché. [18:45:58] https://en.wikipedia.org/wiki/User:Ed_g2s/Sandbox?veaction=edit [18:46:13] very nice work on the pageview analysis! [18:46:22] Thanks edsanders. [18:47:22] matt_flaschen: stream still breaking up for you? it looks good on my end [18:47:59] brendan_campbell, probably my end. I refreshed and it's fine now. [18:48:10] cool cool. [18:48:25] That second result. [18:48:26] * James_F sighs. [18:49:00] James_F, ? [18:49:13] (Specifically, https://en.wikipedia.org/wiki/Barack_Obama_citizenship_conspiracy_theories was the second result when MusikAnimal searched.) [18:49:21] da is danish [18:49:22] It's Danish, fwiw. [18:49:28] harej, jinx! [18:52:11] nice work on the analysis tool [18:52:52] Questions? [18:52:54] James_F: I appreciate our naming structure for that article ;) "Conspiracy theories" may be notable but I appreciate that we call them out as conspiracy theories :P [18:53:37] * James_F has no opinion as staff. [18:53:59] SF mostly just wants lunch I think [18:54:03] we'er like that [18:54:05] *we're too [18:54:14] heh [18:54:18] * Jamesofur says despite actually skipping lunch for 1:1 with Maggie [18:54:18] yaaa [18:54:19] Jamesofur: As in, humans? [18:54:31] :) [18:54:47] guillom: as humans I've found we forget to eat often ;) though not usually on metrics day :) [18:55:02] or maybe that's just me [18:55:05] Fair enough. [18:55:13] "oh wait, it's 3pm maybe I should have brunch...." [18:55:29] tools were still a tiny percentage [18:56:31] yes, tools (and labs) are important, but not sure the data supports moiz' conclusion ;) also, with citoid, VE has a reflinks-like feature integrated [18:56:42] And other tools. :-) [18:56:52] E.g. dab-fixer. [18:56:53] Etc. [18:59:57] Is there any difference in the proprtion of editors who have edit conflicts in V/E re classic editor? [19:00:03] There are some problems with number of edits as a metric. For example, with wikitext, I read articles, and edit section-by-section as I go down. [19:00:03] Comment: one place that VE makes a big difference is with table editing, but that's a bit of a niche activity. [19:00:07] VE doesn't have section editing. [19:00:19] Last I checked. [19:00:23] matt_flaschen: Yup. [19:00:33] \o/ [19:00:37] * Pine applauds [19:00:54] Also adding links with VE is soooo much easier than Wikitext. [19:00:57] Thanks everyone. [19:01:00] If only because of the suggester. [19:01:33] foks, you should talk to aripstra et all about link inspector. Lots of iteration and empiricism went into that. [19:01:42] Noted! [19:01:52] Was a major blocker to us deploying the May 2015 experiment [19:01:55] so worth it :) [19:01:58] I didn't know it had a name! [19:02:06] But that's cool. It's really very useful. [19:06:55] it's also been upstreamed out of VE into mediawiki ooui gadgets, so you can use it wherever [19:10:23] foks: I vaguely think we had 17 distinct major changes to the UX workflow of that tool, BTW. Iteration until it works "right". [19:10:34] I wrote a list at one point. [20:26:57] halAFK: https://en.wikipedia.org/w/index.php?title=Aaron_Halfaker&diff=717630123&oldid=717585822 this will enable dozens of people around the world to say your name [20:38:49] halfak ^ [20:47:44] Emufarmers: there are dozens of people who understand IPA? :P [21:08:43] I only now about that one https://en.wikipedia.org/wiki/India_pale_ale [21:08:50] s/now/know/