[10:24:09] hey guys [10:24:11] allah is doing [10:24:17] sun is not doing allah is doing [10:24:19] to accept islam say that i bear witness that there is no deity worthy of worship except allah and muhammad peace be upon him is his slave and messenger [16:53:23] o/ [16:53:39] Been head-down in working on my presentation for the showcase, so I forgot to wave earlier [19:26:20] hi everyone! The research showcase is starting in five minutes. Watch on YouTube here: https://www.youtube.com/watch?v=nmrlu5qTgyA [19:26:51] details on today's presentations here: https://www.mediawiki.org/wiki/Wikimedia_Research/Showcase#December_2016 [19:27:24] Welcome greenie! [19:27:43] I'll be your friendly moderator today. Please ask your questions in this channel, and I'll take copy them down to ask the presenters during the Q&A session. [19:27:48] T-3 minutes to start1 [19:28:20] Is anyone from WikiProject Women Scientists joining us today? [19:31:22] Research Showcase is starting now! [19:34:13] hi [19:34:30] hi Ziko. here for the research showcase? [19:34:37] yes [19:34:52] the finality of a wiki, great topic [19:35:08] agreed. [19:35:23] lmk if you have any questions for the presenters, and I'll pass them along during Q&A [19:35:34] ok [19:35:37] hey Ziko [19:36:30] problem of many wikis imho that they poorly define their objectives and "finality" [19:40:13] woo-hoo Dr Nettrom [19:40:30] \o/ [19:55:49] women scientists - well, a kind of biography. maybe the specialness lies here [19:56:43] how so, Ziko? [19:57:39] they say that biographies are "easy" article types to start as Wikipedian, as you don't have much to think about the structure [19:58:14] [19:59:54] i wonder how many FAs are biographies, in general [20:00:17] I had a similar question, Ziko. I'll try to ask it if we get time. [20:01:00] and biographies often are of a "serial" kind - you want to have all biographies of a specific group of people, e.g. all members of a legislative body. that may be the reason that those articles have only a lower "quality" [20:01:47] yeah, I'm wondering if we compared bios of non-female scientists, what the difference would be [20:04:22] o//// [20:04:22] second presenter (Andrea Forte of Drexel University) starting now… [20:04:36] J-Mo1, oooh. Can do [20:04:51] Thanks halfak, very interesting! [20:04:58] :D [20:05:18] I like DarTar's comment re: the different developmental dynamics of articles that start during edit-a-thons and other outreach initiatives too. [20:05:18] * DarTar waves at Ainali [20:05:39] * Ainali waves back to DarTar [20:06:02] BTW, Andrea's first Wikipedia work is one of the most cited Wikipedia papers. http://community.hciresearch.org/sites/community.hciresearch.org/files/Bryant05-BecomingWikipedian.pdf [20:06:13] Might actually be *the* most cited Wikipedia paper. [20:06:47] Cited by 718! [20:07:42] somewhat related to this talk https://www.mediawiki.org/wiki/Suggestor [20:08:24] ^ also related to ORES? [20:09:11] halfak: can be [20:09:20] depends, as a feature or a user [20:16:42] Amir1, am thinking that injecting an edit before it's saved would be a great way to help people patrol Tor user edits. [20:17:02] We have the facilities to fully-inject the details about an edit before it is saved right now :) [20:17:18] do you mean injecting a score? [20:17:42] Injecting a revision doc that looks like what the API would return. [20:18:31] You set "parentid" to that current revision of an article and it should all work. [20:18:46] Yeah, I think it's feasible. Not very fun in matter of cache :D but okay for limited use cases (tor users for example) [20:21:06] halfak: i'll probably be working on that over the holidays [20:21:15] \o/ [20:21:58] ANON BIAS! [20:22:23] privacy-enhancing tools are good for potential victims, but also for potential purpetrators? or does it not matter so much for the latter group? [20:22:38] J-Mo1, ^ [20:22:47] got it [20:23:33] temporal features, good point [20:24:37] [20:26:47] in our klexikon wiki (readers are children, writers adults) we accept only accounts with real names, with regard to the readers and reputation of the wiki - but maybe we should allow pseudonyms with the admins knowing the real names [20:28:34] right, harassers often don't care about non-anonimity [20:28:41] needs more measurements [20:28:42] thanks [20:29:57] Ziko I guess there's kind of a precedent for this model on WP, with CheckUser? [20:30:43] J-Mol : I don't understand? [20:30:59] https://www.irccloud.com/pastebin/OPUaSI3B/ [20:31:11] i read the paper and the press release, and i think this statement does not make sense: [20:31:11] "Wikipedia allows people to edit without an account, but it does not permit users to mask their IP addresses and blocks Tor users — except in special cases — so it is still possible to piece together an editor’s identity or location by looking at the things they’ve contributed." http://drexel.edu/now/archive/2016/October/Tor-Wikipedia-privacy/ [20:31:24] halfak: or with suggestor? [20:31:29] in that as a registered user your IP address is private, but a small set of trusted people have access to that information [20:31:35] arlolra, right :) [20:31:40] oh, yes [20:31:46] Open the gate, but build a safe space for review [20:31:50] but in klexikon the real name is the account name and public [20:31:58] ... the IP data (which Tor obfuscates) and the contributions (which are public by defaults) are entirely separate [20:32:19] and in klexikon you cannot give yourself an account, only get it via a mail to an admin [20:33:08] right. I was thinking that your proposal "maybe we should allow pseudonyms with the admins knowing the real names" was similar to CheckUser [20:33:13] ...and IMHO this also undermines the argument for relaxing restrictions on TOr editing. Any thoughts on this discrepancy? [20:34:03] real names and IPs are different things, but on the Internet maybe IPs are more important because you have to validate the real names [20:34:18] HaeB: is this a question for Andrea? [20:34:22] yes [20:34:30] if so, could you re-phrase with some context, and I'll ask it [20:34:54] J-Mo1: see above (posted this in several msgs) [20:35:01] indeed, HaeB, the editing profile might be the weakest link in the privacy question [20:35:08] okay, I'll post verbatim [20:35:36] start at the quote above [20:36:24] as long as editor reputation depends heavily on editing history / edit numers, people tend not to use different accounts for different topics in order to protect their privacy related to topics [20:38:00] andrea, what do you think about allowing only registered editing? [20:38:10] * J-Mo1 will ask this question [20:39:10] I have one clarification to make about data retention after the Q [20:39:45] cool. jump on in. [20:39:50] yes [20:40:00] are you going to finally admit we've been selling it to Putin all these years? [20:40:30] i don't like to use the word "anonymous" for unregistered editors actually :-) confusing [20:41:20] (just as a remark, regarding andreas's and rachel's reponse:) i think that mixes the very practical concern of public edit histories with the in almost all casess quite theoretical concern of subpoenas or checkuser abuse. [20:42:08] (also, i did not see the latter mentioned in the paper specifically, although of course it does not cover all the interview material) [20:46:06] ah interesting [20:46:20] (remark for the IRC discussion) I have looked at various of the previous proposals for enabling Tor editing, and in my observation they tend to show a lack of understanding of what kind of abuse (threat model) the Wikipedia community uses IP data to combat https://lists.wikimedia.org/pipermail/wikitech-l/2012-December/065359.html [20:46:40] Thanks folks! [20:46:41] thanks all, that was all super interesting [20:46:44] thanky you, good bye [20:46:45] it was great [20:46:49] thanks ppl [20:46:50] Thanks [20:46:53] thanks for coming [21:16:08] hm, https://meta.wikimedia.org/wiki/Research:Events hasn't been updatd in almost three monthts... :( [21:41:08] HaeB: That means you have three choices: 1) Wait for me to do it, 2) Wait for someone else to do it, 3) Do it :) [23:31:43] guillom: well concretely i have two choices at the moment: a) continue linking the page in the research newsletter b) stop linking it in the newsletter ;) 3) is not really an option since we failed to get contributions for this issue of the newsletter in general (for the first time in a long while. some are promised, but for the next issue), so i have to [23:31:43] write it up all by myself and don't have time to also search for event updates [23:33:20] HaeB, what do you think of scheduling some co-hacking time for working on the newsletter? [23:33:34] If I had something on my calendar, I think I'd contribute more consistently. [23:34:10] yes, a sprint might be a cool idea! [23:34:39] a big problem especially in recent month has been that the signpost publication schedule has become completely unpredictable [23:35:13] (we have been syndicating them since the newslettter was founded in 2011, so that was providing structure) [23:36:04] in theory of course one could just get the issue ready in advance (most content is not timely), but that's not how these things work out in practice ;) [23:40:31] HaeB, I keep a standing event on my calendar and my volunteers' calendars for Saturday at 9AM PST to hack on revscoring stuff. [23:40:35] It's sort of like my office hours. [23:40:54] I'd be interested in attending a "Research hack hours" that often involves newsletter work. :) [23:41:25] Is there much timezone overlap for the core contributors to the newsletter? [23:46:32] that might be tricky (in the past year or so two of the most frequent contributors have been from asia and africa respectively), but i guess it's worth a try even if not all can show up [23:47:46] i had also vaguely thought about repurposing the research documentation meeting to this (just for myself), but most weeks i can't attend anyway [23:51:32] Oh yeah! I'd do that :) [23:51:50] If you were there asking me to review a paper, I'd be far more likely to choose to do that ;) [23:52:14] Usually, it's just me, DarTar, AbbeyRipstra, and sometimes J-Mo.