[00:23:33] halfak: shallow clone ;) [07:52:00] Hello, I am looking at Kannada stats, everytime, there is a difference of around 2000 articles at stats. wikimedia.org and at dumps. [07:55:10] tuxnani, would suggest emailing the analytics-l mailing list; people won't be awake at this hour on IRC from the team :/ [07:55:54] Ironholds: Thank you! [07:57:16] np :) [08:15:39] * quiddity sends Ironholds to bed. [08:19:49] YuviPanda: Any idea about the problem with article count in Kannada stats? dumps gives one number of number of total articles, and stats/site shows a different number. [08:20:09] tuxnani: sadly none. emailing the analytics list seems the best approach :) [08:20:30] I have sent one. [15:39:35] Guest75872: re: you being shipped diapers, https://www.youtube.com/watch?v=3V35jvY0u7I [15:40:43] o/ Uvi [15:40:48] * YuviPanda [15:40:59] * YuviPanda waves at halfak [15:41:10] U-V-Panda [15:41:33] :) [15:44:19] morning :) [15:44:19] Hey Ironholds. [15:44:19] You were around late last night! [15:44:26] I didn't have anywhere to be so I put some time into python! [15:44:31] wheee [15:44:42] halfak: Ironholds I’ll see you both next week? [15:44:53] Yup :) [15:44:54] * YuviPanda feels slightly guilty for not having done any analytics-y work at all since wikimania [15:45:06] YuviPanda, yup [15:45:13] don't worry, we have a hacking project [15:45:29] YuviPanda, if you want to talk about how we might set up a stream processing framework in labs for testing, that would rectify it. [15:45:29] halfak and I spent some time riffing on the possibilities for event streams [15:45:35] +1 [15:45:38] Event streams :) [15:45:49] conclusion: this is a tremendous force multiplier [15:46:39] halfak: :) indeed. I’m still vaguely confused about what exactly is the end goal, so talking in person might help fix that [15:47:33] YuviPanda, I think that we could make some quick progress by building a simple pipeline and doing something with it. [15:47:43] yeah, agreed [15:47:44] Ironholds, has an idea that would involve "thanks" [15:47:50] halfak: notconfusing emailed wikitech-l about having built something related [15:47:52] oooh [15:48:46] and it's a really simple idea. [15:49:32] But we'd need to at least generate reverts to do it. [15:49:54] I've built stream processing revert detectors in the past. An LRU cache goes a long way. [15:50:09] I can do enwiki pretty nicely with 2GB of ram. [15:55:16] neat! [15:55:22] and, yep, we would [15:57:38] Ironholds, I can block a couple of hours on Saturday to try to hook up a trivial good-newbie-edit detector. [15:57:56] I'm imagining something that would be python only, but it would let us experiment. [15:58:01] totally! [15:58:45] if the registration date is <1mnth from Sys.time(), the user is not blocked, the edit has not been reverted... [16:00:58] halfak: hmm, RAM isn’t too much of an issue with labs, easily get 64G machines... [16:01:26] YuviPanda, great. How's CPU? I don't think we'll need much, but I know that this was a problem recently. [16:01:37] halfak: yeah, we doubled RAM / CPU capacity [16:01:48] now have 2 and something terabytes of RAM available [16:01:59] total RAM, that is. we overprovision but not enough to cause issues [16:02:18] halfak: we’ll also be adding additional capacity in dallas this quarter [16:02:38] YuviPanda, great. I'm hoping to consume some of those new resources. [16:02:40] :) [16:02:52] halfak: yep. We’re going to be restricting them from the start for a while at least, however. [16:03:07] halfak: because we might change the underlying network model at some point [16:03:18] Restricting because of a "might"? [16:03:21] so we could perhaps have requirements such as ‘only puppetized projects’ allowed. [16:03:25] halfak: oh, no ‘might’. we *will* [16:03:28] only question is *when* [16:03:42] YuviPanda, that sounds like a destructive policy [16:04:03] halfak: well, the network model change isn’t backwards compatible at alll. [16:04:29] I don't see why limiting instances will help you manage a network model change. [16:04:41] halfak: it’s network between the instances. [16:04:46] not the underlying physical network [16:04:50] Sure. [16:04:52] Still [16:05:15] halfak: that’s partly because OpenStack has managed this terribly, I think. [16:05:23] halfak: anyway, I’ll check and see how much we can open it up. [16:05:45] halfak: we’ll have toollabs / betalabs partly there, so there’s going to be less capacity problems in eqiad [16:06:28] What's because of openstack? [16:06:38] The requirement for puppetization? [16:06:39] halfak: the network model migration / how it is tied to instances? [16:06:53] halfak: oh, I was only hypotheticating the ‘only puppetized projects' [16:07:23] the understanding is that at some point in time, we will migrate, and there will be heavy downtime. so we want to restrict access to that migration is easier for everyone [16:07:42] So, basically, no new projects except by staff (and the rare volunteers who are competent in puppet) [16:09:53] halfak: no no no [16:09:57] halfak: no new projects in*dallas* [16:10:05] halfak: nothing changes in eqiad at all. [16:10:15] halfak: this only affects people who want to have projects in two data centers [16:11:02] OH! [16:11:12] * halfak was confused and feels better now [16:11:21] :) [16:11:52] halfak: sorry about that :) I went from talking about new capacity *in eqiad* to new setup *in dallas* [16:11:56] so that was confusing [16:12:37] Ironholds: hiyaaaa [16:12:47] morning ottomata :) [16:12:55] No worries. But still. It's becoming clear that (1) I need to become competent in puppet and (2) we need good support for volunteers to pick up and use puppet. [16:12:58] can you respond do the comment on line 158 here: [16:12:58] https://gerrit.wikimedia.org/r/#/c/183551/2/refinery-core/src/main/java/org/wikimedia/analytics/refinery/core/Geocode.java [16:13:09] YuviPanda, python packages have to be converted to debs to work in puppet, right? [16:13:21] what to do about multiple subdivisions (aka regions, states, provincnes) returned for a given ip [16:13:21] halfak: yup. it’s a trivial operation, though. takes about 30s [16:13:30] Good to know. [16:13:47] Is there a, "so you want to use puppet in your labs project" document anywhere? [16:14:02] ottomata, totally! [16:14:36] done [16:15:08] haha [16:15:19] Ironholds: but what should he do? is returning the first subdivision sufficient? [16:16:03] ottomata, hey, you told me to respond, you didn't say anything about subtext :D [16:16:16] that's a good question: what does the command-line tool say about the test IPs he's basing this off? [16:17:47] YuviPanda, ^ see my Q about puppet documentation. [16:18:03] ? [16:18:12] halfak: nope. [16:18:17] halfak: I mostly learnted by stumbling around :) [16:18:29] OK. I have a plan then that involved SOFIXIT. [16:18:30] halfak: perhaps we can puppetize one more thing and write the doc in the process [16:18:35] ^ that [16:18:41] ottomata, ? [16:18:50] halfak: :) [16:19:03] But it will be important that I do (with your help) it so that we can capture a good FAQ. [16:19:21] I'll make a card so that this can be considered official work. :) [16:19:23] halfak: yeah, completely agreed [16:19:26] halfak: yup, yup :) [16:19:49] Ironholds: i dunno, i am asking if you wouldn't mind guiding some specifications here :) [16:19:56] requirements* [16:20:07] question: there sometimes are multiple subdivisions, what should the UDF do? [16:20:13] you are a researcher who cares [16:20:13] :) [16:20:28] I resent the implication that I've ever cared about anything! [16:20:33] maybe halfak cares? i remember he wanted a region udf once :) [16:20:35] I'll look into it ;p [16:22:42] o/ [16:23:07] That was an ad-hoc request, but I can see the general usefulness. [18:57:17] hey jfelipe, mako [18:57:34] we’ll set up the hangout on air in a moment, waiting for Office IT to join us [18:59:16] hi DarTar mako OK, I'm ready for the test [18:59:23] yay, science! [19:14:06] mako, if you’re around come to the 6th floor, we’re testing the A/V [19:20:04] DarTar: oh, sure, i'll come up now [19:24:23] Hello everyone. We're ready for the R&D showcase. Here's the link to watch the streaming: http://youtu.be/fr29FobogSM [19:24:41] I'll be collecting questions if you want to pass them to me. [19:30:04] Eeek! Nothing on my calendar for this event! [19:30:22] Oh noes [19:33:33] it's on staff calender halfak. [19:33:38] Ironholds: [19:33:45] LegacyPageview.java:[94,24] error: method patternIsFound in class Pageview cannot be applied to given types; [19:33:51] rejectedUriPathPages = new HashSet [19:33:55] Pageview.patternIsFound(rejectedUriPathPages, uriPath) [19:34:05] rejectedUriPathPages aint no Pattern! [19:34:12] * Ironholds blinks [19:34:12] whoopsy [19:34:14] For posterity: Terrible puns happened in the hangout chat about my avatar, which is a teddy bear. [19:34:16] this is why mvn test is important [19:34:39] haha, yup, that's how I found it :p [19:35:08] contains()! [19:35:12] * Ironholds fixes, retests [19:35:13] Also, the hangout on the calendar is wrong, check Leila's email. [19:35:14] There goes the stream [19:35:16] Sounds good [19:35:29] done [19:40:11] thanks Ironholds. merged :) [19:40:18] you want it deployed too, I suppose, huh? [19:40:43] would you mind if we waited until we got these other almost ready UDFs merged? Geocoding and your Webrequest one? [19:40:47] that would mean less deploying for me. [19:40:58] Felipe's paper if people want more details: http://faculty.poly.edu/~onov/Arazy_et_al_CSCW_2015.pdf [19:41:29] guillom: awesome, thanks! [19:42:27] I just noticed that the reader-to-leader framework has no arrow for those who leave entirely. [19:42:40] (Once you have you YOU CAN NEVER LEAVE!) [19:43:21] halfak: haha, i never noticed that that :) [19:44:00] halfak: it's like these collective action models with positive feedback loops [19:44:17] halfak: retention problem solved [19:44:37] halfak: Nor for people who found the project! [19:44:45] Arrow from the ether into the leader cloud. [19:44:50] where eventually either everybody joins and/or everybody becomes maximally committted and does anything else [19:45:00] sorry, does nothing else :) [19:45:02] The dude who runs the gothic Wikipedia needs an arrow too. [19:46:18] Where's the "researcher" role? That was 2008, I think. [19:46:40] That's a whooooole different graph. [19:47:05] halfak: am I still the only member of that group? [19:47:07] * DarTar ducks [19:47:13] I'm a member :) [19:47:20] oh wow, I had no clue [19:47:49] I requested the creation of that role way before joining WMF for the AfD study with Giovanni [19:48:25] https://en.wikipedia.org/w/index.php?title=Special%3AListUsers&username=EpochFail&group=&limit=1 [19:48:27] Feb 2008 [19:48:30] DarTar: me too! [19:48:42] Oh wait.. That's my account creation date. [19:48:43] I think User:Whim as well [19:48:45] I'm not sure I'd classify "importer & transwiki" into "level 5", or group it with stewards. I'd more of a level 2.5 or 3. [19:48:50] *Whym [19:48:50] yeah the group was created in 2010 [19:49:02] guillom, that makes sense to me too. [19:49:10] J-Mo: that’s annoying I thought I was a singleton like Jimbo [19:49:32] sorry to break the news: you are not a special snowflake [19:49:41] DarTar: Don't worry, you're still special :) [19:49:41] [19:49:49] <3 guillom [19:49:49] awwwwww [19:49:57] :) [19:52:36] So what about the bot maintainer role? [19:52:41] Or the anonymous editor role? :) [19:52:49] or the bot role? [19:53:27] would the bot (or bot maintainer) be lumped into one of the other cats, based on what the bot does? [19:53:31] hostbot for steward! [19:53:35] marktraceur: they use user rights to define the roles. Bot maintainer isn't a user right. [19:53:43] yes!!!!! [19:53:56] So maybe we need a user right for that. [19:53:59] Or a category [19:54:04] or they need a better definition :) [19:54:16] Or that [19:54:34] so, there is lots of other work on other kinds of roles [19:54:54] i'm think of david mcdonald's group and their work on using barnstars to classify type of work [19:55:05] but their categories were more things like "vandalism fighter" [19:55:19] There isn't a bijection between user rights and user roles; I wonder if the paper explains their choice to use user rights for this purpose. [19:55:36] these are roles that don't map to permissions and mw userrights [19:55:43] yp [19:55:45] yep [19:55:47] here's a paper on roles based on activity patterns: http://www.cecm.usp.br/~eris/pub/artigos/wikipedia/Finding%20Social%20Roles%20in%20Wikipedia%20-%20p122-welser.pdf [19:56:11] i mean, i think all three of these examples are different definitions of "roles" [19:56:35] like, i think ashaw and i talked about userrights in terms of lateral authority [19:56:38] rather than roles [19:57:03] J-Mo, do you think that the growth of the level-2s correspond to the rise of non-content-based WikiProjects? [19:57:04] we looked mostly at sysop/bureaucrat [19:57:47] Is that Jimbo on the right there? :) [19:57:48] I missed the slide…. [19:58:11] marktraceur: right :) [19:58:23] halfak: sounds about right [19:58:24] He's his own level. [19:59:33] Is Jimbo the most core user? [19:59:43] marktraceur: btw, you asked earlier about unregistered users; they're classified in the paper as "level 0: non community members". [19:59:49] halfak: ah, I see. Yes, this is interesting. I'd like to compare the growth rates there. [20:00:08] guillom: Fair enough, yes [20:00:14] ottomata, SWEET [20:00:20] halfak, guess what? [20:00:23] like, i wonder if these suprising results aren't a function of the way that mediawiki organizes user rights [20:00:30] marktraceur: it's a simplification though. [20:00:37] the legacy pageview def UDF is merged. *dances* [20:00:43] like, as a set of permission for what you can do as opposed to an actual progression [20:01:00] Ironholds: w00t [20:01:16] Ironholds: plot that thing [20:01:55] Ironholds: if you want to use the UDF before I deploy (and you don't mind waiting for me to deploy), you can just scp the jar, and use it, right? [20:02:10] yup! [20:02:13] DarTar: no questions from IRC directed to me. I'll be back in 3 min. [20:02:22] no need at the mo, tho, got pageviews data to set up for fundraising et al [20:04:04] "Ah, so We focus on human editors and exclude software bots from this analysis. [...] Since we cannot track subsequent role transitions for participants who initially contribute anonymously [...], anonymous contributors were excluded from our analysis." [20:04:12] ^answers some questions [20:05:04] how did they exclude bots? [20:05:21] ottomata, happy to wait on a merge until after the other UDFs :) [20:05:22] Probably by filtering out users with the "bot" user right? [20:05:25] that makes sense. and there's no systematic way to connect bots with bot-handlers (except through Tool labs). right? [20:05:28] danke [20:05:57] guillom, yeah, except for all the ones that don't :(. Ah well. [20:05:58] J-Mo: Most (all?) bots have a link to their owner on their user page, or in the page history. Depending on the wiki's policies. [20:06:25] Ironholds: we did some analysis on the bot user group method and its accuracy [20:06:26] yeah. it's probably a different template on different projects. [20:06:30] halfak: got a link handy? [20:06:39] Their analysis is enwiki only though. [20:06:49] DarTar, link for? [20:07:03] (DarTar: is speaking now) [20:07:19] Oh! That link. [20:08:32] hey shimgray :) [20:08:54] BTW most of their taxonomy comes from this earlier paper: http://ecis2014.eu/E-poster/files/0590-file1.pdf [20:09:13] https://meta.wikimedia.org/wiki/Research:Identifying_bot_accounts [20:09:15] DarTar, ^ [20:09:21] if aaron were here on the hangout, he might ask [20:09:26] :P [20:09:36] hey hey. btw, sorry for what no doubt looked like privacy trolling on analytics, I didn't expect "someone will probably object" to actually conjure it up ;-) [20:10:03] shimgray, it's a good conversation. I am glad we are having it. [20:10:18] non-topic wikiproject paper Dario is talking about: http://dub.washington.edu/djangosite/media/papers/cscw2014_wikiprojects_final_archival.pdf [20:10:59] heh, no-one ever went bust expecting wikipedians to fret about privacy :-) [20:11:57] hey shimgray [20:11:57] what halfak said, I’ll chime in again on that thread when I have a moment [20:12:19] It's worthy to fret about. One thing that troubles me is that the value to the user for the collected data is assumed to be zero -- yet there's a strong drive for the WMF to be more data-driven. To run experiments in order to know that the software we release is doing what it is supposed to do. [20:12:40] There must be a balance and there ought to be nice ways to opt out. [20:13:11] I suspect if we had the power to define powerful opt outs and to educate our users so that they set them according to their needs, we'd all agree on everything. [20:13:47] But with where we are, we end up disagreeing on how to make due with the imperfect state of affairs. [20:14:06] ah, but then there's opt-out-cost. we don't want ten million tickboxes or no-one will ever opt in or out on any of them. [20:14:24] +1 [20:15:05] is it practical to get a rough number of %age DNT for logged in versus logged out users? if it turns out 80% are logged in (which would not completely surprise me) then we can just bundle DNT in with logged-in and omit both from complex analysis [20:15:33] I don't think so, but I'd suspect the divide wouldn't be that neat; some browsers (screw you, microsoft!) have DNT as the default setting. [20:15:34] That would be problem. [20:15:38] So it gets all sorts of complex. [20:15:57] DNT as default? wow. that must be interesting. [20:16:01] (I'd dislike Microsoft less if they hadn't explicitly done it to screw with non-bing advertisers) [20:18:06] Anyway, we should probably focus on the ongoing showcase. [20:18:12] does someone have the stream link handy? [20:18:17] fine, mooom ;p [20:18:20] :P [20:18:24] http://youtu.be/fr29FobogSM [20:18:29] shimgray, ^ [20:18:41] Filipe Ortega wrapping up a presentation on editor roles [20:18:49] Mako Hill us just about to start [20:18:54] yeah, I';m listening (it's why I remembered IRC :-)) [20:19:04] Gotcha :) [20:21:30] thanks a lot guillom for helping the conversation on IRC and point to the article [20:21:35] jfelipe: i had one other question (feel free to answer after mako's talk): your paper dates the checkuser and oversight roles to 2002, yet the corresponding mediawiki extensions and user rights weren't introduced until 2005 and 2006, respectively... [20:21:43] jfelipe: sure! [20:21:49] jfelipe: great presentation :) [20:21:57] thanks guillom [20:22:33] ...so i was wondering how this was extrapolated (i know that server admins used to do this work informally before, but without specific user right) [20:23:48] HaeB: checking my notes (while listening to mako) [20:24:13] Audio just went bad briefly, but it seems to have recovered. [20:24:21] It was crackly static [20:24:28] halfak: Told cndiv in -staff [20:24:54] lilypad in the wild http://mako.cc/academic/buechley_hill_DIS_10.pdf [20:26:15] Thanks DarTar [20:29:12] one study of collaboration on Ancestry.com here: http://andreaforte.net/WilleverFarrCSCW2014.pdf [20:34:29] Fickin' a it's nate matias! [20:34:37] the one and only! [20:34:47] I mean, I think so. Unless MIT's cloning project finally paid dividends [20:34:52] heh [20:36:18] here's a paper that compares WP coverage of topics women tend to be more interested in than men (spoiler alert, there's less stuff): http://www.wikisym.org/ws2011/_media/proceedings%253Ap1-lam.pdf [20:37:44] HaeB: so it didn't came from dates in which the corresponding MediaWiki extensions were introduced [20:37:52] didn't come [20:38:13] werelate project just mentioned: http://www.werelate.org/wiki/Main_Page [20:38:17] but from textual records (early admins) [20:38:31] just trying to find the link to the document [20:40:10] thanks for the Ancestry paper - that's very interesting [20:41:30] any time! [20:42:02] yes, i've seen that before it's fantastic [20:43:06] jfelipe: i think even the terms "checkuser" and "oversight" weren't introduced much before the corresponding extensions (fun fact: an early version of the checkuser extension was called "Espionage" ;) [20:43:27] HaeB: haha [20:43:38] https://en.wikipedia.org/wiki/Wikipedia:Esperanza [20:43:43] ^ Social network for Wikipedians [20:43:44] for sure I can tell you the dates didn't come out of the blue [20:44:03] oh, goodness, Esperanza. there's a name from the past. [20:44:13] new editors seem to enjoy making profiles: https://en.wikipedia.org/wiki/Wikipedia:Teahouse/Guests [20:45:59] halfak: have you seen teh HBS case on esperanza? [20:46:10] I haven't [20:46:19] halfak: it's super interesting :) [20:46:29] leila +1 for testing on a wiki other than EN! [20:46:50] mako: this? http://www.hbs.edu/faculty/Pages/item.aspx?num=42518 [20:47:22] yeah J-Mo. I'm looking forward to that day. [20:47:34] That mic just traveled really far to go back to Danny [20:47:49] perhaps even not a Wikipedia. How about a project that is more receptive to individual opinions and less rule-bound, like Wikivoyage? [20:48:17] A place where people could remix content. [20:48:28] E.g. collections [20:49:32] totally, J-Mo. [20:49:40] (re last question., hmm. I vaguely remember a long-off proposal for some kind of users.wikimedia.org - common user-profile pages for crosswiki use.) [20:49:44] collections is a neat idea halfak [20:50:22] b.t.w., I've been hearing this idea of some sort of user profile page much more often these days, I don't think it's a totally hypothetical question. :D [20:51:19] https://www.mediawiki.org/wiki/Structured_profiles [20:51:31] not currently in development, AFAIK [20:56:01] Somebody (I don't know names, sorry) said a collections thing was being implemented. Is there a link to that somewhere? [20:57:45] I don't know if any of the things under discussion right now are being implemented. Can you remember any other details? [20:58:21] * halfak googles around mediawiki [20:59:28] DarTar: Timecheck 1 min [20:59:41] Emufarmers, not much public detail yet. The placeholder/notes are at https://www.mediawiki.org/wiki/Mobile_web_projects/Collections_Backend [20:59:52] jfelipe: a followup question would be if that kind of extrapolation could also be applied to the other user rights based definitions... eg the rollbacker right was created to accommodate the reality that there were already many users who did a lot of reverting (and were trusted in doing that constructively), without being admin [21:00:40] no, actually it was not by extrapolation that's for sure [21:00:49] ...could be applied or actually also should be applied ;) [21:00:55] remember we want to track the formal roles actually defined by the community [21:01:22] Oooh ooh I know about that projects [21:01:28] Well, I agree, but that "should" would be tricky [21:01:29] Incubator. Want to revive [21:01:56] At which point in time can we consider that the formal project was created besides the actual definition? A bit tricky I think [21:02:40] http://www.reddit.com/r/moiz [21:02:47] Can confirm -- is locked [21:02:58] hah :) [21:03:13] halfak: yep, I’ll close the broadcast and let people continue the convo if they wish [21:03:41] HaeB: hahaha tracked! [21:03:56] It was actually Jimmy performing those roles in the first place :P [21:04:15] Had to check my e-mail records of July, I was going crazy [21:07:01] DarTar: Yes, it's really exciting but I haven't had dinner (starving) [21:07:04] O_o [21:07:12] thanks for joining us jfelipe [21:07:15] great session today [21:07:25] we want your slides, after dinner [21:07:30] -> commons [21:07:31] you're welcome, hope to see you soon [21:07:45] I will upload them to commons, yeah [21:08:06] thx [21:09:26] WOo! Good show folks! [21:09:34] Thanks mako & jfelipe! [21:09:38] * natematias claps [21:09:39] hey, I'm the Brit here [21:09:43] o/ natematias [21:09:48] "good show". Quit stealing my stuff, viking! [21:10:04] you already did it at Lindisfarne! Did Alfred and Aethelstan not show you what happens? [21:10:06] natematias, I saw you in you in the audience and was pleasantly surprised [21:10:13] :-) [21:10:32] I was in town at the same conference as mako, and decided to drop in. [21:10:34] Once you finish your conversations, it would be cool if you popped back in here to give us some updates on your work :) [21:10:47] will do! [21:11:57] Cool. :) [21:12:25] Oh say Ironholds, I finished the Dawn. It was a fun book. :) [21:12:32] I might pick up the next one in the series. [21:13:18] the Dawn? [21:13:50] Octavia Butler [21:14:02] oh. YES. PLEASE DO. [21:15:13] Also, P.O.S. Doomtree is pretty awesome. [21:15:29] * halfak is recalling all of Ironholds' recommendations [21:15:58] yay! [21:16:07] I've been listening to a lot of his solo stuff recently [21:45:24] DarTar: EL is overloaded. Nuria, Kaldari, Maryana, and I are on it. You can safely ignore it for now. [21:45:40] yes, we just discussed it [21:45:44] cool [21:45:48] thx [21:45:51] np [21:50:44] DarTar, 1:1 is coming. Acquire charger plz. :) [21:53:12] Real nice work team :-P http://dispenser.homenet.org/~dispenser/temp/qualtrics.png [21:53:40] Dispenser? [21:53:43] Where is this from? [21:53:54] halfak: ha ha [21:53:55] That looks like a challenge to read/fill out [21:54:19] "Participate in the future of Wikipedia and MediaWiki" from "Wikimedia Design Research" [21:55:03] Dispenser, gotcha. That's from the design team. I'll hop over to #wikimedia-design if you'd like to join me. [21:55:19] * natematias is back. Hi halfak! [21:55:27] I had difficulties with my connection (was tethering) [21:55:40] Hopping threw so many channels, #wikimedia-office, #wikimedia-research, now #wikimedia-design [21:56:57] Hey natematias. I'm just about to hop into a meeting, but if you want to type at me, I'll be happy to read up when it get done (and I'm sure the rest of the channel would appreciate it too) [21:58:38] halfak: in that case, I'll probably head out to my next meeting at the EFF, but let's plan to find a time soon to talk [21:59:11] ttys! [21:59:39] natematias, OK sounds good. See you! [22:00:57] ttys! [22:18:58] There are literally an infinite number of Wikimedia IRC channels. [23:03:22] Please see https://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Rivers#Recruitment_Letter [23:13:16]