[15:48:07] hey Ting, are you here? [16:02:35] * tommorris spies a sgardner. [16:04:02] tommorris, good afternoon! [16:04:14] (I am in a Board meeting in another channel. FDC vote. :-)) [16:04:37] * tommorris is writing code while being cold. [16:21:09] hey does anyone in this channel have the fundraiser totals chart URL? [16:21:27] (The Board is praising Zack in the other channel, and I want to show them the totals :-)) [16:22:49] sgardner: https://wikimediafoundation.org/wiki/Special:FundraiserStatistics [16:23:17] yay, thanks tommorris :-) [16:23:24] day 17 - $12.85m total [16:23:27] https://wikimediafoundation.org/wiki/Special:FundraiserStatistics [16:23:31] oops :-) [16:23:34] wrong channel :-) [17:53:43] Hello, does the foundation have an opinion on the proposal of ITU control of the internet? [18:54:59] Is there E3 office hours here in 15 mins? [18:55:11] ehr, 5 mins, even. [18:59:59] Hi everyone [19:01:08] Who's here for office hours? [19:03:27] * siebrand raises his hand. [19:03:46] Hey siebrand [19:04:30] (doing other stuff too, was planning on lurking) [19:04:34] Okay [19:04:42] I'll give it a minute, for others to show [19:05:04] * jorm lurks, too. [19:05:21] hello. it happened some weeks ago that our new employee sindy pointed out to me that the registration form for a new account has big flaws, such as that it does not recognize wether the name is already is use... [19:05:22] I'm here! Although I haven't been paying any attention to enwp lately. [19:05:44] WHERE IS MY FACEBOOK CONNECT [19:05:53] * domas contributes to login page redesign [19:06:03] * siebrand greets domas. [19:06:04] Heh [19:06:04] * Isarra murders domas. [19:06:17] And this is why we need a RESOLVED-MURDERED resolution. [19:06:24] Precisely [19:06:30] hi stevenw [19:06:37] who is isarra and why isarra thinks murdering me is that easy [19:06:58] Isarra is crazy people who likes pointy things. [19:07:16] Ziko: which wiki was she registering for? [19:07:17] ...and who usually has better grammar than that. [19:07:22] nl.wp [19:08:11] Yeah the current state of things is something of a hodgepodge. There is validation of the username by the form itself, but also by extensions like AntiSpoof, which tries to keep people from having names too similar to other accounts. [19:09:45] If you take a look at the current form on English Wikipedia, we hacked together a test version of client-side validation, which warns people about errors before they submit for the most part. [19:10:09] Does antispoof work with SUL, too? [19:10:21] Not sure. [19:10:32] guessing that it only checks on first signup [19:11:05] Hi. [19:11:08] Hi [19:11:15] Hi Steven, Ebe123 [19:11:17] not sure what you mean "first signup" Hello71? [19:11:40] As in, first signup, rather than auto-creation of the account via SUL? [19:11:41] StevenW: I'm guessing that it only checks against local usernames on the home wiki. [19:11:42] Sockpuppets [19:12:02] Ah gotcha. [19:12:11] taking a look at the source [19:13:46] nvm [19:13:47] see https://bugzilla.wikimedia.org/show_bug.cgi?id=28747 [19:14:05] i thought there was a plan to do full unification? [19:14:11] and then it will check SUL by default. [19:14:41] Yeah not sure. We should ask James Forrester on that one. ;) [19:14:42] #28747 says it checks global DB for spoofing [19:15:05] Anyway. The current plan for us is to redesign registration from a visual standpoint, but not muck with the server-side validation. [19:15:43] We A/B tested a version with decent client-side validation against the current state of affairs, and the measurable gains in people creating new accounts was quite small. [19:16:52] We'd end up with a few hundred extra registered people on English Wikipedia, for example, but it might take an extra month to make AntiSpoof and the rest of the current set of checks play nice and provide people validation before they submit. [19:17:48] Thankfully we might still get an account creation API, thanks to a volunteer dev who has one waiting in the wings: https://gerrit.wikimedia.org/r/#/c/18127/ [19:18:52] Ebe123: actually there's nothing automatic that checks for sockpuppets. AntiSpoof and the like just look for usernames that are too similar to existing ones. [19:19:05] Ok [19:19:59] Has anybody tried out the new form? [19:20:18] How long ago did you put it in? [19:21:14] It's been at 50% for a couple weeks, but only at 100% since Thursday. We decided to turn the test version on fully to see how it performs at scale, while we build the permanent version. [19:21:43] How does it perform? [19:21:50] Ah; I was going to say I created a new alt account ~Nov. 8th and I didn't notice anything different. [19:22:14] * The_Blade goes to test it out now... [19:22:47] Isarra: no new bugs, but the jury is out on whether we'll see a larger or smaller number of additional people registering. [19:23:01] Neat. [19:23:12] Part of the reason we did it was also that since the start of the fundraiser, there have actually been large spikes in new registrations. [19:23:15] which jury? [19:23:31] ha. The jury of data. [19:23:34] http://toolserver.org/~DarTar/reg2/ [19:23:47] You can see the overall numbers from enwp there ^ [19:24:34] * StevenW thinks that if domas wants Facebook Connect, he should submit a patch. ;) [19:24:41] and see how it goes [19:24:54] Wow, that /is/ different. Looks like an improvement to me. [19:25:02] Brb [19:25:02] Glad to hear it! [19:26:15] stevenw: there're quite a few plugins by now, I think [19:27:06] For sure. I was just joking though. There are people who would murder us if implemented Facebook Connect or other social signup buttons. [19:27:11] eww Facebook [19:28:04] indeed [19:28:45] stevenw: there're amish people out there too [19:28:54] :) [19:29:03] Why would it be such a bad thing to use facebook connect and the like if people want the option? [19:29:13] isarra: exactly [19:29:15] Facebook? [19:29:17] No [19:29:19] ! [19:29:22] Although there might be privacy concerns with that... [19:29:33] isarra: what kind of concerns, if that is voluntary? [19:29:39] Yeah it doesn't jive with the current edition of the privacy policy, AFAIK. [19:29:42] I don't know! But there might be some. [19:29:47] Okay. [19:29:53] That's a reason. [19:29:58] what kind of? [19:30:11] technically you can use whatever third party site, with voluntary acceptance of use [19:30:18] no need to embed objects for everyone [19:30:34] e.g. load remote assets for login only when someone requests them, set a cookie to use remote auth, etc [19:30:48] but what qualifies as request [19:30:59] facebook integration would be bad [19:31:01] on login page clicking a button 'sign me in with X site' [19:31:04] soapy: why? [19:31:06] and imo even a facebook button looks ugly [19:31:08] Mmm, perhaps this would be something well worth discussing, but perhaps this isn't the time or place. [19:31:17] when is the time and place? :) [19:31:24] The technical parts are far from impossible and could be worked out. But the big "if" is volunteer acceptance. [19:31:24] I have no idea. [19:31:45] stevenw: if 5% people would use Google, 5% would use twitter and 5% would use facebook to log in [19:31:51] why do the rest 85% have a way about that? [19:31:56] a say [19:31:56] that is [19:31:57] well for one thing facebook officially requires you to use your real name, so that would be bad for anonymity even if voluntary [19:31:57] buttons for social media are not important. but it is important that we do something against barriers for participation. so the new log in form does not make a big difference in numbers, i understood? [19:32:04] also though it would make us a commercial website [19:32:08] This leads me to what we're working on next though. [19:32:11] soapy: _voluntary_, I go by full name in many places [19:32:26] From a volunteer engagement perspective, we don't actually need a ton more accounts registering. [19:32:44] people have accounts elsewhere [19:32:50] it would be nice, but what we need is more people registering _who contribute_ [19:32:51] people are getting used to their identities being centralized [19:32:55] how would third party usernames look like internally? [19:33:03] If we wanted that, all we'd have to do is turn off Filter 159 [19:33:07] right now, on English Wikipedia for instance, only 25-30% of accounts ever edit. Even once. [19:33:17] What about on Wiktionary? [19:33:28] How do the proportions compare cross-project? [19:33:32] Not sure Isarra. [19:33:41] Might be something to look into. [19:33:44] yes. [19:34:19] Some projects I seem to remember it is higher. Commons for example, where a lot of the people with new accounts are editors from other projects. [19:34:33] Meta I'd think too. [19:34:38] right [19:34:49] stevenw: if you want editor percentage to be higher, make login/account form more complicated [19:35:03] if you want more users, make it easier and more beneficial to create accounts [19:35:12] what is the aim, again? ;-) [19:35:31] that is quite true. Making it harder does increase the "quality" since only the most dedicated make it through the gauntlet. [19:35:41] Back. [19:35:58] this is interesting, some day I will make a summary of most of contributions of people in this debate [19:36:00] X? NO! [19:36:17] tells a lot about the community :) [19:36:36] that kind of resistance to change isn't necessarily a bad thing [19:36:38] stevenw: you want to increase participation/ [19:36:42] ? [19:36:44] ergh [19:36:54] make it that it is an action filled with negativism [19:36:58] i wonder why people create an account and then don't use it. possibly, because they see only after registering the wiki syntax in the articles? [19:37:17] We have a little bit of knowledge about the Ziko. [19:37:37] hehe [19:37:42] From past surveys, from looking at stats of where people were on-wiki before registering, and from a new survey. [19:37:54] looking at your userpage I see that you work for Facebook [19:38:08] He does? O_o [19:38:12] omg, really? [19:38:22] hahah [19:38:43] >_< [19:38:47] domas also has been around Wikimedia for a little while ;) [19:38:50] or to use gadgets and/or set preferences [19:38:57] Anyway, there are quite a few people who are readers or donors, who either just wanted to be "a part of Wikipedia" or thought they would get extra reader-oriented features. [19:39:23] And there's another contingent of people who sign up with the thought that they might edit some day, and want to have an account just in case. [19:39:47] or maybe some people forget their passwords [19:39:51] that too [19:39:53] because they didn't use SSO [19:39:55] (HAHA!) [19:40:33] the fact that we don't require email, and it's the only way to retrieve your pswd, does mean we get some of that kind of registration [19:40:54] tbh, I had some of these discussions for years with various people at various wikimedia-related entities [19:41:03] and usually it is "yeah, it would be a good idea, but community..." [19:41:26] we should do X! yeah, sure, but community won't allow us! [19:41:44] thats how you end with community of 1000 people [19:41:52] one of the things about working at the Foundation, is that you either have to guess when that is just not true, or you can go ask people. Like in office hours. ;) [19:42:01] even though it's a small sample obviously [19:42:26] anyway, you want to make login easy? make it frictionless [19:42:33] I'll be sure to publish the survey results, since people seem curious about who all these non-editing accounts are. :) [19:42:49] I have a non-editing account [19:42:56] so that people can look at my page and make useless remarks [19:43:20] The people love you domas. Don't deny them their remarks. [19:43:29] ok! [19:44:49] Preferences are scary on enwp. Do people ever get scared off by them? [19:44:58] Or do they just sign up and not use them? [19:45:05] you can't take lots of them away, community will fight you [19:45:09] not *that* scary [19:45:25] lots of the preferences make your wikimedia experience very very shitty [19:45:26] I counted over 40 gadgets in one horrible long list. [19:45:43] Scary. [19:45:45] at least nowadays some of them cause a warning [19:45:47] don't they [19:45:57] I doubt people actually leave in any large numbers because the preferences are horrible. But it certainly would be nice to reduce the clutter. [19:46:07] there is actually a thing going on in the developer community to remove preferences. [19:46:21] * StevenW goes to find that RFC on mediawiki.org [19:46:22] for the most part, i want to kill them with fire. [19:46:36] should probably at least reorder them [19:46:49] https://www.mediawiki.org/wiki/Requests_for_comment/Core_user_preferences [19:46:51] are there stats on skin change? [19:47:06] Yes. Though that's a whole 'nother discussion entirely. [19:47:09] JORM! Fancy meeting you here [19:47:15] * Jan-Bart slips Jorm a dollar [19:47:17] hey, it's jan-bart. [19:47:21] * jorm pockets said dollar. [19:47:21] I didn't get that greeting from jan-bart [19:47:27] :( [19:47:33] DOMAS! [19:47:38] YOU ! [19:47:39] yo [19:47:39] last time i saw domas i gave him a running, jumping hug. [19:47:39] HERE! [19:47:49] it made him uncomfortable. [19:47:55] you think so? [19:47:57] * Jan-Bart takes options from Domas… gives them to Jorm [19:48:06] jan-bart: don't have any, they're RSUs [19:48:20] * domas grins [19:48:25] Ziko, to answer you from earlier ("so the new log in form does not make a big difference in numbers, i understood?") [19:48:31] * Jan-Bart takes whatever the #$*&(# RSU's are and gives them to Jorm…. just in case [19:49:04] The answer is that it makes a smallish difference, though at the scale of thousands of registrations that is enough to justify some change, I think. [19:49:27] The numbers are all at https://meta.wikimedia.org/wiki/Research:Account_creation_UX [19:49:30] yes [19:49:34] once I log in on enwiki, I get this sandbox link at the top [19:49:35] what is that? [19:49:39] why would I need a sandbox? [19:49:44] I prefer diamonds [19:49:44] Ah, im not late am I? [19:49:44] It's a custom thing for English Wikipedia [19:49:58] and pt.wiki [19:49:59] do we have numbers about how many people try to make an account, fail because of an already used name, and then leave? [19:50:02] gwickwire: we end in about ~10 mins. [19:50:08] that was a community added gadget, the sandbox link. i hate it. [19:50:26] jorm: I'd guess [19:50:31] the sandbox is a custom thing for most wikis, i think. [19:50:32] Ziko: we measured the overall error rate, but not the specific reason for the error. [19:50:37] Is the DR one still going to go on, or is this to be the DR one? [19:50:44] theres still one [19:50:44] Ziko: it also inaugurates what i expect will be a period of split-testing UI changes for the entire login / signup process [19:50:48] jorm, the sandbox really does not make sence because the newbie is not the only one who is editing there [19:50:48] jorm: Stop saying things I agree with. It's freaking me out. [19:50:50] it's there for pl.wiki, and i never heard anyone complain about it [19:50:54] its in 10 mins gwickwire [19:50:54] gwickwire: DR discussion is in about 10 mins. [19:50:56] :) [19:51:05] :| [19:51:08] From the sandboxes of new users, I've probably learned half of all Pakistani business phone numbers [19:51:17] Ziko: we have data on how many people try and register a username too similar to an existing one (triggering anti spoof rules), but that's probably not what you're asking [19:51:21] Ebe123: StevenW got the spot I wanted :P [19:51:30] :( [19:51:34] jorm: what do you think that when I go to my contributions, I have a link that allows to block myself? [19:51:45] And I suppose throwing the Stevens in a pit to fight it out would have been improper. [19:51:45] and 'email this user' [19:51:46] The_Blade: http://www.youtube.com/watch?v=zyu2jAD6sdo [19:51:50] the problem with the sandbox link is that it's interpolated into the toolbar on DOMContentLoaded, so you get this annoying thing where you think you're clicking on on your talk page, but it shifts to the left at the last second and you end up somewhere else. [19:52:04] DarTar, that's going into the direction [19:52:05] Heh. [19:52:11] ori-l: thats what community wants you to do [19:52:11] :) [19:52:13] ori-l: And it's there by default! [19:52:14] so shudup and listen [19:52:17] Isn't it great? [19:52:26] gosh, I still have my abilities [19:52:32] not really [19:52:34] pointed out two things and got people discussing about that! [19:52:41] hm? [19:52:49] domas: if i get it removed, you buy me a beer at wikimania 2013 [19:52:52] hello hello71, you must be new here [19:52:52] what was the initial reason for a new login form, a human who saw the flaws or something in the data? [19:52:52] deal? [19:53:07] Ziko: we haven't published error data, we might do so now that we have the new version up at 100% [19:53:21] mechanism to start doing a/b testing on post-account creation activities. [19:53:30] I think we have it for the third test, DarTar? Just not on-wiki yet. [19:53:36] stevenw: how many people don't login after seeing login form? [19:53:47] the % of login shows vs number of successful logins? [19:53:52] ha - correct, for acux_3 [19:53:58] No idea. Yet. [19:54:12] stevenw: so why do you have a meeting about it?! [19:54:20] oh you mean registration? [19:54:22] domas: do you mean sign up or login? [19:54:24] both [19:54:43] for login we haven't tracked anything (yet), just for signup [19:54:45] we didn't measure login bounce rates [19:54:46] heh, maybe I should have a session with you about logical event aggregations/statistics/tracking [19:54:50] 2013 is soon [19:54:52] and you don't have such data [19:55:06] domas: feel free to swing by for a chat ;) [19:55:11] (and it would be straightforward to implement) [19:55:35] mind you, we may have some surprising news [19:55:49] Speaking of login, here's the docs about our design ideas for that. https://www.mediawiki.org/wiki/Wikimedia_Foundation_Design/Login [19:55:53] like what? that you are going to have a bigass monster to answer simple questions? [19:56:10] no, no cephalopods [19:56:42] no nonsense event logging, I don't know if StevenW already mentioned that [19:56:46] domas: i think you're looking for another team :) i do have bounce rate on login form data somewhere -- we collected that for a while [19:57:01] agree you don't need mapreduce to increment a counter [19:57:22] ori-l: I'm talking more about general direction about how to instrument large set of software to have real-time data for things like that [19:57:34] there's stuff like profiler now that could get you some of that data [19:57:44] but I think technical mediawiki profiling and logical site behavior should be on two different paths [19:57:50] so you could answer these questions easily [19:57:55] and not MR'ing [19:58:07] in a few :) [19:58:15] domas: i'd love to give you a quick overview of what we have in place and what we'd like to build, and to get your input [19:58:19] sure [19:58:33] let's chat about it afterwards (or now, elsewhere, if you wish) [19:58:49] where? [19:58:53] #wikimedia-e3 [19:59:30] Yeah for anyone who wants to chat, that's our team's channel. We're in there pretty much all the time, since ori-l is a vampire and doesn't sleep. [19:59:54] Okay we've got about two minutes before Steven_Zhang kicks us out. Does anyone have any last questions? [20:00:03] Now for DR [20:00:11] lol, I'm not kicking people out :) [20:00:21] that's TBloemink's job ;) [20:00:28] Aw.. Why not? [20:00:31] ori-l is a vampire, eh? This explains so very much. [20:00:42] * TBloemink is evil [20:00:55] * gwickwire is not evil. [20:01:00] Thanks for chatting, all. Please bug us in #wikimedia-e3 or on-wiki if you have further questions/comments. [20:01:13] OK, I think that's my queue. [20:01:18] Hi all :-) [20:01:22] Hi [20:01:27] Cue. [20:01:31] I mean, hi. [20:01:44] Hello. [20:01:45] So, first, an update on my fellowship since the previous office hour session [20:01:47] hi [20:01:55] And how was it [20:02:10] What did you accomplish [20:02:14] My last office hour you were in, silly :) [20:02:16] (WP:WQA) [20:02:33] ahhh its actually office hours… that explains so much... [20:02:33] So, thus far, a survey was done on volunteers ([[WP:DRSURVEY]]) [20:02:45] as well as participants, that one [20:03:12] Should we get WP:GEO closed? [20:03:21] An analysis was done on active dispute resolution forums [20:03:27] http://en.wikipedia.org/wiki/Wikipedia:Dispute_Resolution_Improvement_Project [20:03:48] hi [20:03:52] And changes were made to the dispute resolution noticeboard, to make it more effective and efficient [20:03:55] DRIP, eh [20:03:57] (Also on that page) [20:04:03] heh, yeah, DRIP [20:04:04] :P [20:04:34] The DR wizard in use at DRN has had some positive effect on cases in general [20:04:45] How? [20:04:50] Has anyone here seen/used it before? (What are your thoughts?) [20:05:01] I've seen it, not used it [20:05:08] DR wizard? I'm not sure what that is/ [20:05:17] Ebe123: Cases weren't as TL;DR, cases resolved faster etc. [20:05:48] Seen not used. My opinion of such things is biased by my technical background, so not useful for something meant as a user interaction component. [20:06:00] The statement "cases resolved faster" might be bad [20:06:14] It isn't a good thing [20:06:18] survey results are kinda scary. [20:06:21] in some cases [20:06:53] These were resolved properly, but I agree that we shouldn't race to a finish [20:06:55] :) [20:07:05] But the key thing is it makes it simpler [20:07:12] How? [20:07:14] No complex wikicode needed to file a case. [20:07:22] The 2nd party must use the wiki [20:07:24] Coren: ArbCom could take a hint from that ;) [20:07:41] Ebe123: there's only so much we can do. [20:07:59] Steven_Zhang: Sorry for the dumb quesiton, but how was success of a case determined? [20:08:13] odie5533: Not a dumb question :) [20:08:16] If the parties came to a agreement [20:08:32] Steven_Zhang: Is it reasonable to expect that the intersection of people who will use DR in general and those who will reach ArbCom is significant? [20:08:35] And I followed that up at the article/article talk [20:08:51] Coren: the longer a dispute takes, the less likely it is to be resolved [20:08:58] that's not the case all the time [20:09:21] But the further it gets up the DR chain, the more likely a banhammer will be needed. [20:09:50] Steven_Zhang: Wait. How did you reach that conclusion by contrast to, for example, "The less tractable a dispute is, the longer it will last"? [20:10:14] "But the further it gets up the DR chain, the more likely a banhammer will be needed." the banhammer would be needed at the start [20:10:28] No, not necessarialy [20:10:34] How? [20:10:39] If its something minor and the participants come to a resolution early [20:10:45] then it can be resolved early [20:10:57] if it's an israel/palestine dispute for example [20:11:04] it'll probably go to ArbCom [20:11:13] because they have ingrained opinions [20:11:19] Hasn't it already [20:11:27] The next arbcom might not have ingrained opinions, though. [20:11:46] No, the editors have ingrained opinions, not ArbCom [20:11:50] Oh! :) [20:11:51] though some may argue for both. [20:11:54] Anyways [20:12:07] hey, did anyone suggest facebook login yet?! [20:12:18] No [20:12:20] As a result of the trial at DRN, it has been proposed that this be implemented in a universal DR wizard, for all forums [20:12:28] http://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Reforming_dispute_resolution [20:12:36] Which was successful :) [20:12:43] It will take time [20:12:50] That's currently being developed by a dev. [20:12:52] Then another RfC will be needed [20:13:03] Nah, not exactly. [20:13:09] But the underlying issue affecting DR remains. [20:13:13] CONSENSUSCANCHANGE [20:13:16] Shortage of volunteers. [20:13:23] Lol @ the facebook login [20:13:28] Is dispute resolution intended for incredibly silly things that shouldn't even be disputes at all? [20:13:30] Not always bad [20:13:44] Not always Isarra [20:13:54] Ebe123: not enough volunteers leads to disputes being unattended [20:13:59] or to volunteer burnout [20:14:12] But not always [20:14:22] Ebe123: Redundancy is good. [20:14:24] Where do you describe what changes were implemented to DRN between May and August 1? [20:14:30] We shouldn't have too many [20:14:31] To try and learn more from existing volunteers, I'm currently undertaking a second survey. [20:14:38] Too many is not bad. [20:14:42] Too little is bad. [20:15:02] odie5533: http://en.wikipedia.org/wiki/Wikipedia:Dispute_Resolution_Improvement_Project [20:15:04] A dispute of 2 editors with 10 volunteers coming [20:15:05] Steven_Zhang is an undertaker? [20:15:08] Lead paragraph. [20:15:21] KFP: undertaking, conducting [20:15:23] etc :) [20:15:59] So, this survey will tell me more about the motivations current volunteers have for doing DR, and their ideas. Hopefully we can make changes to attract more volunteers. [20:16:08] Just a question on the survey, was there a question that would show how many volunteers actually were involved in cases previously? [20:16:20] Lets close http://en.wikipedia.org/wiki/Wikipedia:Geopolitical,_ethnic,_and_religious_conflicts_noticeboard [20:16:24] The first survey was on all participants [20:16:35] Ebe123: i have an agenda to get through first :) [20:16:45] Steven_Zhang: They aren't described in much detail. So two changes were implemented: 1) the Volunteer Guide box was added to the right of the page and 2) a bot was added that does something? [20:16:45] I will then [20:16:53] the second survey I targeted only volunteers. [20:17:18] odie5533: the bot does maintenance on the page, keeps an eye on cases and flags them for attention if needed. [20:17:20]