[02:37:49] Ironholds: are you around? [14:53:35] morning [14:54:42] hello iron holder [14:55:05] o/ Ironholds [14:55:24] Hey harej :) [14:55:38] hello half alaska [14:56:11] pew pew pew [14:56:31] half kalashnikov [15:00:55] morning halfak :) [15:01:07] I just wrote up a diatribe about R that slams the CRAN maintainers. Great idea. [15:01:28] Hmmm... [15:01:40] But don't they make the decisions about what gets into CRAN? [15:03:29] <^d> #til when Ironholds wakes up he doesn't shower or eat, he diatribes [15:03:38] halfak, yup! [15:03:43] and that's why nobody talks shit to them [15:03:45] Ironholds: it's entirely reasonable to not want to come up with hypotheses before you do your data collection because you have no idea what correlations might exist and you want to base hypotheses on what's actually there, right? [15:03:58] harej, that's a halfak question! [15:04:06] halfak, luckily, we now have a route-around for CRAN [15:04:15] devtools::install_github("github_repo_containing_an_R_package") [15:04:17] This is the reputation you've built for yourself. I think of you as someone who knows statistics. [15:04:22] ^d, oh, I've been up for hours. [15:04:35] halfak, oh, outcome of the meeting with Nate, BTW [15:04:49] I'm gonna be helping out with the thanks research, he's putting me in touch with some students interested in geo stuff, and ah. [15:05:00] Apparently there's an MIT class that wants guest lecturers on social systems... [15:05:06] ...so that's probably happening. [15:05:16] harej, IMO, hypothesize all of the time -- especially *before* a quantitative analysis. It will help you be critical and curious of your results. [15:05:24] Ironholds, sweet! [15:05:58] Ironholds, guest lecturer, invited speaker, CRAN face-puncher. [15:06:04] Okay halfak. What if I were to say that my experimental hypothesis is how it would work in Ideal Land, but I am pretty sure the null hypothesis is actually what is happening? [15:06:49] harej, sounds like social science. Ideals are often not met :\ [15:06:57] * ^d collects data on the effect of finishing his coffee [15:07:02] harej, still tremendously valuable! [15:07:05] +1 [15:07:08] I was explaining this to a...person last night [15:07:18] science is not just finding the answer to things we don't know [15:07:24] science is also validating things that "everybody knows" [15:07:37] To know is most valuable of all. [15:07:45] because what "everybody knows" may not be true, and occasionally you'll find out that what "everybody knows", is totally wrong! [15:07:51] So, harej what are you hypothesizing about? [15:08:03] halfak, yeah, I'm really not sure what in the hells happened to my life. [15:08:31] My current hypothesis: Ha = "WikiProject average thread size correlates with editing activity in WikiProject subject area as defined by banner usage", and H0 = "No relationship whatsoever," and I am almost certain it's H0. [15:08:47] (Is hypothesis small-h or big-H?) [15:08:59] <^d> Remember 4H? [15:09:19] What about 4H? [15:09:30] <^d> You were talking about H's. [15:10:10] harej, IOW, ha is that there's a relationship between a project's content editing activity and its meta activity? [15:10:17] and h0 is that there's no real relationship? [15:10:21] Right. [15:10:24] gotcha [15:11:23] harej, it may be that there's too much extra-project activity about to capture the real project activity. [15:11:48] i.e. edits from randoms to project pages vs. projects-focused efforts from project members. [15:11:56] Well, rather than measuring overall edits, which is an imprecise proxy, I'm measuring the average number of edits in a single talk page thread, suggesting back-and-forth. [15:12:43] As opposed to someone spamming announcements, which is a reasonable use case but not indicative of collaboration. [15:13:52] Gotcha. I was thinking about the article editing activity though. [15:14:03] The other side of the correlation. [15:14:06] Ah. [15:14:11] Do you have a writeup? [15:14:18] No, I'm trying to work on that right now :P [15:14:20] I'd love to see plots/stats [15:14:22] Cool :) [15:14:41] May I suggest that you throw it on a research page on meta? [15:14:45] The stats collection hasn't happened yet and might not happen for a little while. As you said, I should construct a proper hypothesis first. [15:15:02] Na. I think you have one. [15:15:12] Now I do. :-) [15:15:15] There's no "proper" beyond, "i wrote it down" [15:15:26] Gotcha :) [15:15:34] I may have other hypotheses. [15:15:46] Can I still have something as my hypothesis and be almost confident it's wrong? [15:15:54] yes [15:15:55] :) [15:16:00] Perfect! [15:16:13] Hypothesis aren't supposed to be right. They are supposed to be interesting and testable. :) [15:17:44] Gotta get to packing. Back in a 15 minutes or so. [15:21:25] halfak, Hadley likes my diatribe [15:21:27] brb victory lap [15:25:40] aand twitter is going apeshit. sweet. [15:27:21] Also applies to Wikimedia-l [15:27:48] harej, hmn? [15:27:58] "Don't use Wikimedia-l" [15:28:11] aha [15:28:24] <^d> Friends don't let friends.... [15:28:38] <^d> halfak: My coffee hypothesis was correct. Finish it -> more awake [15:28:52] ^d, follow-up research question: what happens if you drink a second cup? [15:29:53] <^d> Have to test another day, I'm not thirsty now [15:32:13] I think we might be introducing a categorical bias based on your thirst patterns. [15:32:44] Since I was asked to create a Research page on Meta... I see some of these things mention IRB review. I'm not affiliated with a university so I don't have any IRB I could answer to, as far as I know. (Someone on our team is affiliated with a university but I am not sure how much work he will be doing.) What do I do in this situation? [15:33:53] harej, that template is overly specific to a certain type of research documentation. You can safely ignore that bit. [15:34:10] I *will*, however, be interviewing human subjects. [15:34:35] ahh. So there are some norms that I'm failing to get adequately documented. [15:35:41] The first part of the research is coming up with a ranked list of subject areas by activity and correspondingly a list of wikiprojects by activity (and correlational data between the two). the second part is interviewing the people who are named as contributors to those subject areas / projects (which anyone could get from edit histories anyway) [15:35:44] Basically, if you document your project (interview protocol, consent form, private data storage, etc.) on meta and invite wikipedians to comment -- and no major concerns are raise and not addressed --, then you are good to go [15:37:43] Are there nice boilerplates I can use for those things? ^_^ [15:38:03] See https://meta.wikimedia.org/wiki/Research:Means_and_methods_of_coordination_in_WikiProjects for a good, recent example. [15:38:12] Also, related to your work :) [15:38:54] I've been pointing those guys to WikiProject X :) [15:39:18] * halfak runs off for more packing [15:39:22] Yes, and I'm meeting with Dr. Gilbert in Seattle [15:43:11] Not a doc yet. ;) [15:52:19] so... [15:52:25] who do I talk to about last-minute travel-related problems? [15:52:45] Doreen probably [15:52:52] Also :( [15:54:04] My UC is flaring up and I'm kinda worried. [16:00:12] Ironholds, how long are you planning to stay in SF? [16:00:19] iff I make it, until the 28th [16:00:19] Assuming all goes as planned. [16:00:22] kk. [16:00:27] (see the mathematical if? SCIENCE) [16:01:03] Just so, you know, I don't think that you might still stay until the 28th if you don't make it. :) [16:01:44] hahah [16:02:03] I suspect that I'll either make it until the 28th or not at all :/. My gut is not a happy creature right now. [16:02:16] (disgusting biological details redacted for the sanity of the reader) [16:10:52] hi guys, is there some easy way to query for articles that have been featured on the main page of en.wikipedia? [16:11:03] * Ironholds thinks [16:11:15] not easy, like a single function easy. But it's doable! [16:11:26] actually I could probably just provide you with a list of page titles if that would be helpful [16:11:37] then you can retrieve the content for each one [16:11:41] would that work? [16:12:36] yes, that would be great! [16:12:53] cool! drop me an email so I know where to send it? okeyes@wikimedia.org [16:13:45] sent [16:14:06] so that would be a list of all articles that were ever featured on the main page, right? [16:14:11] ta! [16:14:27] yep :) Or at least, all articles with the pertinent (category|talkpage template0 [16:15:35] cool :) is it also possible to get the date of *when* they were featured? [16:18:47] Ironholds, there's a date on category links. [16:19:04] Assuming that the category wasn't remove/re-added, it should be pretty close. [16:19:11] sensible! [16:19:14] FaFlo, ^ [16:19:24] hokay, let's see what we can get [16:19:31] o/ FaFlo [16:19:56] nice! [16:20:55] so they are indeed put in a specific category? bc I couldn't find that [16:21:23] yeah, a hidden one; it's on the talkpage [16:21:26] (query building) [16:21:56] i see isee [16:21:58] Ironholds, what's a good way to see hidden categories? [16:22:27] halfak +1 ^ [16:22:33] I think there's a gadget or pref for displaying them? That might be bollocks, though. [16:22:57] It's hard to distinguish "stuff technically identifiable" and "stuff I know because I've been on Wikipedia for a decade in ~1month" [16:23:12] heh. [16:23:13] http://quarry.wmflabs.org/query/1633 [16:23:18] ^ That's my strategy [16:23:44] So that category must be "Featured_articles_that_have_appeared_on_the_main_page" [16:24:05] And cl_timestamp should *probably* contain the rough date. [16:24:44] Hmm... But the date for this particular article is 12/31 [16:24:47] That's not right. [16:24:54] It's on the main page today./ [16:25:11] I wonder if that category gets added when they *decide* to put it on the mainpage. [16:25:24] possibly! [16:25:33] FaFlo, ^ [16:25:33] FaFlo, dataset generated, will email :) [16:26:22] great, thx a lot :) [16:26:27] YGM! [16:27:06] so how did you generate that now? [16:27:47] is that reproducible ? :) [16:42:34] FaFlo, see the sql file attached! [16:42:40] it should be reproducible through the labs dbs, I think [16:43:46] Ironholds: didn't get the email yet, will check [16:45:20] huh! [16:45:30] definitely sent it [16:49:11] let's just wait a bit, maybe it's rush hour on the transatlantic cables [16:59:16] so, it arrived [16:59:24] aaron gave me a similar result with http://quarry.wmflabs.org/query/1633 [16:59:44] but we figured that it is not necessarily the same date when it was on the front page, right? [17:00:16] yeah, it's the date the tag was added, not the date it appeared. Womp womp :( [17:00:34] the only thing I can think of to extract the date it was added, is grab Talk:page_name, and parse the date out of the featured article template [17:01:27] ok, i see [17:01:54] well, anyway, thanks a lot Ironholds :) [17:02:07] np! [17:09:12] Ironholds, halfak, I've created a Research page. I have no idea if it makes any sense. https://meta.wikimedia.org/wiki/Research:WikiProjects_and_Subject_Area_Activity_(English_Wikipedia) [17:10:37] harej, Looks good to me. I'll give it a read an ask some questions. [17:10:53] It doesn't have to be totally fleshed out right away, right? [17:11:04] Indeed. It's a wiki after all :) [17:11:19] One of the pitfalls of grounded research is that you kinda have to make it up as you go along. [17:11:41] It's important that it is in a "good" state when you invite others to comment your recruitment/interview activities. [17:11:54] Indeed. I think that is just fine. Do you have an interview protocol that you could post? [17:12:11] I should have one soon. [17:14:35] Cool. [17:15:16] Also, I'd recommend noting that you guys have been through the IEG process under WMF Policies, Ethics... That process represents a pretty good vetting. [17:38:31] I'm leavin' on a jet plane. Don't know when I'll be back again. [17:38:33] o/ [17:38:37] safe trips [17:39:01] Ironholds: do you know if LZ is around? [17:39:22] I don;t [17:41:05] wondering if standup is happening today, given that people are traveling [17:41:20] I could use 20 mins to run to the office or I’ll be stuck at home for all morning meetings [17:48:52] gah, Ironholds: you guys really want me to stay at home today [17:48:57] yo lzia [17:49:02] haha [17:50:42] Ironholds, lzia: how about the three of us reconvene? I’d really like to get to the office [17:50:52] sure; I'm quite literally not going anywhere [17:51:12] Ironholds: aren’t you meant to be on a plane? [17:51:40] meanwhile lzia is trolling me and accepting all standup invitations [17:52:18] I’ll be back online in a moment /waves [17:52:56] sure [17:53:10] someone just send me the link when things are going on [18:30:31] leila, lmk when we're meeting [18:32:18] Ironholds: I don't think we are [18:33:05] okay, dario suggested we were. nevermind then. [19:34:49] hey Ironholds, we’re canceling the check-in [19:35:33] I’m here with Maryana and Jon and we’ll mostly talk about quarterly goals / any outstading requests [19:35:37] okay [19:35:46] hope you have a fun week [19:44:07] You know what you need to install in order to build python's matplotlib. [19:44:11] ^? [19:44:12] EVERYTHING [19:44:24] You must install everything else in the entire world first. [19:45:10] * halfak installs the 92nd package [19:47:14] and this is why we ggplot2 [19:47:32] I should just switch to bokeh now. [19:47:43] naw, bokeh is balls [19:47:44] I bet it has a similar set of dependencies though. [19:47:47] *holds up PUN sign* [19:48:00] bokeh == ggplot2 for python [19:48:07] How does bokeh + balls = pun? [19:48:34] Oh. :P that's not a ball. It's more of acircle [19:48:39] haha [19:49:04] aw, twitter, stop it [19:50:14] twitter is being bullshit again? [19:50:42] no, twitter decided to go apeshit in a really good way over something I wrote [19:51:04] Ironholds, please link to some of this ape fecal matter. [19:51:21] Uh oh. Boarding has started. [19:51:30] I'll have to ask for links from the bay. [19:51:32] okie! [19:51:35] o/ [19:51:51] OMG matplotlib installed [19:51:57] also https://thescienceweb.wordpress.com/2014/02/05/r-help-mailing-list-to-use-cage-fighting-to-resolve-conflicts/ [21:29:13] J-Mo: Did "WikiProject Pulse" ever become an actual thing? [21:30:00] nope. but the stuff that Michael G. is working on is an actual thing, and it is a successor to Pulse. [21:30:50] Terrific. WikiProject X is happy to review your proposals and do the dirty work of actually implementing them. [21:53:53] Are there view stats for portals? I know I could go to stats.grok.se and look up stats for individual portals, but I am more interested in what the most popular portals are. [21:53:59] (Portals on enwiki, to be clear.) [22:11:55] DarTar: yay for https://meta.wikimedia.org/wiki/Research:Wikimedia_referrer_policy [22:12:01] should be a simpleish change, tho [22:12:15] YuviPanda: yup [22:12:51] DarTar: I could write a patch sometime this week [22:12:53] I just wanted to make sure people inside and outside WMF are comfortable with the change from various angles (Legal, Tech, Analytics etc) [22:13:15] Jon offered to do the same, so I guess it’s whoever gets there first :) [22:13:24] I haven’t pinged him recently about this though [22:13:24] :D [22:13:34] I haven’t written any MW yet [22:13:38] well, for a while [22:13:40] :D [22:13:48] I guess it’s a patch that I could write [22:13:53] but yeah [22:14:03] if that’s what you’re asking, I trust you [22:15:22] YuviPanda: if/when you have a patch ready lmk, we should link it from that page [22:15:26] yeah [22:26:25] DarTar: can you add a bug? [22:26:43] YuviPanda: sure [23:04:15] YuviPanda: I’m creating a phab task and cc’ing you. Project? [23:06:00] YuviPanda: Wikimedia site requests? [23:09:58] YuviPanda: I left it uncategorized and unclaimed for now (and copied you) [23:34:09] YuviPanda: plenty of comments trickling in on the talk page if you’re curious [23:46:43] DarTar: i submitted a patch. doesn’t actually set it on our wikis but sets up the infra [23:46:52] I saw it, that was fast :)