[00:52:00] OK. I'm out of here. Have a good one! [00:52:01] o/ [16:00:47] mornin' halfak. :-) [16:01:00] o/ leila [16:01:02] we're in the meeting, in case you're in some other room [16:01:02] :D [16:01:20] Your ping beat my calendar ping [16:01:38] :D [17:45:17] o/ guillom [17:45:19] I was hoping to riff on one of your edits to the VisualEditor page on mediawiki.org in my presentation today. [17:45:28] Could you take a quick look at my slides to tell me if I did OK? [17:45:30] https://docs.google.com/presentation/d/1ZGz9HnTbDKUHX378rYgAwImuAtgQyG2YxCuwofdYGEk/edit#slide=id.p [17:45:33] hey halfak [17:45:39] * guillom opens. [17:45:59] See slides 4,5 and 89 [17:46:31] On 89, I talk about how, once we solve other barriers for newcomers, we'll reap the benefits of VE. [17:47:51] halfak: Are you sure I'm the one who originally wrote that? [17:48:02] There's an edit where you add it. :) [17:48:05] I don't remember; but it doesn't mean I didn't do it. [17:48:09] I'll go look it up again quickly. [17:48:22] I may have stolen it from elsewhere :) [17:48:42] https://www.mediawiki.org/w/index.php?title=VisualEditor&diff=416093&oldid=416080 [17:48:57] looking [17:50:10] halfak: oh damn your presentation is today?! [17:50:14] * YuviPanda ws hopign to not go to the office today [17:50:18] looks like that's not happening then [17:50:21] Yup. [17:50:21] * YuviPanda preps to go to office [17:50:25] But it'll be recorded! [17:50:27] :) [17:50:37] I can watch live I guess [17:50:41] not same as being there ofc [17:51:33] I also fell into 'customize my linux laptop' mode. currently at the i3 tiling window manager... [17:52:18] Oh man. I haven't done that in a while. [17:52:32] Actually fixing things in my dev environment. [17:52:33] ha. [17:53:08] I suppose I did install a new OS to get CPU scaling working a few months ago. [17:53:15] That was a HUGE pain, but well worth it. [17:54:03] heh [18:03:06] o/ DarTar [18:03:18] Is there a showcase setup call I should be joining? [18:03:58] halfak: check with Leila, in VPE Staff meeting [18:04:05] kk [18:04:07] o/ lzia [18:04:08] ^ [18:08:20] halfak: are you around for some AV testing? [18:08:22] :-) [18:08:30] Yes [18:08:32] (IT people will be back in few min) [18:08:35] Was just asking about that [18:08:37] :P [18:08:37] great! thanks! I'll ping [18:08:40] :P [18:08:42] kk [18:15:58] Aaron, can you join here: https://plus.google.com/events/c001rf6survcafoe767c40bbj8k [18:16:11] noo! [18:16:12] wait [18:16:13] :D [18:16:15] k [18:16:19] * halfak waits [18:16:23] halfak: Found the source: https://strategy.wikimedia.org/wiki/Product_Whitepaper [18:16:31] It was written by Howie there :) [18:16:45] guillom, lol. Too late to fix slides. [18:17:04] I'll just have to say that it was Howie, but I thought it was you! [18:17:06] No worries; at least we know :) [18:17:44] halfak, you should have invite [18:17:46] halfak: Note that the original text was: Removing the avoidable technical impediments associated with Wikimedia’s editing interface is a necessary, /but not sufficient/ precondition for increasing the number of Wikimedia contributors. [18:17:58] Interesting [18:18:00] :) [18:27:37] * halfak stuffs face with cookies because he forgot lunch [18:27:46] guillom, change has been made to slides [18:27:53] :) [18:28:05] Also, I can see you :D [18:28:07] halfak: oh, good! Thanks. [18:28:18] hmmm, dye hair before or after research showcase? [18:28:22] hmm, I know, *during*! [18:28:27] Lol. [18:28:30] Starts in 5 min [18:28:32] 3 min [18:28:56] youtube link? [18:28:59] halfak: I was just joking with James that I wanted to play with an IR spot near my head, so that the only thing the camera would see would be a big white spot :) [18:29:10] Nerdy anonymity. [18:29:19] the trick is to get bright red hair. then you perfectly blend in [18:29:33] YuviPanda, it's in an email. [18:29:38] I don't want to move my setup. [18:29:49] https://www.youtube.com/watch?v=vGyrVg_qKSM [18:29:51] YuviPanda: ^ [18:30:07] * YuviPanda stands by [18:31:04] halfak, you here or in sf? [18:31:11] in MN [18:31:17] k [18:33:20] thanks guillom for sharing the link. :-) [18:33:31] Everyone, welcome to the showcase. If you have question, please post it here. [18:36:17] welcome Ziko. :-) [18:36:20] hello [18:37:13] YouTube link just in case: https://www.youtube.com/watch?v=vGyrVg_qKSM [18:37:21] nice funnel - i am now thinking how to relate it to the social and content components of the wiki [18:37:50] hi fuzheado [18:37:55] Hi Ziko! [18:38:30] wow, all sorts of non-regulars [18:38:31] hello fuzheado. [18:38:58] CSD = Christopher Street Day. AFD = Alternative für Deutschland. [18:39:10] :D [18:39:27] Actually, when I edit English Wikipedia, I feel like a newbie, because I often don't know the abbreviations different from German. [18:39:46] hi leila ! [18:39:59] Gah my Youtube screen is frozen so I’m not seeing the acronym city [18:40:03] hey fuzheado; great work on the guerilla videos from Wikimania! [18:40:16] guillom: thanks, hope they were usable! [18:40:31] +1 (I made a video myself, can I add it?) [18:40:40] Ziko - absolutely! [18:41:02] Its not sharp but I consider uploading it to Commons [18:41:10] fuzheado: in case it helps, the slides are here: https://www.mediawiki.org/wiki/File:Visual_editors_effect_%28Showcase,_July%2715%29.pdf [18:41:31] leila - nice, thanks for that. I’m hearing the audio fine, but the slides aren’t changing for me [18:41:55] ah! disconnect and reconnect fuzheado? it sometimes happen to me on Firefox. :-\ [18:42:51] nowadays we still have WYSIWTF [18:43:04] heh [18:44:39] I was impressed how well VE worked on table editing for the Videos page on Wikimania, so it’s been a lifesaver for me [18:46:00] oh - and the speed? my experience was that the pages build up very slowly [18:46:32] It was zippy for me [18:46:39] It's improved with time [18:47:21] possibly, the VE will not have much effect on vandalism. [18:47:52] if they want to wreck something, they already can change something in a infobox wikisyntax bubble [18:48:23] VE is suprisingly fast on a slow (3G/2G) connection ;) [18:48:56] Even as a veteran editor, I abandoned a lot of editing tasks because of complexity (ie. tables) and VE has brought me back into editing again [18:50:18] Q: maybe one week is not enough time to observe a difference? [18:50:40] Giovanni_C: will ask. can you expand more? [18:50:54] you mean in terms of "productivity requires more time to be observed"? [18:51:06] more reverts - what exactly for? [18:51:07] Yes! [18:51:14] got it, Giovanni_C. [18:51:45] Ziko: is that a question for Aaron? I've missed it I think. [18:51:54] leila: yes [18:52:13] direct link to 2015 results btw https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_registered_editors/May_2015_study [18:52:17] what kind of edits of what kind of editors - regulars, casuals? [18:54:26] Oh but what tough nut to crack — “Social/Motivational!" [18:54:28] Ziko: Can you expand? Not sure I understand your question. [18:54:29] final statements sounds very sound [18:54:30] clap [18:54:50] *clap clap [18:55:02] if there were more reverts (a little bit more) - what kind of reverts? because of simple vandalism? [18:55:18] Ziko: I think there were a little fewer reverts. [18:55:33] (only slightly) [18:55:46] oh sorry, then i got it wrong [18:56:07] I can still ask if the type of reverts were different, Ziko. [18:56:17] hey leila - what wikis did this run on? [18:56:26] enwiki atgo [18:56:34] hi atgo .:-) [18:56:35] thx [18:56:37] hi! [18:56:38] so the conclusion / assumption is that the newbies in this sample encountered some these social/motivational barriers and that they were the reason that they did not contribute more. how coul this assumption be tested empirically in this setup? [18:56:53] maybe less revert because less screwing up with the code? [18:57:25] To isolate the “Social/Motivational” factor, it’d be interesting to do the A/B testing on GLAM engagements, and see how different groups using VE and not-VE behave [18:57:29] Ziko: Another hypothesis would be: fewer reverts because less unsourced content (VE makes it easier to add citations). But that's just a hypothesis. [18:57:30] The assumption that "anyone can edit" doesn't seem justified, from what I've heard. Dario mentioned that at one point, 25% of new account registrations immediately attempted to edit the Justin Bieber article... which is protected from anyone who is not autoconfirmed, aka is not ten years old. [18:57:40] I think that most research or statements begin with the idea: editing should be made easier for the newbies. but it is good to look only at the effects for the regular ditors, how to help them. [18:57:52] guillom: sounds reasonable [18:58:06] guillom: interesting… possibly [18:58:27] only = also [18:58:27] i cannot hear the questions at all. [18:58:42] Yeah hearing Dario very dim [18:58:42] I think it's fixed subbu [18:58:50] Dario = Better [18:58:51] yes, i hear dario [18:59:00] great, thanks for confirming Ziko. [18:59:19] yes, hear him now, better than before. [18:59:30] :-) [18:59:52] Could it be the satisfaction from using a VisualEditor (+usability) influences positively middle term and long term contribution? If so, what period of time should we measure to be sure? [19:00:10] I made up a kind of hypothesis for the motivation to edit: the potential editor thinks positively of the wiki but thinks that at a specific point he can improve it. [19:01:21] Marmick: I agree that there can be a long term effect. Someone learns to edit wiki code, then does not edit for weeks, is frustrated that when he comes back he has to learn again... if VE, then he does not have to learn again so much. [19:02:22] Ziko: I think so! Positive responses and ease of use from a system ensure a future use of it. [19:02:23] o/ [19:02:48] The test was done only with newly *registered* editors? [19:02:58] Hey golks. Thanks for watching. [19:03:01] Ziko, yes [19:03:04] thanks halfak [19:03:07] halfak: awesome [19:03:22] Ziko, I'd really like to run a test with anons, but it's been overruled for complexity reasons. [19:03:32] Also, it's hard to tell which anons are new. [19:03:50] but the contrast was also only with registered newbies? otherwise, that might have an impact and make a difference [19:04:01] Regretfully, people seem to not like the idea of me implanting a tracking device in their hand so I can track them cross-device ;) [19:05:06] BUT you're right. This is something we don't know. I think there are a lot of reasons to believe that the pool of anons is generally more experienced that newly registered users. [19:05:30] Even though anon->registration is a strongly assumed path, I don't think it's that common. [19:05:36] I suppose that regular users are much more sympathetic to registered newbies [19:06:35] I donT know the design but that may explain the difference with regard to reverts [19:07:11] Ziko, anons get reverted less and (where they can create articles) deleted less. [19:07:29] marmick, I think that we should look at 1-2 months for retention measures to start. I have some work that suggests this measurement is relatively stable. [19:07:41] https://meta.wikimedia.org/wiki/Research:Surviving_new_editor [19:08:52] is this the IRC channel for the Research & Data showcase? Calendar item doesn't say [19:08:58] spagewmf: yes [19:09:00] What Thomas et al are presenting is being documented here: https://meta.wikimedia.org/wiki/Research:Wikipedia_Knowledge_Graph_with_DeepDive [19:09:08] halfak: thanks for the link. it makes sense. [19:09:24] spagewmf, sorry about that [19:09:26] what does "Recall 70%" mean? [19:09:32] Will be more diligent next time [19:09:40] we should include that in the calender event, spagewmf. thanks for bringing it up. [19:09:48] Recall @ 70% = 70% of the Trues are marked True [19:10:00] but in general, regulars are more hostile to IPs? [19:10:18] Ziko, that's a good question. I'm not sure we know that, but essays suggest some hostility. [19:10:20] hey halfak: any idea where I would find the event logging table(s) for Notifications? IIRC, you said one was set up when notifications were rolled out years ago. [19:10:26] spagewmf: https://en.wikipedia.org/wiki/Precision_and_recall [19:10:36] Ziko, And strong assumptions about IPs that are generally wrong. [19:10:41] J-Mo: do you mean Echo? [19:10:42] spagewmf: https://en.wikipedia.org/wiki/Precision_and_recall recall shows the % of relevant items that are selected. [19:11:05] spagewmf Yes! tho I've been told that the old codename is depricated :) [19:11:40] Why were IP users less reverted? Because of the kind of their edits? Because they did simple tippo edits? While the registered do more complex edits? [19:12:31] J-Mo: http://ee-dashboard.wmflabs.org/dashboards/echo# used to graph some of the Echo events :-( [19:13:12] spagewmf, halfak I think I found what I'm looking for on stat1003! [19:13:40] Cool. J-Mo I'd have had to do digging anyway. [19:15:04] J-Mo: from the EchoConfig settings in wmf-config, looks like we log to Schema:Echo, Schema:EchoMail, and Schema:EchoInteraction [19:16:14] yep, that jibes with what I'm seeing spagewmf. Unfortunately I was especially interested in a table called EchoPrefUpdate, and it looks like that schema is no longer in place. [19:17:52] Ziko: https://meta.wikimedia.org/wiki/Research:Wikipedia_article_creation already proved that usually newbies are more evil than unregistered users [19:17:57] (brutally simplified) [19:18:09] taking question for Thomas et al. :-) [19:18:29] thanks Nemo_bis [19:18:31] Thanks for linking Nemo_bis [19:18:33] Ziko: a possible explanation is that if you want to do something unfair you'll register first, knowing it gives you a cheap advantage over unregistered users [19:18:40] J-Mo: the commit comment was " Echo preference change is already captured in the general [19:18:43] preference change, this is just a duplicate [19:18:53] when it was removed [19:19:35] ah… excellent to know. I'll look in user_properties... [19:20:15] fundamental question: why to people register. and when. [19:20:44] leila: Question: Did they actually add those properties to Wikidata, or was this a proof of concept? And if they didn't, can the Wikidata contributors import them (e.g. semi-automatically?)? [19:20:51] Ziko: yes, and that changes drastically across different language editions [19:21:27] So that answers my question :) [19:22:05] Ziko: you don't need to know the reasons for everyone; it's enough to know that a sizable minority of new users are from people wishing to "alter" consensus, or post something they would be ashamed of with their main account, or try to make their propaganda survive quality checks [19:22:13] question: did you look at dbpedia's extraction methods for comparison? [19:22:19] see e.g. http://dbpedia.org/page/Aerated_Bread_Company [19:22:48] (https://en.wikipedia.org/wiki/DBpedia ) [19:23:21] J-Mo, I think you might want https://meta.wikimedia.org/wiki/Schema:PrefUpdate [19:23:58] yes! that helps. thanks halfak. [19:24:07] No prob bob :) [19:24:32] (... https://en.wikipedia.org/wiki/Template:Infobox_company actually has a "founder" field) [19:27:01] I think Founder is https://www.wikidata.org/wiki/Property:P112, how do I then find statements of that property? [19:28:13] ... I guess http://wdqs-beta.wmflabs.org/ and the WTF of sparql [19:31:56] thanks guys [19:53:44] * halfak might actually have a 1:1 with DarTar today [19:53:52] This always seems to get delayed. [19:56:36] halfak: DarTar was showing the researchers around after the showcase [19:57:57] No worries. the meeting isn't scheduled for hours. I might have jinx'd it by talking about it now though. [19:58:03] Thanks for letting me know spagewmf :) [19:58:56] halfak: I'm sorry I missed most of your talk, I would have had a dozen questions. [20:03:08] spagewmf, can still ask 'em [20:03:12] Either here or on the talk page. [20:03:19] https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_registered_editors [20:03:33] Or more specific to the recent study: https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_registered_editors/May_2015_study [20:05:14] halfak, "We found that newcomers with VE enabled by default were reverted significantly less and blocked slightly less often than newcomers with VE enabled." <-- needs editing. [20:05:30] visual editing or just plain editing? [21:00:35] hey halfak, coming in 2 [21:00:42] (sending out one note to the list) [21:01:30] kk [22:03:04] DarTar: are you in 32? [22:04:14] lzia: I am [22:04:20] heading there DarTar [22:04:23] kk