[03:37:18] evening people [03:41:13] evening lzia :) [03:41:47] evening Ironholds. :-) how's it going? [03:42:14] second verse, same as the first! [03:42:29] ? [03:42:43] it goes as it always goes ;p [03:42:48] how are you? [03:43:57] haa. ... doing well. [03:44:46] anything I can do to help? :D [03:47:11] haha! :D not really. [03:47:14] I'm taking it easy [03:47:18] ;-) [03:48:53] good! [04:11:30] lzia, guess what I'm writing? [04:27:17] One step closer to streaming persistence measures. [04:27:24] And bed. [04:27:30] See you SF folks in the morning. :) [04:28:40] halfak, night! [04:28:54] also, I wrote a vectorised URL encoder/decoder that doesn't blow up on strings with ASCII NULs [11:25:38] Ironholds: you might like… http://thedoomthatcametopuppet.tumblr.com/ [12:40:36] YuviPanda, seen it! Is great :D [12:40:55] Ironholds: :) wonder what other mixtures would be nice [13:09:02] whee, my URL decoder/encoder works! [13:13:17] okay, no it doesn't [14:54:50] morning ottomata :) [14:55:55] morniing! [14:56:26] how goes? [15:10:29] yuss [15:10:42] 100k URLs decoded in 143ms [15:11:00] that is a two-order-of-magnitude improvement over the existing implementation [15:11:14] only moreso, because my implementation doesn't explode with out-of-range characters, it just ignores them. [16:05:39] Ironholds: it goes! busy morning (mondays usually are, so many emails sometimes!) [16:05:55] yup :D [16:06:06] also, i just drank my regular coffee sized jar of espresso [16:06:10] i am very caffeinated. [16:06:15] how goes up there? [16:17:44] ottomata, it goes well! [16:17:50] I wrote a URL toolkit for R in C++ earlier today [16:18:02] parsing, decoding and encoding, component extraction and parameter-to-value mapping so far. [16:18:32] one would think "hey, why do we need this?" - because R is stupid and half the documentation is "there is something that does this in baseR, it just breaks in these scenarios" [16:21:11] ha, awesome [16:21:17] sounds like someone is becoming a programmer [16:21:36] there is a small step between C++ and java! [16:21:43] and actually, it might be an easy step! [16:21:51] C++ is a paaain imho [16:22:00] :) [16:23:40] should we tell Ironholds about INTERCAL? just in case he needs more pain, that is [16:24:21] * Nettrom assumed Ironholds was part of the conversation, apoligies if he wasn't [16:24:22] I like C++! [16:24:25] what's wrong with it? [16:31:30] Often, you discover when it's too late. :P [16:41:11] Nemo_bis, oh, totally [16:41:26] I had wonderful fun this morning with std::string::npos versus -1 [16:41:31] C++11, screwing me yet again. [16:47:51] morning leila! [17:04:55] leila, do we have the checkin? [18:06:44] morning J-Mo [18:07:24] morning Oliver. How was your weeeknd? [18:08:17] * YuviPanda waves at J-Mo [18:08:31] long time no see! :) [18:08:42] hi YuviPanda! [18:09:16] J-Mo, second verse, same as the first. I wrote a lot of code, though. [18:09:21] and managed to bruise the shit out of my arm [18:09:55] ouch. sorry to hear that [18:10:23] eh, worth it! [18:10:28] And I upped my draw. Victory. [18:13:48] say, where's halfak? [18:13:51] he's normally conscious by now [18:14:27] hmmm. that's strange. he basically lives in this chan. Hope everything's all right. [18:15:57] Ironholds: guess what I’m doing this week? [18:16:28] debian packaging a C++ app! [18:16:37] stripping completely naked and going into the woods with a 60-inch recurve bow [18:16:45] hunting unceasingly until you find a tiger [18:16:50] that’s next week [18:16:55] my bow is on the way. [18:17:04] killing it and taking it back home between your teeth to demonstrate...I guess, that you don't mind hairballs and hypothermia [18:17:10] it’s an old english longbow, though. so hunting the french this time. [18:17:11] oh, you can borrow mine [18:17:20] longbows are a pain! Go for a recurve [18:17:24] J-Mo, I'll text him [18:17:29] no. must re-create AoE [18:17:51] AoE had recurves [18:17:57] also longbows [18:18:06] Ironholds—I'm sure he's fine. Halfak's a tough cookie. [18:18:37] or to put that another way: [18:18:47] breaking his arm couldn't get him off IRC, but /something has/ ;p [18:19:06] I'll text just to be sure. Nobody is allowed to hurt Aaon [18:19:07] *r [18:19:15] I need him to write the rebuttal statement for our WWW paper. [18:31:30] doh [18:31:35] J-Mo, apparently he is indeed at work [18:31:41] specifically, just got to the office ;p [18:58:49] hey halfak! You live! :D [18:59:06] Just barely [19:00:52] oh? [19:03:40] ping Helder [19:03:52] hi [19:03:56] meeting? [19:04:33] yep [19:12:58] Hey Helder [19:31:13] hey jonas_agx ! :-) [20:01:13] halfak, we doing our cross-coast head-putting today? [20:01:21] Yup [20:01:22] :) [20:01:34] yay! [20:01:36] okay, there in 1m [20:02:34] kk [20:12:55] yall seen this? so cool! http://ncase.me/polygons/ [20:13:41] lzia, you have a package upstairs :) [20:14:00] no I don't Ironholds [20:14:01] ;-) [20:19:09] leila, amazon disagrees! [20:20:57] I mean: I don't have anymore. ;-) [20:21:10] you really don't have the heart to remain as a secret santa. ;-) [20:23:43] Helder: let's talk this week about the revision service, it's a good idea schedule with Danilo as well [20:25:13] jonas_agx: I just sent you an e-mail [20:57:34] is diff text stored in the DB [20:57:35] ? [20:58:09] i.e. can i call the diff text as a query on the database, or must i make an api call with old and cur ids? [21:00:33] replicas DBs don't have text or diffs, only metadata, you need to use API [21:01:37] or dumps, if you want to parse a large amount of diffs [21:02:30] danilo, thanks [21:02:56] leila, wait, what? :P [21:03:01] the draw date was <= 20 December! [21:03:16] So I got it so it'd show up before then ;p [21:03:36] isn't part of the point that I don't know who has sent this to me? :D [21:03:40] cuz it seems now I know [21:04:12] and now that you've revealed this information, it's easier for others to know. ;-) [21:05:15] oh yeah. I reduced the list of valid tuples! Nooo! [21:05:18] *falls on sword* [21:05:26] and no, the point is nobody knows who's buying until stuff turns up [21:05:30] that's the secret! [21:06:01] from Wikipedia " Sometimes the gift-giver will personally give the recipient the present, thereby revealing their identity. " [21:08:07] is that so? [21:08:47] then we may as well reveal the information ahead of time. :D [21:09:04] but you know? who am I to question this process. Thanks! :-) [21:13:15] no problem! [21:13:19] they are some of my favourite books [21:13:34] Possibly mis-spelled words in DESCRIPTION: Vectorised (3:8) [21:13:38] lies. CRAN /lies/. [21:14:11] see. that's what I like. I'm happy I didn't constrain you too much. ;-) [21:31:07] leila, good! I hope you enjoy them :) [21:50:41] eeee [21:50:44] I LOVE OPS. [21:59:08] halfak, want some cool news? [21:59:12] yes [21:59:25] I emailed Ops asking about GeoIPV2. Their response was "we totally want that we just can't find someone to test it". [21:59:40] heh [21:59:41] so, we're getting IPv6 geolocation soon-ish so they can watch it blow my face off. Or something. [22:00:04] Wait... didn't we already have that? [22:00:20] ...bahyahaha oh you sweet man. [22:00:40] yes. In the sense that someone has handed us a revolver with five full chambers and we have a chance at life. [22:00:54] It works, but provides a valid country code in the old files about 1 in 200 times. [22:01:01] which is so rare as to be largely worthless. [22:49:01] Ironholds: I.. think I’ve successfully packaged a C++ application! [22:50:39] YuviPanda, congrats! [22:50:42] what did you package? [22:50:49] Ironholds: wdq.wmflabs.org [22:50:52] is written in C++ [22:50:56] cool! [22:51:09] I'm working with your department on some C++ acktually. It's gonna be fun. [22:51:16] yeah, I remember [22:51:17] good luck :) [22:51:19] and I'm about 30 seconds away from releasing my first C++ library on to CRAN. [22:51:26] you do? How? :p [22:51:29] it feels fairly trivial to package a C++ thing if it has a makefile [22:51:34] Ironholds: you were talking about it here earlier? [22:51:43] oh, cool [22:51:48] and yes, it is, but makefiles are a pain to write [22:51:58] oh, you think you won't need to write them by hand [22:52:07] hey, better than auto* [22:52:10] and then you find that boost puts it's static libraries in a REALLY STUPID PLACE that NetBeans can't find. [22:52:18] and then you scream [22:52:34] heh [22:53:04] then you wonder why you're using NB in the first place [22:53:16] indeed. [22:53:19] before remembering that KDevelop overwrites your preferred terminal client. [22:53:28] * YuviPanda murmurs Vim [22:53:53] or, https://www.jetbrains.com/clion/ [22:54:12] JETBRAINS 4EVA [23:01:30] heh [23:05:35] hey, J-Mo [23:05:51] hi fhocutt! [23:05:56] what's happening? [23:06:31] not too much, but I got the revamped bot basically working! [23:06:38] woot woot! [23:07:10] (still needs logging, config, etc., but the basic workflow of going by "added to category" seems to be working at a first pass) [23:07:25] nice [23:08:02] Cool! fhocutt where do I see the most recent version of your code? [23:09:31] halfak, the *most* recent version is still full of debugging cruft, so I haven't committed yet, but I've been putting the rewrite in here: https://github.com/fhocutt/grantsbot/tree/catuserfilter/matchbot [23:10:04] forked from a J-Mo project? [23:10:14] J-Mo, you told me you didn't do much code. Why would you lie? :( [23:10:21] Ironholds, yep. [23:10:30] when did I say that? I don't write much GOOD code… [23:10:47] Again... ;p [23:11:02] fhocutt, said it before, but: awesome to have you re-on-board, by the way :) [23:11:09] thanks, Ironholds! [23:11:15] I'm enjoying working on this [23:11:20] btw, Ironholds, did you see my fork? https://github.com/jtmorgan/geomo [23:11:39] J-Mo, no! But, see, I couldn't write that. [23:11:44] I know /some/ of those reserved words. [23:13:02] oh, some of you may be interested--I'm putting together a tech talk for early Jan on API client libraries and developer experience [23:13:18] AWESOME [23:13:36] says the author of his language's MW API client library [23:13:57] Let me know when it is? I should really attend. [23:14:26] it will *probably* be 1/6 [23:14:35] which language, btw? [23:14:42] R! [23:15:01] It was actually really disturbing; I ended up seeing requests in the requestlogs from something I'd built, that weren't from me. [23:15:34] Ironholds, excellent. [23:17:16] leila, re survey [23:17:24] yes, Ironholds [23:17:28] I'm going to be really nice and take advantage of my awesome internet connection and just put it on stat1003 [23:17:43] because otherwise you have to download a >100mb CSV, and ain't nobody got time for that [23:17:46] cool! thanks! I'm downloading it, but stat1003 is good [23:18:15] uploading! [23:19:01] dang, today has been a good day. [23:23:47] 3Quarry: Add date when query was last run - https://phabricator.wikimedia.org/T77941 (10Capt_Swing) 3NEW p:3Normal [23:27:56] 3Quarry: Make the query description window expanable - https://phabricator.wikimedia.org/T77945#832191 (10Capt_Swing) [23:29:29] Oh noes! I've been outed by a bot ^^. Nice feature. Does wikibugs only ping this chan for certain phab projects? [23:30:01] yeah, analytics-engineering, iirc. [23:30:07] I actually put it on ignore because it kept annoying me ;p [23:40:59] 3Quarry: Allow comments on queries - https://phabricator.wikimedia.org/T71543#832306 (10Capt_Swing) If we add comment functionality, we should also add some notification mechanism. The query author is unlikely to notice that someone has commented on their query if they don't get a ping of some sort. [23:41:28] 3Quarry: Allow comments on queries - https://phabricator.wikimedia.org/T71543#832312 (10yuvipanda) [23:42:32] 3Quarry: Allow comments on queries - https://phabricator.wikimedia.org/T71543#832317 (10Capt_Swing) yep... ugh. That's not going to be trivial. [23:47:16] hey halfak: did you end up documenting the PMID dump on the DataHub? I can’t find it [23:47:36] I didn't. It's just sitting on our datasets machine. [23:48:39] thank you for reminding me to document the session datasets :D [23:51:48] * halfak is going to keel over soon. [23:56:15] 3Quarry: Allow comments on queries - https://phabricator.wikimedia.org/T71543#832409 (10yuvipanda) Yup. only option is if flow is flexible enough, IMO.