[00:59:57] Ironholds, yt? [01:00:15] leila, yep [01:00:41] do you know which table has the variable that specifies whether a user can be contacted by email or not? [01:01:24] (actually, is there such an option, that the user say that I don't want to be contacted by the WMF? or it's just about other users contacting him/her?) [01:01:52] * Ironholds thinks [01:02:03] there's a user preference to enable/disable email [01:02:07] so user_properties [01:02:17] If you want to email individual users, you should talk to legal though [01:05:37] legal has approved [01:06:46] then yeah, I think it's in user_properties. [01:06:50] the emaila dddress is in user [01:14:22] thanks, Ironholds [01:14:27] np [14:56:31] morning! [15:13:14] halfak! [15:13:19] CHI REVIEW DAY! *dances* [15:13:38] Hey dude. So I've been hearing. :) [15:13:38] the paper I'm first author on uniformly got "probably reject, but might accept". I don't know how big of a deal this is. [15:13:40] Did you hear back yet? [15:14:31] "it doesn't make explicit design recommendations" - yep, we should do that. "it doesn't contain more assessments!" - it's a NOTE. We ran out of space. [15:15:49] So I'm going to write up a rebuttal, the basic argument of which is "we agree entirely with your rebuttals, but we don't think it's appropriate to build design recommendations into the paper yet because as you note this is a very nascent area of interest and it's hard to determine from the merely correlative evidence what people should do around design. However, we do want to include ideas for further avenues of research that could p [15:15:50] roduce such recommendations." [15:15:56] no idea if this is a good idea ;p [15:24:14] morning heatherw :) [15:26:11] morning :) [16:08:45] yuvipanda, I gots a thing for you you might enjoy. [16:09:21] hmm? [16:09:23] tlel [16:09:25] me [16:09:38] ever debianised C++ before? [16:09:42] ah [16:09:45] thankfully not [16:09:48] :D [16:09:58] do need to at some point, though. [16:10:29] would you like an example use case? :D [16:11:44] if you debianise a library I'm working with and otto puts it on the stats machines, we get hyperfast UA parsing [16:11:53] faster-than-the-python-version fast. That's pretty fast. [16:12:43] heh [16:12:52] source? [16:13:16] ottomata would probably be faster, though, depending on how quickly you want it [16:14:10] yeah, but I need to tweak it [16:14:21] and it's gonna be https://github.com/ua-parser/uap-cpp with some tweaks [16:14:55] namely that screw having it as a command-line thing, I want it as an importable library [16:14:58] not entirely sure how to do that. [16:15:02] but hey, we'll find out [16:17:17] heh [16:17:22] do you want to use it in python? [16:17:29] or just straight C++? [16:18:22] the latter, there's already a python library [16:18:31] right [16:18:39] what I actually want to do is integrate it with R through RCpp, in the same way I integrated our geolocation code [16:18:43] because HOLY SHIT it's so fast. [16:19:03] heh [16:19:08] So I started off by experimenting with compiling that library as a component of WMUtils (which would load it), but swiftly ran into make problems. [16:19:20] yuvipanda, no seriously - https://gist.github.com/Ironholds/38ba1e27017e544925df [16:19:26] c_geo_* is C, geo_* is python [16:19:26] yeah, I heard :) [16:19:30] oh [16:19:31] yeah [16:19:32] that [16:19:43] nice [16:19:48] it's pure regex matching without boxing overhead [16:19:50] so would be fast [16:19:54] so yay [16:19:56] Ironholds, I think your rebuttal strategy sounds solid. [16:20:10] This is a common fight around notes. [16:20:36] "Needs more stuff" -- "But there's no room for more stuff. This is the most important stuff." [16:20:45] yay! [16:20:52] yuvipanda, boxing overhead? [16:21:00] with python, that is [16:21:21] Also, if you have good reviewers, they will appreciate the argument re. design recommendations. [16:21:24] ahh [16:21:30] halfak, I hope so. One of them appears to be an idiot. [16:21:36] That'll happen [16:21:39] "the paper opens with.." "no it doesn't. I know, I wrote it." [16:21:39] Reviewer #3? [16:21:46] naw, #2. [16:21:59] Who also totally invalidated the idea of anonymity by including a load of idioms in their text. tsk. [16:22:19] "aiya!" as an exclamation is a cultural identifier. [16:22:43] * yuvipanda does the needful [16:22:45] still, the feedback is good overall. Although I was sad to see that most people ticked the "passing familiarity only" box, it seems. [16:22:51] yuvipanda, okay, so I stole THAT line, but... [16:23:31] Ironholds: the bot that announces deployments also says do the needful [16:23:55] neat! [17:14:02] aargh [17:14:15] yuvipanda, the debianisation of the C++ library; something you might be interested in working on? [17:14:25] I don't want main.cpp, just the header files and parsers. [17:14:31] main I can construct more usefully meself [17:14:33] * Ironholds pops knuckles [17:14:58] depends on how quickly you want it, Ironholds :) [17:15:06] give me a timeline and *then* I will tell you [17:15:09] of course I'm interested, etc. [17:15:17] "end of the weekend so I can hack on it"? [17:15:36] ah, yeah, probably not :) [17:15:43] okay. Then, whenever you can! [17:15:48] we can rely on python until whenever you're free :) [17:17:03] alternately...hmmnmnmnmn. [17:17:07] :D [17:17:12] also file a phab ticket? [18:32:43] darnit, I have a thing for ellery and he ain't online [18:42:28] ewulczyn: i need to restart hive-server, i see you are runnign a qquery [18:42:37] can you let me know whent it is done (and not start another) so I can do that real quick? [18:45:23] ottomata: ok, go ahead [18:45:56] ewulczyn, wanna chat retrieving the donate clickthroughs, btw? [18:46:20] Ironholds:yes [18:46:37] k, tahnks ewulczyn [18:46:40] restarted, you may continue [18:46:48] I am adding to my todo to find a beefier home for hive server! [19:01:27] * Ironholds fumes at email [19:02:21] * yuvipanda pats Ironholds [19:26:44] Ironholds: are you going to be presenting at the next Monthly Metrics? Or Toby? [19:27:05] I won’t be around (at a workshop) but happy to help prepare the deck of course [19:30:06] DarTar, December? [19:30:22] Make Toby do it? It sounds better from him. [19:33:23] I’ll drop you both a line [19:45:33] Ironholds: are we good here? https://trello.com/c/VdOwP6yD/521-state-of-traffic-data I want to close this card [19:45:39] see my note [19:47:06] yup! [19:48:41] can you quickly add a note on this bot overview thing and mark it as completed if it is? [19:49:18] Ironholds: or just tick it as complete [19:49:26] totally [19:49:40] boom! Done [19:49:59] danke [20:46:51] yay! [20:47:00] the hackathon in NY is working on gender data around Wikipedia. [20:47:09] I did my duty and pointed them at https://boards.greenhouse.io/wikimedia/jobs/32299?t=k6n8vo [20:50:06] also, halfak|Mobile, I got you a few more utils users [20:50:42] Python utils? [20:50:47] yup [20:50:56] they wanted to detect reverts. I went "cool! I know just how..." [20:50:58] Wool :) [20:51:12] -l [20:51:23] Just about to take off for Chicago [20:52:01] have fun! [20:52:02] I get to hang out work Geiger tonight [20:52:28] And convince comms people to study Wikipedia tomorrow :) [20:52:48] Here we go [23:03:21] lzia: https://docs.google.com/a/wikimedia.org/spreadsheets/d/1CLz7QS0HOBrX-OE2d8-Xa9DAkyM9ZlRaBHjTcTglies/edit#gid=320674717 [23:03:26] DarTar: ^ [23:03:38] thanks, Deskana