[18:55:44] * guillom waves. [18:56:14] o/ [18:57:12] * Samwalton9_ waves [18:59:09] Good morning. [19:00:01] heyall [19:00:10] Hey @milimetric [19:00:35] o/ [19:01:04] good evening [19:02:40] and we've got a stream. : D [19:03:05] do we? I only see the splash screen [19:03:09] interesting, there's a "live chat" sidebar on the youtube view. does that bridge into hangouts or is that yet another separate chat? [19:03:24] brion: I think it's separate. [19:03:26] Would one of the channel overlords change the topic, please? [19:03:28] heh [19:03:35] Yeah, real efficient, I know. : P [19:03:37] nevermind...someone just walked past - it's live :) [19:03:45] lol [19:04:01] debt - Maybe it's pre-recorded. (*X-Files theme song plays*) [19:04:22] I (Tilman) am going to be relaying questions from IRC today [19:04:42] i_jethrobot: LOL! maybe! [19:04:47] ...line them up here to asked at the Q&A here in the end [19:05:08] Is there… a link to the youtube stream? [19:05:12] Welcome to [19:05:15] Tilman is the new James! [19:05:15] https://www.youtube.com/watch?v=6fF4xLHkZe4 [19:05:18] could you turn up the sound, pls? [19:05:23] Today's theme [19:05:36] Oh I think my browser messed up [19:05:37] halfak: is it cutting out for you? [19:05:43] Yeah refresh got it [19:05:47] Works OK on my end. Ah, OK. [19:05:50] sound is just really quiet for me. [19:05:54] (Wikilove the extension or the concept?) [19:06:03] happy anniversary ottomata [19:06:11] YairRand, live wikilove. [19:06:11] * funcrunch waves to i_jethrobot [19:06:23] * i_jethrobot waves back [19:06:27] As in "I want to share some wikilove with because of " [19:06:33] happy anniversary everyone! [19:06:33] also, the overview page with the agenda and all the links: https://meta.wikimedia.org/wiki/Wikimedia_Foundation_metrics_and_activities_meetings/2017-01 [19:06:42] funcrunch: I have some good news to tell you later [19:06:47] sound could be louder still, pls? [19:06:53] guy-ls [19:07:03] +1 to a bit louder [19:07:23] i_jethrobot: Cool! [19:08:08] Hello everyone! [19:08:21] hey! Thanks for the topic update. [19:08:26] \o/ [19:09:46] Half tof the people are chatting here and half in the youtube stream chatroom [19:09:58] gained up the audio, hope that is better for yall [19:10:06] rfarrand: yes, and I wish we could get everyone in the channel of record, which is this one. [19:10:41] i might be able to disable youtube chat for future metrics meetings [19:10:45] if that would help [19:10:51] Not everyone is comfortable with IRC, Pine. [19:11:00] YouTube chat is much easier for people to engage with. [19:11:09] ^ [19:11:18] True, but not everyone is comfortable with Google/YouTube chat, either ;) [19:11:19] And this is a community meeting, so I don't think forcing people to learn how to use IRC is a good idea :) [19:11:43] brendan_campbell: didn't we use to show this IRC channel on one of the three screens in the room? [19:11:43] HaeB, this is true, but it is infinitely easier, since it is right there by the video. [19:11:51] foks: it is most definitely not a community meeting. It's a public meeting, but it's run by WMF. :) [19:11:55] maybe we need some way of forwarding all comments in both directions [19:12:05] foks: only if you use a google account and are logged in [19:12:11] Pine, I would argue the WMF is a part of the community. :) [19:12:16] ^ [19:12:18] lol tried to past a link to freenode webchat into the youtube chat and it told me no URLs are allowed. [19:12:19] HaeB, yes :) [19:12:20] foks: I disagree very strongly. [19:12:29] But let's save that discussion for another day. [19:12:33] Pine, that's not fair and it's mildly offensive [19:12:56] I mean, a lot of us *are* community members. [19:13:03] Yeah, this conversation is too big for now. [19:13:08] Let's save that discussion for later. :) I'm happy to talk about it, but it's going to be a distraction from the main agenda. [19:13:14] milimetric! [19:13:20] gooo dan! [19:13:22] Speaking is Dan An-- [19:13:24] This work is super important for analytics work I have been doing for years. [19:13:27] Andreescu [19:13:28] oh, it's on the video. [19:13:30] :) [19:13:44] Pine: you are in the minority with this opion among the editor community https://commons.wikimedia.org/wiki/File:Editor_Survey_2012_-_Foundation_part_of_community.png [19:14:07] HaeB: I'm not going to discuss this further during this meeting. I'm happy to discuss later if there's interest. [19:14:27] Oooh, data lake. [19:14:32] Sounds...majestic. [19:14:55] https://meta.wikimedia.org/wiki/Research:Wiki_archaeology [19:14:59] This shit is complicated [19:15:35] I bet neilpquinn is a big fan of this. He's been doing lots of metric development. [19:15:51] brendan_campbell: can you double check the level of the audio coming from the office to the stream? the volume of folks in the office is wayyyyy lower than folks speaking on hangout (eg dan) [19:16:04] This is super useful for the audience work, and easy and good explanation from Dan! [19:16:39] halfak: absolutely! Analytics has been doing amazing work on this :D [19:17:11] (catching up, throwing out random opinions...) making IRC the only option is only going to serve to lock people out. We really ought to upgrade to a modern chat platform as the default for these, if we want to make them more accessible to a wide portion of the community. [19:17:21] +1 [19:17:27] As much as I love IRC. :) [19:17:37] Also this is really interesting work. [19:17:44] ragesoss, generally +1. The old IRC folk are hanging on tight though. [19:17:48] I'm a big fan of Chatzilla [19:17:51] this is the most gloriously de-normalized table. :) [19:17:53] +1 on the IRC thing. [19:17:53] for IRC - easy to use [19:18:10] ragesoss, I'm a fan of migrating so long as we can all agree on where we are migrating *to* [19:18:11] there are IRC bridges for all the likely options. [19:18:13] awjr: already did. thanks [19:18:23] awesome, thanks brendan_campbell :D [19:18:28] yuvipanda was working on a mattermost server for a while. [19:18:32] https://xkcd.com/1782/ [19:18:39] mattermost is like an open source/open protocol version of Slack [19:18:40] YairRand, I was looking for that ;) [19:18:46] -1 on IRC, still a standard channell [19:19:04] halfak: yeah, it'll take someone doing some sustained work with both community and technical foundations to make it actually happen. [19:19:22] milimetric, does "Analyze revision text" include ORES scoring on past revisions? [19:19:23] I think yuvi got demoralized by people who insisted that they would not move off of IRC. [19:19:43] but slack is also nice, just used it for the first time... [19:19:47] halfak: also, Matrix and Rocket.Chat. [19:20:05] That could answer questions like, "Were past edits more or less damaging as estimated by ORES", have community standards got stricter over time (yes, but quantifying would be good) (more reverts for same level of damage). [19:20:21] matt_flaschen, could do that but it would be pretty intensive CPU-wise [19:21:02] matt_flaschen: is this a question for the Q&A? [19:21:04] halfak, sure, it's interesting though, and there's no deadline. It could even be a separate ORES instance to avoid slowing down the realtime use cases. [19:21:05] i'm quite excited about future prospects of the data lake [19:21:06] HaeB, yes. [19:21:08] ragesoss, halfak: I know yuvipanda uses Matrix now, which is why I switched to it. I'm currently connected to this over their IRC bridge. Instructions at https://meta.wikimedia.org/wiki/Matrix.org :) [19:21:08] +1 to the chat suggestions. [19:21:16] nice dan! [19:21:18] +1 matt_flaschen [19:21:46] Super quiet on stream [19:21:47] thanks neilpquinn. I sense a likely contender :) [19:21:56] Office still pretty soft. [19:21:58] brendan_campbell: still really quiet from the office [19:22:06] adjusting... [19:22:13] ty <3 [19:22:21] how does Matrix compare to IRCCloud? [19:22:22] brendan_campbell: thx, alright now for me [19:22:47] * milimetric claps for Anasuya (so awesome) [19:22:51] kudos to abartov on the new "Cuteness approved" image ;) [19:22:57] I tried out Matrix brieflly, but setting it up was too hard. Maybe that's just because it would be used as basically a client for IRC, instead of a default chat platform. [19:23:27] ooh inset video [19:23:27] (I've got chat on the brain right now, because I'm working on integrating chat channels with Rocket.Chat into the Wiki Ed Dashboard) [19:23:43] milimetric, that was awesome. I'm stoked to play with the historical names, groups and block status :) [19:23:48] WCA approved! [19:23:51] jaufrecht: pretty well, I'd say. The interface is a bit less polished but it has all the same features (permanent backscroll, mobile apps, inline URL previews) and is under active development. [19:23:52] jaufrecht, IRCCloud is just a fancy proprietary client for IRC. [19:23:57] thanks brendan_campbell! [19:24:04] jaufrecht, Matrix is an open source protocol meant to be implemented by multiple apps. [19:24:07] i love the picture-in-picture of the presenter + slides [19:24:22] I'm pretty interested in Matrix, hadn't heard about that option before. [19:24:27] yeah, picture-in-picture is great. [19:24:44] matt_flaschen: let me rephrase: "How does the user experience of accessing Wikimedia IRC chatrooms via Matrix (and Riot, I guess?) compare to accessing them via IRCCloud's webpage?" [19:25:18] \o/ in ORES, we have ukwiki and ptwiki pretty well supported. [19:25:31] We've got a task for tamil, I'll check on the status of that [19:25:49] +1 to neilpquinn re IRCCloud, it has taken away most of the pain of using IRC for me [19:26:07] I've used Slack too (in a non-Wikimedia context) and am not missing too much [19:26:32] like all things, Slack is only useful if there is critical mass [19:26:38] IMO, we need an OAuth-based chat system, so that any editor can just log in without worrying about account setup or configuring a client or anything like that. [19:26:44] I came from a Slack shop and I think the key to good Slack is good slack management. It doesn't work if there's not a system in place. [19:26:54] I really liked that it was easy to use on mobile. [19:27:33] Ours worked really well. [19:28:08] ragesoss: +1 [19:29:07] Of course, I have crazy dreams like a chat channel automatically attached to every wiki page, too. [19:29:13] amazing [19:29:22] 200k edits!! [19:29:31] That's well over 500 a day. [19:29:38] This community development info is making me very happy. [19:29:45] no sleep for that user! [19:30:01] I really hope they're using semi-automated tools like gadgets, or have a really good ergonomist. [19:30:28] whoever that person is, we should get them a free massage! [19:30:59] This is super interesting/encouraging [19:31:00] Very insightful comment there about coming to people's countries because even with (limited) scholarships it's not effective for many people to come to Wikimania, etc. [19:31:02] BTW see also https://phabricator.wikimedia.org/T150732 "Provide a group chat system for mentoring" (I'm a bit skeptical though regarding that one though regarding communication channels balkanization. FWIW, IRC is workig very well for communication between my Outreachy mentee and me) [19:31:04] It can be more cost-effective for WMF too. [19:31:13] +1 matt_flaschen [19:31:51] On the topic of training, I did some work recently to hack up the Wiki Ed Dashboard to be able to pull content from a wiki and handle on-wiki translations as well.https://dashboard-testing.wikiedu.org/training/support-and-safety/keeping-events-safe [19:31:56] https://dashboard-testing.wikiedu.org/training/support-and-safety/keeping-events-safe?locale=es [19:32:00] :) [19:32:19] or rather: https://dashboard-testing.wikiedu.org/training/support-and-safety/keeping-events-safe/situations-you-might-encounter?locale=es [19:32:24] ragesoss, on-wiki translations meaning same-wiki translations (like Meta announcements)? [19:32:32] ragesoss: I've been meaning to go through the drashboard training, as I plan to integrate that into LearnWiki. [19:32:35] matt_flaschen: yes! [19:32:43] That's on my very long to-do list. [19:32:55] ragesoss, that's very cool. It integrates with Translate extension? [19:33:03] matt_flaschen: yes. [19:33:24] Yes, increased resources for this, please. [19:33:36] +1 Pine, love the talk abartov [19:33:36] matt_flaschen: https://meta.wikimedia.org/wiki/Training_modules/Keeping_events_safe/drafting/situations-you-might-encounter [19:34:03] This could be used to make translateable trainings for wikidata, Lua modules, etc. [19:34:04] Work in progress! :) [19:34:05] How should the Foundation decide whether to spend more resources on expanding this program vs all the other things the foundation could do? [19:34:18] jaufrecht, Annual Plan, basically. [19:34:25] (and that whole process) [19:34:26] and the strategy process [19:34:34] abartov: Nice work : ) [19:34:44] but also: by using Analytics, jaufrecht :) [19:34:47] abartov, great work. Really great to hear about cool programs like this. [19:34:49] abartov: high-five [19:35:11] abartov: totes great! [19:35:49] milimetric, oh, yes, also this :P [19:36:02] abartov: if you want help running a tech-talk recording of how to contribute to wikidata I will help you! [19:36:04] * foks , answering the wrong question since 2016 [19:36:14] I'm envisioning something like somebody giving a presentation like this and saying, "Our pilot shows that this program has a cost/benefit ratio of between 3 and 10." and then that could be compared to other programs. Is the current strategy process coming up with anything that we could use to make apples-to-apples comparisons? [19:36:16] Um.... is that applause or static? [19:36:17] the Seattle people were appropriately introduced with a plane landing first :) [19:36:22] Pine, it was static, I think [19:36:23] ^ [19:36:24] : P [19:36:31] I almost thought that was an intentional sound effect. :) [19:36:37] the mic is the laptop mic, so a lot of ambience [19:36:50] it was cool, sounded like the HBO intro [19:37:05] The timing was good, brendan_campbell. [19:37:09] especially, between very heterogeneous choices. What adds more knowledge to humanity, a new server farm in Africa or training six medium-size wiki communities in Wikidata etc? how would we make that choice? [19:37:14] Yeah, sorry, the a/v setup is the best we could do here [19:37:17] commons upload audio will be much smoother btw...having some issues with the remote folks [19:37:28] also: welcome to the strategy team, they're great (got to hang out with them at all-hands). Yall should talk to them if you haven't [19:37:28] guillom: it provided an amusing sound effect moment :) [19:37:29] thanks, everyone. I'm really pleased to see there's interest in this work. [19:37:30] no worries guillom i've adjusted [19:37:41] guillom texting ;) [19:37:52] rfarrand: absolutely! I've been meaning to do the Wikidata talk here at WMF to get a good A/V recording. [19:38:05] OK, will send you email and start Phab task :D :D [19:38:19] jaufrecht, very hard to make comparisons across different areas like this. It's hard enough to measure, "Is this project effective?" (it can usually be done, though), let alone "more effective". These aren't isolated systems happening in a lab. [19:38:20] jaufrecht: are these 3 questions still about asaf's talk, or about the strategy presentation? [19:38:32] guillom, it sounds pretty great considering [19:38:35] Things happen in parallel (people get training while there are perf improvements making edits faster, they're not isolated). [19:38:38] Asaf's presentation made me think of them. [19:38:45] abartov / rfarrand: make sure to get a 3D recording of that so we can have Asaf in VR [19:38:51] jaufrecht: unfortunately, I don't see WMF making that kind of comparison, at least not across departments. [19:38:58] milimetric: you don [19:38:59] (And I've closed IRC on my laptop to avoid the pings) [19:39:05] brendan_campbell: can we up Julia's mic a bit, please? [19:39:11] pine yes [19:39:13] I should rephrase, "very hard to make quantitative comparisons", it can still be done, but sometimes requires management, not just hard numbers. [19:39:14] *Pine [19:39:16] rfarrand: not texting, IRCing ! [19:39:17] jaufrecht: and this year's annual plan process promises to, once again, be super-compressed, so I am not optimistic about the chance to properly weigh this. Here's hoping I'm wrong! :) [19:39:24] milimetric: you don't want abartov to have MORE stalkers than he already has do you? [19:40:04] haha [19:40:49] I don't think we'll ever get it to, "program A is a 6 and program B is a 4 so give the money to program A", but is it worth it to start going further in that direction? Or is that so impossible that any effort to quantify like that is wasted and counterproductive? [19:40:50] abartov: question for you. Where there reasons that people brought up, to you or others, that the pilot you just presented about was not good in specific ways? I would not be too surprised if you say you haven't heard those since from what you described, it seems to be part of what WMF does. [19:41:08] i'll set up an automixer next week. will make this massive change in input gain easier to manage [19:41:09] abartov: please take a look at the Wiki Ed / Programs & Events training framework, when you get a chance. With wiki-based translatable content, I think it could be a really good fit for helping more language communities get into wikidata and such. [19:41:35] leila, HaeB: can we have that question as a part of the Q&A at the end of the meeting? (I'm interested too.) [19:41:43] sure, Pine. [19:41:46] :) [19:42:02] James_F: are you keeping track of questions? :) [19:42:08] James is out. [19:42:13] HaeB is James_F today. [19:42:14] He's on vacation :) [19:42:18] ow, Tilman is doing it. :) [19:42:23] brendan_campbell, too soft in office. [19:43:16] Sudden gain! [19:44:21] ragesoss: will do! [19:44:33] leila: I didn't quite understand your question, particularly its second sentence. could you rephrase? [19:45:16] leila: the first part I understood to be asking whether there was any negative criticism of the program. I can answer that, if that's what you meant. [19:45:22] abartov, basically I'm asking: did anyone come to you and say for reason x, y, and z this kind of pilot is a bad idea? And if so, what were the reasons/view points, they had. [19:45:27] This audio is almost impossible [19:46:16] leila: will ask the rephrased version [19:46:17] :D marktraceur [19:46:22] thanks, HaeB. [19:47:01] marktraceur: dan was very quiet in the office. could you hear me? [19:47:01] I totally could not hear Danny [19:47:06] marktraceur: It's not very bad, it just becomes too loud, or too low on my end. the too low part is okay, the too loud part is a bit more problematic. ;) [19:47:19] I can hear milimetric fine, but HaeB is quiet for me. [19:47:22] HaeB, basically not at all [19:47:32] And then every once in a while there will be weird static. [19:47:38] will be better next time [19:47:45] HaeB: the last question you asked, we pretty much didn't hear it. [19:47:51] (it was from Guillaume and the office being unmuted at the same time) [19:48:05] milimetric: (aaaa. yeah.) [19:48:09] the question was basically can we have ORES scores computed on historical revisions [19:48:34] Oh! Yeah! CPU intensive. I want both damaging and article quality models on old revisions :) [19:48:37] Yes, from above: [19:48:44] Question was, "Does "Analyze revision text" include ORES scoring on past revisions? That could answer questions like, "Were past edits more or less damaging as estimated by ORES", have community standards got stricter over time (yes, but quantifying would be good) (more reverts for same level of damage)." [19:49:28] I love Dario, "something happening in the public debate" [19:49:37] :D [19:49:38] ok will move closer to my mike on the next one and speak louder [19:50:17] HaeB: please continue to speak as you normally would. i will adjust [19:50:19] thanks Dario, we're *definitely* excited to solve the redirect confusion as part of this work [19:50:32] brendan_campbell ok [19:50:44] Yeah I was slow to mute us. Doing this from afar with the mouse on my lap :) [19:50:56] leila: got it. Nobody opposed the idea on Meta or wikimedia-l when announced, as far as I recall. The concrete pilots were all eagerly received. There was some criticism on technical details (e.g. the real-time interpreter in one case was not good enough, and was very distracting.) [19:51:09] guillom: no worries, this is why I played starcraft for 18 hours a day in college :) [19:51:20] :D [19:51:27] Time well invested! [19:51:28] abartov: thanks, but you should say it again. :) [19:51:56] abartov's talk: https://phabricator.wikimedia.org/T156409 [19:52:38] abartov, I'm a fan. I didn't know until know that the "emerging" targets had been decided upon! [19:52:57] Now that I know that tamil is on the list, I'll be making sure that we put some focus there. [19:53:38] abartov: thanks for your response, here and on the mic. That's what I expected, but it was good to hear it from you. :) From my experience, this kind of work is essential for the work we do, and I'm happy to see that th results you have seen support it. [19:54:03] milimetric: Nice, those apm come in handy! [19:54:05] I'll try to see if I can get some research interest in uk/pt/ta-wiki. [19:54:35] Who is the shoutout for? [19:54:42] Maggie [19:54:43] From Victoria [19:54:46] Oh! Great [19:55:17] Leon on YouTube points to tools.wmflabs.org/redirectviews [19:55:34] yeah, e.g. http://tools.wmflabs.org/redirectviews/?project=en.wikipedia.org&platform=all-access&agent=user&range=latest-20&sort=views&direction=1&view=list&page=Twenty-fifth_Amendment_to_the_United_States_Constitution [19:55:42] +1 about participation at All Hands. [19:55:48] allhands++ [19:56:01] halfak: yeah, it's been determined for almost a year now, but we haven't been able to announce publicly because we've been waiting on the ED to determine just how the rest of WMF owns that definition. I.e. what does this new definition mean for e.g. Product. So far, it keeps getting deprioritized. :( [19:56:06] musikanimal: we'll get you better data for that soon, too, but that tool's awesome [19:56:28] clap clap clap : ) [19:56:54] gotcha abartov. IMO, I just want to know what other people think and have the freedom to apply a grain of salt. [19:57:21] Priorities often come down to chance. [19:57:33] So knowing allows me to aim things in a better direction. [20:02:48] audio level will be more of a constant next time, thanks for hanging in there [20:05:53] thanks brendan_campbell, you're doing a really great job with these big calls [20:06:17] also: that static was mostly just us not managing the hangout mute/unmute quickly enough [20:08:27] thanks, i just have to adjust a couple of things. learn something each time