[03:50:40] i'm confused https://www.mediawiki.org/wiki/Talk:Wikimedia_Research/Research_and_Data#Legacy_or_not.3F [13:31:17] Friday thought: When coastal cities flood. There will probably be an industry of divers and dredgers that salvage stuff from old cities. What will they be after? [13:31:55] It's not like the water level will rise in an instant. So it will be something that is left behind but becomes scarce or valuable in a post global flooding Earth. [13:32:09] a surprising amount of money and strategic resources [13:32:23] the Federal Reserve Bank of NY has a large amount of gold bullion from the US Reserve [13:32:45] They'll probably truck that out as the water level rises though. [13:32:52] and outside of gold reserves, two of the Petrolleum Reserve facilities are in (respectively) Baton Rouge and Lake Charles, Louisiana [13:32:56] true! [13:33:13] * halfak imagines a drilling platform above an old city [13:33:27] I imagine you'll have a class of ad-hoc divers who act as almost freelancers, getting particular things people left behind and want to check for [13:33:34] "granddad's gold watch, we forgot to grab it etc etc" [13:33:56] I wonder if there will be some abandoned security boxes in the banks [13:33:59] outside of nostalgic things, bank contents generally? I'm thinking Safety Deposits rather than cash. [13:34:01] snap [13:34:11] :D [13:36:27] Two fun Hugh Howie books come to mind. "Sand" is about sand divers who use special tech to swim through the sand into cities that are buried thousands of feet below. "The Shell Collector" is about a reporter in love with the sea post global flooding. [13:37:27] It was fun to read The Shell Collector and have people discuss global warming/flooding as a problem they were born into. [13:37:30] oooh [13:37:37] fun = scary & interesting [13:38:28] have you read Seveneves yet, btw?# [13:38:46] Nope. Haven't gotten to it. I'm still trying to finish off Quicksilver. [13:38:50] I'm 90% in it. [13:39:06] Seveneves pretty good? [13:39:29] Very! [13:39:42] but I think you'll like the premise; it's very practical science based on an impractical premise (the moon explodes) [13:40:04] So it's: how do you rebuild humanity from 'everything you can get into space in 2 years and no support otherwise' [13:40:58] Sounds pretty good. I just finished The Martian too. I highly recommend it. [13:41:02] I read it in one day. [13:41:07] It was hilarious. [13:42:38] yeah, I read it one night; liked it quite a lot, although it felt kind of...engineer porn-y [13:44:06] Yeah. Totally is. I really enjoyed that aspect. [13:44:22] I kept feeling like "These are things an engineer would never feel the need to explain" though. [13:44:31] I mean, to other engineers [13:47:06] For the sake of the log, I feel like I should point out that "Engineer porn" is building an improvised Mars rover trailer out of random parts of other things and discussing the testing strategy to make sure it works. [13:48:22] yep [13:48:53] or juggling food supplies and vitamins so that you can sustain yourself for 3 years on multivits and potatoes grown with the aid of split rocket fuel [13:49:44] OMG the rocket fuel. "I'm writing this log from the rover because I fled the hab in fear of exploding." [13:51:44] yes! [13:51:46] that was the best bit! [14:47:54] hey lzia, I'm not sure my schedule can afford a half-day link recommender hackathon after all. Is there a way I could still effectively move the project forward if I were to contribute at the 1-1:30 hour level today? [14:48:34] o/ J-Mo [14:48:34] hi J-Mo. This is totally optional for you, if you just want to sit around and do your things, and also have an eye on the conversations. [14:48:38] hi halfak. [14:48:44] * halfak digs into teahouse stuff this morning [14:48:49] G'morning lzia :) [14:49:06] thanks lzia, what would be the best way for me to help right now? [14:49:25] so, J-Mo, do what you think it's best to do. We will talk to you about user testing after today. I can ping you later in the day, or next week. [14:49:26] ("wait for me to contact you with marching orders" is a fine answer) [14:49:52] that sounds good to me. right now, you're mostly working on prototype, right? [14:50:42] J-Mo: right now, nothing. :-) The goal for today is to make http://tools.wmflabs.org/navlink-recommendation/ design ready. It won't be production ready, but it would be ready enough for us to be able to start testing. [14:51:31] nice. I'll take some time today to look it over and start thinking about user study stuff then :) and we can touch base again next week. sound good, lzia? [14:51:38] (also, thanks for letting me off the hook ;) [14:51:41] There will be another half-day hack probably next week, J-Mo, and if we achieve our today's goal, user testing can also start towards the end of next week or whenever you have time. [14:51:50] excellent! [14:51:59] great, thanks, J-Mo! [14:52:29] (totally, J-Mo. the optional invitation was more of a FYI, and in case you really like to sit in there. ;-) [15:03:34] Morning, scientists. [15:49:31] o/ guillom [15:49:43] hey halfak [15:51:20] Just saw your 'morning :) [16:26:28] J-Mo: how many more interviews do you have for article rec? I'm wondering when is a good time to review the results and I'd like to block that time in my caleneder. [16:26:32] calender* [16:26:53] madhuvishy: around [16:26:55] ? [16:26:57] hi leila [16:27:03] YuviPanda: yeah [16:27:08] hellooo yuvipanda. you are not back, are you? [16:27:10] at standup etc but can chat [16:27:15] leila: no flying in a few hours [16:27:24] ooki, safe trip yuvipanda. [16:27:36] leila: I finished reading all the feedback and have been using it in discussions at the offsite but haven't written it down will do that [16:27:41] leila: at least one more, scheduled for next Thursday (a re-schedule) [16:27:42] madhuvishy: yaeh, need to puppetize article rec :D [16:28:06] should have some results written up by the following monday [16:28:18] YuviPanda: cool, its a similar flask app no? [16:28:26] madhuvishy: ya, simpler actually since it's just flask [16:28:27] is there a fab based deployment already? [16:28:28] and no celery [16:28:31] no [16:28:32] yuvipanda, I want to help you with the documentation. Do you want to block some time next week for doing it together? [16:28:33] there's nothing. [16:28:34] okay [16:28:37] leila: ya [16:28:46] J-Mo: okay. thanks. I'll wait until after that one then. [16:28:46] code's in wiki-ai? or elsewhere? [16:28:53] leila: can you poke madhuvishy and get her pointers to the code? [16:29:05] madhuvishy: also do a new project request? I'll create it as well [16:29:27] YuviPanda: on labs? [16:29:39] madhuvishy: https://github.com/ewulczyn/translation-recs-app [16:29:41] madhuvishy: ya [16:29:59] madhuvishy: on wikitech.wikimedia.org on left sidebar there's 'request project'. it's currently running in the 'reportcard' project which bugs me :D [16:30:06] okay, i'll call it translation-recs? [16:30:08] ya ya [16:30:30] madhuvishy: can you call it just recommendation or art-recs? [16:30:32] :D [16:30:48] art-cre-rec, madhuvishy. ;-) [16:31:15] Security by obscurity! [16:31:56] totally, guillom. [16:32:27] leila: fine with me - art-cre-rec i will never spell right :P art-recs or recommendation - your pick :) [16:32:50] art-recs madhuvishy. :-) [16:33:02] alright! [16:33:23] (since we will have a need for link-recs soon, too. ;p) [16:34:01] Then you can create "standard article recommendation", and it'll be st-art-recs. [16:35:33] stuart resc [16:35:35] or somthing like that [16:35:51] leila: we are trying to get revscoring/ores deployed this month, btw :) [16:36:13] oh! that's a great news, yuvipanda. :-) [16:36:22] leila: it's going through security review now [16:36:50] Where's Oliver when I make nerdy puns. [16:48:03] I want to get the list of user groups on a particular wiki with at least one member [16:48:09] https://nl.wikipedia.org/wiki/Speciaal:GroepsrechtenWeergeven [16:48:15] there is too many that are without use [16:48:29] what would my uarry uery look like? [16:53:57] guillom, grooooan [16:56:22] every time you groan god kills an isis member [16:58:23] madhuvishy: created project! [16:59:07] YuviPanda: yup saw :) i'll start looking at the code and setting stuff up - later today may be [16:59:41] madhuvishy: kk [17:00:14] madhuvishy: setup a fabfile too? also not sure if ellery's in the office - maybe catch up with him afterwards, show him how the fab stuff works? leila ^ [17:00:30] madhuvishy: basically, be the devops/techops person for this for now :D [17:00:35] YuviPanda: i'm not going to office today - assumed no one would be there [17:00:41] madhuvishy: yeah not today [17:00:45] madhuvishy: eventually etc [17:00:54] yeah alright [17:00:59] i'll catch up with him [17:01:58] leila: are you all thinking of keeping the source in ellery's github? or gonna move to a different org on github/gerrit? [17:02:03] YuviPanda: ^ [17:02:24] (gotta go to flight) [17:03:40] okay let's catch up later [17:22:36] madhuvishy: for now it's on Ellery's github. do you have a recommendation? best practices? [18:36:44] TIL "operationalized" is not a word [18:36:54] Of course, it's not, but yes it is. [18:37:17] A good example of the nonsense we make when inventing language. [18:40:08] Did you know literally means figuratively? [18:40:26] http://edition.cnn.com/2013/08/15/living/literally-definition/ [18:41:19] Yeah... that, I think is silly. "Literally" is used as "super figuratively" as hyperbole. Changing the definition makes it no longer hyperbole. [18:41:26] What is the English language becoming? [18:41:50] Next: People start using figuratively to mean literally [18:42:06] I figuratively am in pain! [18:42:07] :p [18:44:54] ToAruShiroiNeko_: https://twitter.com/gpaumier/status/566394361525448704 [18:48:40] unfiguratively. I like that [18:50:14] Created User:Unfiguratively [18:52:24] I may make that a bot to automatically corrects edits that incorrectly use figuratively at some point. [18:52:36] I will single handedly fix the intertnet. :p [18:52:46] :D [18:52:57] guillom do you have a suggestion for me regardling the uarry thing I mentioned? [18:53:08] I want to get the list of user groups on a particular wiki with at least one member [18:53:20] I haven't see that. [18:53:32] Relocating to my desk now, brb. [18:53:34] seen* [18:54:51] ToAruShiroiNeko_, query link? [18:55:14] Otherwise, "SELECT DISTINCT ug_group FROM user_group;" [18:55:29] yes but would that check if it has at least one user? [18:55:38] things I had in mind were a tad bit too complicated [18:55:40] Can't exist in that table without at least one use [18:55:41] r [18:55:46] hmm [18:55:53] You know, you are right [18:55:59] I was overthinking as ususal [18:56:14] you sl good. :) [18:56:32] still no q on keyboard [18:56:36] very annoying [19:20:52] * Emufarmers reads guillom literarily. [19:21:21] * guillom writes on Emufarmers. [20:37:30] halfak it was SELECT DISTINCT ug_group FROM user_groups; btw :D [20:37:51] ... yeah [20:38:23] abusefilter, arbcom, bot, bureaucrat, checkuser, epcampus, epcoordinator, epinstructor, eponline [20:38:23] flow-bot, ipblock-exempt, rollbacker, sysop [20:38:38] for nl wiki [20:38:42] wow thats a lot of groups [22:19:39] halfak: in a flight and reading the workshop proposal [22:19:41] OMG SO EXCITED! [22:19:57] :D! [22:20:25] *I'm* excited about getting something submitted a whole 10 hours before the deadline with no stress. [22:20:34] Had way more stress trying to get people to come to meetings. [22:21:12] Anyway, it's gonna be good. :) [22:21:12] hah [22:21:26] halfak: sorr I didn't manage to get my bio in but I saw the email go by saying bios are optional [22:21:41] Yeah. Worked out just fine in the end. [22:21:50] I feel bad for trolling everyone for a bio. [22:22:02] We might still have the organizers insist that we put them in. [22:22:22] heh I bet I'll feel slightly insecure about mine :P [22:22:24] but oh well [22:23:21] halfak: I'm thinking I'll target 5 new datasets in quarry next week [22:27:51] YuviPanda, cool! I'm gonna have a really awesome one soon. [22:28:01] I crossed a major milestone with value-added research [22:28:15] (Assuming I don't find another problem :/) [22:29:20] halfak: nice. I'll maybe start with a standardized phab process for requesting a new dataset be added. [22:29:48] primary things required would perhaps be URL to it, a schema (that you can use a helper script to generate) and any indexes you'd want, *and* a license [22:30:00] I think having an explicit license from day 0 is important...