User talk:Philippe (WMF)/VisualEditor change management

From Meta, a Wikimedia project coordination wiki

Moin, this looks carefully thought through, Philippe, thank you for sharing. It might be worth considering to give movement internal institutional stakeholders, like community media interested in understanding and covering the deployment, a specified port to call. Would streamline these additional information demands with formats quite different from the overall distribution requirements, regards --Jan (WMF) (talk) 16:58, 20 May 2013 (UTC)Reply[reply]

Too aggressive in bringing in IPs/new users[edit]

I really like the concept of visual editor, and have been doing most of my editing on Enwp (where possible) using the alpha version. However, I think a week between experienced users and new users is insufficient, even if the critical bugs of working with templates and adding references are addressed. It will take several weeks for experienced users to become comfortable enough with the changes to be able to not only use the new process but to explain complex tasks like adding references. Aside from clear vandalism, the biggest reason for reversion of newbie/anon edits is lack of referencing, and that is not going to be fixed by VisualEditor; that is a separate user education function, and if the "teachers" aren't familiar enough with the topic to explain it, then everyone fails. Risker (talk) 04:33, 18 June 2013 (UTC)Reply[reply]

So if you got to change the calendar to whatever you want, what date would you choose for adding IP editors? Whatamidoing (WMF) (talk) 15:41, 26 June 2013 (UTC)Reply[reply]
Given the usual drop-off in editing that occurs in the English Wikipedia about this time of year, I'd suggest four weeks from when all of the key features are available. Even as I write, I see plenty of experienced editors stumbling around with references and infoboxes and lots of bugs being reported, several of which are being addressed relatively promptly. I'd say probably five weeks from now, or the beginning of August, for any useful A/B testing to take place, and we may well be scaring off more newbies than welcoming them right now. There simply aren't enough hands on deck who are familiar enough with the changes to assist the hundreds of new editors that register every day, let alone the IP editors. Risker (talk • contribs) 15:54, 26 June 2013‎ (UTC) .Reply[reply]
I generally agree with Risker - August/September would be best for the full rollout, in order to give time for many more of the bugs to be fixed (at a non-breakneck pace), and time for the old-users to learn the new systems, and time for the docs to be written/updated - but I've already made those views clear elsewhere so I won't belabor them here. ;) Quiddity (talk) 16:52, 26 June 2013 (UTC)Reply[reply]
Simply put, we've waited 10 years for this. A few extra weeks, while the 'regulars' have a chance to explore and understand the inner workings of this new tool, isn't going to kill anyone. The "workflow" issues that everyone seems to be worried about actually affect power users far, far more than new users; as we probably all realise, new editors consider the conditions under which they started editing to be "normal". And I'm pretty sure that there will be no significant change in the vandalism levels: people who will vandalize will do it regardless of the interface. I can't emphasize how important it is to write the parallel set of help pages and get them translated into key languages, though. I've been editing Wikipedia for 8 years and I still rely on the help pages for certain features. I see you, WhatamIdoing, and your colleagues, as the people who can best lead that effort (no, not do it yourselves, but lead others in doing it). Risker (talk) 17:10, 26 June 2013 (UTC)Reply[reply]
Thanks for the thoughtful comments. Are you more concerned about the amount of time in between registered editors getting to use VE and the IPs getting to use VE, or about the time of year? I was thinking that the relatively lower number of users in July was actually a benefit: fewer editors means less work for the people who patch up the problems created by inexperienced users. Whatamidoing (WMF) (talk) 08:25, 28 June 2013 (UTC)Reply[reply]
Well, it's a bit of a combination, although my focus is on the time between registered editors and IPs. It's pretty obvious that the underlying objective was to have the roll-out complete before Wikimania (which is the intellectual "new year" for Wikimedians), but I think that's putting process over ensuring that the new product is embraced by the editing community. It's important to attract and retain new editors. It is also important to retain experienced editors. And while I believe that VisualEditor can play a key role for both groups, it will only do so if it can provide at minimum the functionality of the existing editing process. It's got to be better to be embraced, and right now I'm very much getting the sense that all kinds of solutions are being tacked on without really ensuring sufficient compatibility and usability, just to say "yes, it can do this too!" I'd rather everyone just slowed down a little bit and created a more realistic timeline. This is intended to be a keeper for the next decade; I'd rather see every solution and bug resolved with a well-considered patch rather than allowing the bad habits of "oh, I'll just stick this bit of code in until someone has time to do something better" patches that are endemic in the MediaWiki code.

Right now, it seems there is good use of the feedback page on English Wikipedia; I'm not sure where else it is being trialed, but I hope you are getting similar responsiveness on the other projects. Unfortunately, what I see is the same dozen or so voices over and over, instead of once or twice from dozens of people. It's becoming increasingly difficult to follow what has and has not been fixed, what will have to wait until later, what is expected to be ready before becoming default. The one thing that would be really useful for experienced users would be a FAQ that compares how one does a task using "edit source" vs how one does the same task using VE. That will make things so much easier when trying to explain things to newbies. Risker (talk) 21:07, 29 June 2013 (UTC)Reply[reply]

I've never heard that Wikimania had anything to do with the timing.
Take a look at this list of recent changes using VisualEditor. Does that really look like the same handful of people over and over? I think it's important not to confuse "most people posting feedback" with "most people using VE". Most people aren't giving us any feedback. That might be because they assume that any problems are their own fault, or because they don't notice them, or because they don't understand how to report them, but the result is that a handful of dedicated communicators appear to dominate the public feedback, even though they don't dominate the user base.
How would you encourage people to embrace VisualEditor? I don't think that people can embrace what they can't use. You seem to be ruling out "turn the button on so they can actually use it without taking special action to enable it", so how would you get it in front of people so that they have the option of embracing it? Whatamidoing (WMF) (talk) 08:04, 30 June 2013 (UTC)Reply[reply]


I'm worried that there is an explicitly stated goal to identify people who are helpful so that there will be a list of people who can be cajoled into helping with other things in the future. VisualEditor is a big deal that is going to change the way editing is done for most users, most of the time. It is something that we're all going to have to muddle through as a community. There's a big difference between that and, say, a deployment that will break a tool used by 50 editors. I'm afraid that if people know they're likely to be tapped for more projects, they'll be less helpful. We see it already in other areas. Risker (talk) 04:40, 18 June 2013 (UTC)Reply[reply]

On the other hand, some people will be very happy about the possibility of forming an ongoing relationship with devs and other staff members. For people at smaller communities, a closer relationship is a means of protecting and promoting your project, not a way of adding to your task list. WhatamIdoing (talk) 12:03, 20 June 2013 (UTC)Reply[reply]
AFAIK, WMF already keeps such lists of people. --Nemo 21:31, 20 June 2013 (UTC)Reply[reply]
Not systematically, for this purpose, no. Philippe (WMF) (talk) 05:43, 13 July 2013 (UTC)Reply[reply]
Well, it has been mentioned in the past (I remember Sue saying something about it). A recent example mentions the CRM: mw:Volunteer coordination and outreach/ECT Feb 2013 quarterly review#Timeline. --Nemo 21:24, 28 July 2013 (UTC)Reply[reply]