[00:49:01] halfak: hey! around? [00:49:21] halfak: so anon editing got turned on for mobile (FINALLY!). Do we have quick / reusable-code ways of looking at revert rates? [00:50:11] For all projects? [00:50:28] halfak: yes, yes. [00:50:38] it was a bit unexpected, but comminuty decision, etc [00:50:43] and it’s been kind of live for a while :P [00:51:24] YuviPanda, if I were going to going to look at these stats, I think I would want to compare historically. I'd use an XML dump or run a job in hadoop using joal's new XML input format. [00:51:41] I have some code that will work, but it uses the API and it's slow for that reason. [00:51:57] https://pythonhosted.org/mediawiki-utilities/lib/reverts.html#mw-lib-reverts [00:52:17] The base of this library contains general revert detection methods. [00:52:20] halfak: hmm, something like ‘1 week before, one week after’ might not be too slow for API even? [00:52:22] not sure. [00:52:30] I’m just mostly curious as to what the revert rates are [00:52:49] YuviPanda, you can do better than one week before/after. [00:52:53] they’re definitely causing a massive dive in new editor creation (http://mobile-reportcard.wmflabs.org/#monthly_reports-graphs-tab) but I don’t think that’s a bad thing [00:52:54] See mw.lib.reverts.api.check_rev [00:53:09] It takes an API session and a rev_id and returns you a revert if applicable. [00:53:23] halfak: oh, I see. [00:53:35] YuviPanda, where is this dive? [00:53:38] halfak: oh, and this runs only on teh API? [00:53:47] Well, I have one for the DB too. [00:53:51] mw.lib.reverts.db.check_rev [00:53:53] :) [00:54:29] Was it april 1st that we deployed the change to editing for anon/mobile? [00:54:29] halfak: edit-monthly-unique-editors, edits-monthly-successful, successful-edits-newbie [00:54:49] halfak: yeah, something like that. it was mergedon 30th march and rolled out a bit staggered. I can get exact dates later if you want [00:55:04] halfak: I remember you guys did a lot of work around this around Wikimania, so thought I should tell you :) [00:55:22] We need more time to know if there was even a large effect. [00:56:05] I've been monitoring itwiki. They did this back in Nov. [00:56:05] https://meta.wikimedia.org/wiki/Research:Anonymous_mobile_editing_in_Italian_Wikipedia [00:56:24] See https://meta.wikimedia.org/wiki/Research_talk:Anonymous_mobile_editing_in_Italian_Wikipedia/Work_log/2014-12-31 [00:56:48] I have to updated those graphs. [00:57:08] They extract the seasonal component. [00:57:20] That's what the gray lines are. [00:58:34] halfak: you totally should :D [01:11:11] halfak: do you think you’ll have time to do any of that anytime soon? [01:13:58] YuviPanda, good Q. [01:14:43] I just did an analysis to grab reverts that used a dataset from the 4th of March. [01:15:05] I could run the same code on the next dump when it is ready. [01:15:38] For registration rate, that's actually related to something I'm working on tomorrow, so that'll get done pretty soon. [01:15:52] I doubt I'll be able to finish it tomorrow though. [01:16:08] Anyway, I've got to run. Have a good one. o/ [01:21:38] halfak: \o/ ty [14:05:42] morning analytics [14:19:56] Hey hey Ironholds. [14:20:00] Wrong channel :P [14:20:14] hah [14:20:18] hey halfak :). How goes? [14:20:44] Good. I made a thing last night. [14:21:29] It will take a yaml configuration file and turn it into OOjs UI form fields -- in a single page HTML app -- to help you develop. [14:21:33] Wanna see? [14:23:29] Ironholds, ^ [14:24:06] totally! [14:24:13] I'm just replying to Hadley's enhancement requests to one of my libs [14:25:39] show me the yaml config file! [14:25:40] * halfak jumps in the batcave [14:26:03] Nop [14:26:04] Wait [14:26:08] DOn't go in the batcave [14:26:12] Analytics meeting in progress [14:26:13] has [14:26:25] gotcha! [14:27:20] * halfak calls Ironholds [16:01:49] Ironholds: hey; are we meeting today? [16:02:22] I think it might be useful, to see where we are and what we're each doing [16:04:27] guillom, we are! [16:04:28] sorry, I got distracted by having abuse hurled at me. 2 minutes? I need a smoke [16:05:02] Ironholds: no problem [16:05:09] Take 10 if you need :) [16:05:23] We don't have 30 minutes' worth of discussion anyway, I imagine [16:24:19] Hey YuviPanda [16:24:19] I've got a researcher who wants to stand up a very simple flask app. Got that flask in a bottle stuff ready? [16:24:19] Oh. I see you are away. I'll go bug Coren. [16:24:19] surely it'd be flask in a flask ;p [16:24:19] Ironholds, or "pocket flask". [16:24:19] But that suggestion failed :( [16:25:57] aww [16:42:38] halfak, can we host at http://www.wwsbd.org/ ? [16:42:38] What Would Shaowen Bardzell Do? :P [16:42:38] sb? [16:42:39] You and the Bardzells. :P [16:43:41] If we're going to pick feminist scholars, I want to highlight Sandra Harding. [16:43:45] that seems fair [16:43:50] Because we're all building "successors" [16:44:09] Which is one of my critiques of Bardzell's work [16:45:13] I'm honestly not sure what'd work best as a domain, unless jokes are the ideal [16:45:20] although I'll note that http://www.notbuilthere.org/ isn't taken [16:46:21] How about http://sociotech.science/ [16:46:59] wait [16:47:02] .science is a TLD now [16:47:08] Seems to [16:47:11] *so [16:47:15] oh my glob [16:47:24] http://hciisa.science/ [16:47:50] sa? [16:48:00] HCI is a.science [16:48:15] heh. [16:48:47] Goddamn. Someone owns wiki.science. [16:48:49] I like the idea of a .science. actually it looks like hci.science isn't taken [16:49:26] I'd feel back taking something too general though. I don't want to squat. [16:49:29] that's fair [16:49:59] but going too far in the other direction (sociophysics or sociotechnical) makes for less discoverability. Some happy medium? [16:50:42] halfak, peer-production.science? [16:50:52] the value being peer production, and the joke being it's a collaborative site [16:52:03] I actually really dislike the term "peer-production" [16:52:17] But I think the general direction is good. [16:52:58] hmmnmn [16:53:04] maybe we need more minds. Nettrom! [16:53:35] if you were hosting a collaborative site of blogs and visualisations and home-pages for people in our sphere, what would you call it? [17:03:11] plus.hci.science [17:04:05] or facehci [17:04:48] facehci? [17:05:02] The facebook of hci! [17:05:49] * guillom is being completely unheplful. [17:06:04] :P [17:12:15] my work will be hosted at barely.science [17:12:42] ^ thus [17:12:43] This [17:12:45] this is the one [17:12:48] hI [17:12:51] :) [17:12:55] :D [17:12:58] halfak: hi [17:13:02] o/ YuviPanda [17:13:29] Did Lego point you to docs? [17:14:29] The 5 sentences on wikitech? Yes. [17:14:40] Is that "flask in a bottle"? [17:17:54] YuviPanda, ^ [17:18:25] Mostly [17:18:36] You write your flask app [17:18:40] Put it in app.py [17:18:56] Put that in www/python/src [17:19:06] Do webservice uwsgi-python start [17:19:07] Don't [17:19:09] Err [17:19:11] Done [17:19:17] Need to be easier of course [17:19:32] No worries. That makes sense. [17:19:51] halfak: but no fucking around fcgi and lighttpd config [17:19:54] So there is that [17:20:31] I think I'd like to write a "So you want to run a flask app" page. [17:20:44] YuviPanda, yes. I have a burning hate for lighttpd and it's weirdness [17:21:23] It is shared yes [17:21:31] Yup yup [17:22:07] halfak, barely.science? [17:22:09] I like it [17:22:10] It would be nice to have an simple example flask app that makes a connection to the database and does a few other nice things. [17:22:11] I like it a LOT [17:22:12] :) [17:22:18] want me to register? [17:22:20] halfak: yeah [17:22:25] I want to have these templates [17:22:30] Simple [17:22:34] Simple plus API [17:22:37] Ironholds, working on it now. [17:22:40] Simple plus db [17:22:41] :) [17:22:47] +1 YuviPanda [17:22:50] Or even better have these be composable in some way [17:23:12] hahah [17:24:14] oo. Looks like techonology is a TLD now too. [17:30:19] :> [17:36:10] Emufarmers, looks like barely.science is already taken. Did you beat us to it? [17:44:30] halfak: wasn't me! [18:01:18] J-Mo, would you be interested in a group hosting solution for websites and blogging around science and HCI and community dynamics and such? [18:06:08] I like the idea, Ironholds [18:06:25] and what would you call it, knowing that .science is a TLD now? [18:06:28] we're iterating on names. [18:17:24] halfak, systems.science? [18:18:09] Too general. [18:18:16] open-systems.science? [18:20:07] then we run into the "okay was it one word or a dash" problem, would be my only blocker [18:20:18] open.science is too general too. Hrmrmrmr. [18:20:23] +1 [18:20:23] HAH [18:20:25] halfak [18:20:27] halfak [18:20:30] notrocket.science [18:20:30] I'd like open.science better [18:20:34] ha! [18:20:49] "where do you lot hang out?" "oh, it's notrocket.science" [18:21:08] I think someone should scoop up that domain regardless. [18:21:16] J-Mo, I got socio.technology [18:21:29] In case we want that. Otherwise, I'll find a use for it. [18:21:30] will do [18:21:42] omg halfak that is SUCH A DOMAIN WIN [18:21:47] :) [18:21:56] * J-Mo high fiving a million angels right now [18:22:40] socio.technology/jmo? [18:22:51] or jmo.socio.technology [18:24:06] socio.jmo [18:24:15] * J-Mo wants his own domain plz [18:25:10] wikipedia.jmo [18:25:17] jmo@jmo.jmo [18:25:19] lol [18:26:01] Am now the proud owner of notrocket.science [18:26:07] jmo@j.mo [18:26:16] ^ Macau [18:26:57] j.mo is not registered [18:27:04] J-Mo, ^ [18:28:30] ohhh… nice. which registratr do you use to get macao? [18:28:56] OH MY GOD [18:29:06] halfak, notbrain.surgery, pointer to Molly, is also not taken! [18:29:06] don't you register that, Ironholds. [18:29:08] * Ironholds squees [18:29:18] we could have notrocket.science and notbrain.surgery [18:29:24] how has /nobody thought of this/ and had 2 dollars. [18:29:52] The *.technology ones are expensive. I paid $25 for it. [18:30:09] .surgery are 60 [18:30:16] halfak: which registrar? [18:30:18] I guess they figure scientists are broke, developers are maybe broke, and surgeons are loaded [18:30:37] J-Mo, I did godaddy, but I'm still looking for one that will do *.mo [18:31:30] * J-Mo eeeexcellent… [18:32:02] if we don't do notrocket.science I am darn redirecting that to whatever we do pick. [18:32:43] ooh secret redirect easter egg. I like that. [18:33:38] J-Mo. Bad news. "Registrants must have a registered business in Macau, with the same name as the domain they wish to register" [18:33:56] Or.. now you have a reason to start a business in Macau named "j" [18:34:31] that seems like a solvable problem. [18:44:03] but seriously, does anyone object to notrocket.science as a base domain? [18:45:05] Ironholds, doesn't really say who we *are* [18:45:21] We're not rocket scientists? [18:45:23] But fair point. [18:45:36] behavioural.science? ;p [18:45:43] But I still find it very amusing and satisfying [18:46:00] crowd.science? [18:46:02] ooh [18:46:03] I like it [18:46:06] I really like it [18:46:10] collective.science [18:46:17] I like both! They work both ways [18:46:19] collaboration.science [18:46:23] "we are a group" "we are studying groups" [18:46:26] yes [18:46:29] I'd rather crowd or collective; shorter. [18:47:33] should we straw-poll it? [18:48:13] +1 Seems like we should me making a list somewhere. [18:48:27] makes sense. I'll start an email thread with relevant humans? [18:48:56] I'm thinking you, J-mo, me, R-stu, maybe protonk, whoever else is hosted on UMN or UW stuff [18:49:01] I wonder if Amanda would be interested? [18:50:33] ditto leila, but leila seems to avoid having much of a web presence. ninja.science ;p [18:50:51] ;-) [20:06:18] halfak, http://www.minnesotacricket.com/ [20:06:25] definitely picked the wrong snowpocalyptic state [20:11:56] Ironholds, you play cricket? [20:13:09] I used to! [20:13:29] There's not much call for cricket in Assachusetts, though. [20:56:26] I'm just going to put this here: https://twitter.com/hashtag/IAmAScientistBecause [20:59:38] guillom, aw [21:25:25] Ironholds: my brother plays Cricket in buffalo [21:25:26] apparently [21:25:28] ven though it’s cold [21:25:43] Cricket feels like a very ‘non minnesota’ meeting [21:42:28] hey, folks play cricket in MN too [21:42:38] if you come here, I'll show you where ;) [22:03:21] Ironholds, ^ [22:03:26] See Nettrom's comment. [22:04:35] hehe [22:04:48] halfak: is there enough non-snow covered area for a 22 yard pitch? [22:06:20] Nemo_bis, yay! [22:10:24] Yay! [22:10:26] But why? :) [22:10:42] oop, wrong match ;p [22:10:53] was aiming for Nettrom but he left. Boo Nettrom leaving! [22:51:22] halfak, https://github.com/halfak/Multiquery very cool [22:51:33] I wrote a similar R thing (global_query) about six months ago and it has saved my life so many times [22:51:36] :) I've been meaning to wrap that up for a while. [22:51:51] Now I see a typo in the README :( [22:52:01] * halfak aggressively ignores [22:52:10] The yak does not need shaving! [22:52:44] yay, I'm special! [22:52:57] I asked a medical friend about wtf I get multicolored beards. [22:53:07] Apparently it's because my genetic material comes from different phenotypes [22:53:15] I'm basically three people in one body. Biology is incredible. [22:53:26] Same thing with my eyeballs [22:53:29] Also I should use this in salary negotiations "you're getting three analysts for the price of one, here" [22:53:37] the heterochromidia? [22:53:44] Yeah. [22:53:47] * Ironholds high-fives [22:53:53] yay for being multiple people! [22:53:57] We are megatrons [22:54:07] Or captain planet [22:54:19] If I'm going to be a captain I would like to be Captain Marvel [22:54:21] she's hella-badass [22:54:22] ... not megatron... voltron [22:54:31] Ironholds is a chimera. [22:56:01] Ironholds: By the way, Erik asked me today why you were importing the data locally, and not using Magnus's WDQ tool, since it has all the data and is very fast. He might ask you later, so I'm giving you a heads-up. [22:56:04] guillom, actually that'd be far more extreme [22:56:08] cool! Thanks :) [22:56:56] ("later" as in "in 5 minutes") [22:57:07] yup ;p