[14:49:17] did anyone keep logs from the metrics meeting? [14:51:02] I did but I was 15 minutes late [14:52:11] Ironholds: did you ask jorm to re-do the Main Page from scratch? [14:54:44] jps: no? [14:54:49] tommorris: do you have an opinion on sending people to the reference desks if they submit AFT5 feedback with a question mark in it? [14:55:06] jps: not a strong one. sorry, busy, writing. ;-) [14:55:09] Ironholds: someone should [14:55:26] jps: Brandon's currently busy working on...lesse. five. projects. [14:55:46] he's a bit busy. [14:55:49] well then you should ask Brandon's boss to file more personnel reqs [14:56:02] which will be highly useful come July when the budget is [14:56:14] maybe the exacting standards for designers are too high. [14:56:42] I highly doubt it. We've got four of them and are not having any trouble attracting applicants for a fifth post ;p [14:57:05] being massively over-capacity? We call that a monday. I'm an external contractor who does not, in the board's eyes, exist, and I'm working on five projects. [14:57:23] the people who do exist and are considered permanent are just as over-capacity. [14:58:59] do you suppose Erik would go for a pilot project to rewrite mediawiki in nodejsdb? [15:00:06] would he pay money for a project to completely eliminate the value of our entire platform and features' departments domain knowledge, eliminate compatibility with third-party and secondary extensions and put a big query mark over the value of the work we're doing now? [15:00:11] I'd suggest that would be a "no". [15:00:30] maybe after the visual editor is working [15:00:56] no, I can say with some certainty it won't be something we put money into after that either. [15:01:21] maybe Ward would be more into that [15:01:38] possibly, although I have no idea if he has any money [15:01:40] * Ironholds shrugs [15:02:24] a backwards compatible MediaWiki written in a proper programming language would be nice. ;-) [15:02:27] it wouldn't eliminate anything which already exists. Are you afraid of fud from Encarta? [15:02:49] but such a project would be out of the scope of the Foundation. [15:02:51] no, I'm afraid we'd pay a lot of money for the development of something we have absolutely no use for [15:03:19] like, if you're asking "would we pay for it *to replace MediaWiki*", that's almost certainly "no", for all the reasons given above [15:03:27] if you mean "would we pay for it because it's kind of a neat thing to do", no. [15:03:37] We don't put money and engineering resources behind something just because it's kind of neat. [15:03:37] a replacement MediaWiki would be good. but so would a puppy. [15:03:41] it has to serve a purpose [15:03:49] I suggest you mention it casually to developers and ask whether they would prefer to maintain PHP for decades [15:04:39] you could keep the existing database schema [15:05:12] maintaining PHP, a language they are all (by necessity) experienced at using, versus potentially learning an entirely new language and removing all value from the software we already have [15:05:21] what you are asking is "build a new wiki". [15:05:29] at the same time distracting efforts and/or funds from the more pressing problems we have [15:05:31] which is great. go for it. [15:05:38] but it's not within the remit of the Foundation. [15:05:46] you're free to bring it up on wikitech-l or something, but the WMF would not touch it. [15:05:55] anyway, I don't recommend it until the visual editor is working. Then you will have a bunch of dazed javascript programmers wandering around like zombies wondering why the big blue room is so bright sometimes but so dark others [15:06:13] they will need to be distracted [15:06:21] they could fix Liquid Threads. [15:06:28] lemme find that nice bug I filed a while back [15:06:31] jps: trust me, we'll find something for them to do. [15:06:47] more pop-up windows? [15:07:18] ah, here we go. https://bugzilla.wikimedia.org/show_bug.cgi?id=40501 [15:07:22] our plan calls for interwiki notifications system, a replacement for talkpages, wikiproject formalisation and userpage formalisation [15:07:24] sooner or later they will want to work server-side, and a clean re-factoring off the existing database schema would save everyone else ages in the future [15:07:27] I'm sure we have things we can put them to use on. [15:08:27] I'm not saying scale back existing unified watchlist, comms, or community efforts, but only comms are javascript-intensive [15:09:26] erm. I don't think I was interpreting what you were saying as meaning that [15:09:46] after they get the visual editor and bigbluebutton integration or whatever the talkpages replacement turns out to be [15:09:56] It's called Flow, but sure [15:10:11] okay, so you want to know if, after 2014, we'll be interested in working on this? [15:10:33] VE + Flow + Echo + GlobalProfile + whatever crazy-ass thing jorm thinks up next is a lot of time. [15:11:05] cool [15:11:15] ask us then ;p [15:11:15] I'm okay with eventualism [15:11:23] it's not necessarily even "eventualism" [15:11:42] 2014 comes around I predict we'll have a heck of a lot of things to work on than eliminating 12 years of domain knowledge. [15:11:59] quite frankly once the VE kicks in we cease knowing what our concerns will be. [15:12:07] well, ask some developers and sleep on it, don't make up your mind now [15:12:22] I'm sure that this suggestion will be filed appropriately. [15:12:41] okay. Reedy, how do you feel about re-implementing MediaWiki entirely in nodejsdb? [15:12:49] and presumably porting all the extensions over as well. [15:13:04] I predict the Visual Editor will be a huge success but it will make vandalism appear to increase for at least three weeks [15:13:44] it'll increase the amount of vandalism. but it'll make it possible for people who aren't wikinerds to edit. snakes and ladders. [15:13:44] short-term vandalism is not my concern [15:14:00] well, vandalism is an odd one. It depends how you look at it. [15:14:13] So, people who know they can edit + are vandals = vandalise already. Vandalism does not require wikisyntax knowledge [15:14:32] sorry s/vandalism/test edits/ [15:14:35] my worry on the vandalism front is that it will remind everyone that gee, we can edit, because this'll be live 20 minutes before the BBC and techcrunch pick up on it [15:14:53] my /greater/ concern is that good-faith-but-not-up-to-par edits will dramatically increase [15:15:08] our framework is not built to tolerate a dramatic influx of newbs. We saw what happened last time that happened and it wasn't pretty. [15:15:27] the only solution to that is to allow me to chop the heads anyone who doesn't meet the basic level of competence required to continue living. [15:15:29] pretty soon the New York Daily News will be running headlines "Wikipedia: Peak of Laffer Curve Below 50%" [15:15:30] and that's before we get to purely technical concerns like "if the edit volume triples overnight does the abuse filter die" [15:15:55] and we risk entering a vicious circle where increased vandalism overwhelms our small user group who can actually /deal/ with it. [15:16:03] oops I guess below is a preposition they wouldn't capitalize that [15:16:15] we've got 700 admins and they've got a power law going on. Ditto edit filter managers, vandal fighters, NPPers. [15:16:26] we only need around 200 people to burn out before everything goes to shit [15:16:42] and, of course, for each one who burns out, the odds of others burning out increases [15:16:53] * Ironholds is actually calculating the numbers on this rather thorny problem of at-risk usergroups now [15:17:37] well there are two big knobs that nobody has tried yet due to insufficient boldness: asking for people to try to edit on the Main Page, and asking them to attempt the reference desk (where they are unlikely to damage article space) when they have a question on AFT5 [15:17:48] so, you know that I'm strongly in favor of both [15:17:53] the AFT5 one is unlikely to be a problem any time soon [15:18:31] and asking people to edit on the main page would be swell but...eh. Why bother. The VE will get us a dramatic influx of volume. [15:18:44] Quite frankly the problem is not "we don't have enough people who want to contribute" [15:18:56] if it was, inviting people to edit more prominently would be swell [15:19:04] the problem is the barrier *for* that user group is too high. [15:19:10] well please measure all three. It shouldn't take long to see what you can get out of it. The last thing you want to do is have two go at the same time, because they will convolve [15:19:20] all three? [15:19:43] (1) VE (2) Main Page recruitment (3) AFT5?2RD [15:20:11] well, we'll measure 1. It's kind of hard to measure 2 /without/ breaking things, and measuring 3 is probably a no-go [15:20:19] if two of them happen at the same time, the stats would be way off [15:20:19] we're looking at ways of reducing the volume from AFT5 [15:20:25] well, yes. [15:20:39] you want to be able to compare to control before and after [15:20:45] yes, I did basic research skills too [15:20:48] when the VE goes live, people are going to want to keep it up [15:21:16] the other two can be tested with easy transient measurements, and the existing clicktracking infrastructure [15:21:20] I hope [15:21:22] anyway [15:21:36] I have expressed enough hopes, goals, and aspirations for today [15:21:42] ya [15:22:32] Did anyone keep an IRC log for the metrics meeting? I'll donate my 15 minutes late version to someone who had to leave early