Wikipedia talk:VisualEditor/RFC

Flow

@Cryptic C62:'s comment [here, this RFC is already quite messy, without adding Flow into the mix, and has had very poor attendence so far. Compare with the German RFC, which has reached 400 votes in approx two days: de:Wikipedia:Umfragen/VisualEditor Opt-in. (350 want it opt-in only; no deploy to IPs). You say we should make our unanimous opinion known regarding Flow - unfortunately the Flow part of this RFC is not going to do that. Even if it attracts 100 people demanding wikitext is retained, those hundred will be a small sample size not reflective of the broader community wishes, and those hundred will be disregarded as stale data in two years time when WMF has an actual product called 'Flow'. Including Flow in this RFC reduces attention on parts of the RFC about the Visual Editor. Visual Editor is the issue now; not source editing.

The WMF developers have been scaring the villagers with talk of source editing being discontinued, and Juliusz Gonera (also a dev?) recently put out a blog post which used the unfortunate language "while still sticking with markup editing for now. " (emphasis added) which has been copied by tech rags. I agree we need to be firm and clear that wikitext is required - but not on an RFC about VisualEditor.

fwiw, I believe we need a simpler discussion tool, but it needs to be quite functional (unlike the WP:Article Feedback Tool), and needs to coexist with wikitext. I see this new discussion tool as being useful for newbies, for people who need to comment while they are mobile, and suitable for times when more complex syntax is not required. Wrt newbies, due to the increasing depth and coverage of WP content, increasing we're going to see people commenting first, rather than edit a page. This should be encouraged, and supported with a simple interface for their first contribution. A thoughtful potential contributor is more likely to comment when they see a POV bias in the text, if the article is above ~B class. A reluctant or hesitant contributor wants a low-risk entry point, and comments are the way to do that.

Flow is still vapourware. I am concerned that (some?) WMF personnel hold the belief that wikitext can be decommissioned over time, and are wasting developer resources on this silly objective. If that is the case, there are two problems: the data being used to support development of a wiki text killer, and the design of 'Flow'. We need to question the data, and get involved in the discussions about the design of Flow. If they don't have design docs, we can write them now. ;-) That way we dont get a half baked product dumped on us later on, like has happened with Visual Editor. Then we can complain if the WMF a) doesnt explain/justify the requirements they have, or b) hasnt built Flow per specification! John Vandenberg ^(chat) 06:02, 28 July 2013 (UTC)

Yes flow is vapourware, but the risk is that because they intend to develop flow in the medium term they won't invest in improving the existing systems. It is somewhat easier to tolerate the odd $100,000 to a million being wasted on such topdown initiatives provided we also have a small proportion of the development budget available for useful work such as implementing the various bugzilla requests to reduce the number of edit conflicts.

As for getting involved and making it work, some of us who took part in the beta testing of v/e have very bad experiences of trying to work with WMF developments. At the very least I would hope to have an assurance that if something fails beta testing it won't simply be implemented because the schedule was more important than the project. ϢereSpielChequers 10:37, 28 July 2013 (UTC)

I feel your pain mate. I've sunk a solid week into raising bugs that should never have been made it to a test environment, let alone production. Beta testing and focus groups arnt what VE product needs. It needs experienced architects and designers, and proper user acceptance testing with a very large test plan (or automated test suite regularly run on all browsers), and ... meh, I'd better stop there ... but what is done, is done. I dont see VE being rolled back on en.wp, even if de.wp and others are able to halt the deployment.

Regarding Flow, there is still time; we have far more say if we write the requirements, and be the architects and designers, and collaboratively write the software spec. We've just witnessed what happens when we don't. btw, in general the devs are fixing lots of bottom-up bugs too. p.s. can you give me a link regards to the edit-conflicts problem - I'm guessing its drama boards where they are happening, and Im glad to say I havent had to frequent them as much for over a year, but I have had two edit-conflicts on WP:VEF in the last week that should have been automatically resolved, and I 'feel' that they were previously automatically resolved, but I cant say for certain that its a regression. John Vandenberg ^(chat) 12:27, 28 July 2013 (UTC)

Thanks, can't find any of them in Bugzilla so they've probably been archived or renamed (I'm sure a big part of our fear of Flow et al is that the devs don't use MediaWiki themselves they use Bugzilla. I'd love to see Flow rolled out first as a bugzilla replacement). So I've started new ones. Yes the Drama boards are an obvious area for edit conflict, some easily prevented. But the shooting gallery that is Newpage patrol is far worse, and whilst we can assume that most who've made it to the dramaboards can handle edit conflict, it is part of the process that comes across really badly to new article creators. ϢereSpielChequers 19:15, 28 July 2013 (UTC)

Bugzilla:48719 is the Tracking bug for edit conflicts. (And yeah, it took me a while to find that. :/ )

The mw:Feature_map (section D:4) discusses synchronous editing as a goal, and I believe this is something parsoid is hoping to work on? Ah, volunteer only at the moment, see mw:VisualEditor/Software_design#Architecture and mw:Future/Real-time collaboration and mw:VisualEditor/API/Data Model/Surface. Keywords: "edit conflict" and "real time" seem to get the most results at mw and meta.

Re: bugzilla improvements, I'm sure I read something more specific, but can't find anything further than m:Wikimedia Engineering/2013-14 Goals#cite_note-1 at the moment.

Re: Flow specs, the mw:Template:Flow_Navigation navbox contains most of what is available. The "Use cases" page is one of the better overviews.

Re: wikitext - they have to support it, for editors without javascript, or editors who use screenreaders (eg Graham87). I agree that some of the words used in discussions or recorded meetings is a little too... 'enthusiastic'... about the hoped-for 100% adoption rate - I'll continue to use wikitext by preference, because I want to see piped-link-targets and template-parameters all at once, and because I need WP:Popups in my previews, and etc. Plus serious performance issues in large articles.

Discussing the other Flow questions over at Wikipedia talk:Flow would be good, so that other editors can learn from the answers, and contribute to the questions/feedback, and to prevent fragmented/duplicated discussions. HTH. –Quiddity (talk) 02:11, 29 July 2013 (UTC)

As for why this was not as well participated in as the de-wiki one, well, the WMF staff removed links to it several times, and I suspect the de-wiki probably had a more centralised alert to its existence than we had. Adam Cuerden ^(talk) 22:15, 28 July 2013 (UTC)

Another factor is probably WP:CONEXCEPT giving them an excuse to do that ("oh, we don't have to listen to you anyway, so we'll try to make it as hard as possible for you to complain legitimately to use about it; after all, we won't listen to you even if you do succeed in complaining."). Double sharp (talk) 11:21, 29 July 2013 (UTC)

Not VE's fault?

VE is a good idea since the focus is making Wikipedia easier to edit by all. The discussion behind it has been terrible. It is even hard to find a place to give feedback (maybe I am an idiot for commenting here or it should have been made more obvious that there was a venue for general feedback). It looks like a bunch of people who know basic html are trying to debate both the engineering and usability of a project that teams at some of the most prominent tech companies would have a hard time undertaking. The RfC is a joke. The Foundation should be ashamed of the product and the community should be ashamed of their inability to do anything about it. Any potential new editors will move along since it is obvious that Wikipedia has turned into a joke. I hope that Wikimedia gets acquired by a company that actually has the resources to support such an important endeavor while still keeping ads out of articles. 67.40.177.104 (talk) 06:45, 30 July 2013 (UTC)

When was the last time a company acquired a 501(c)(3) charity? —Jeremy v^_^v ^Bori! 19:27, 30 July 2013 (UTC)

Barely anyone disputes that it would be better to have a visual editor, the problem is the disappointment that it was rolled out before enough of the bugs were fixed. Having bugs in new software is perfectly normal, unfortunately this got implemented a little early before enough of the bugs were fixed. ϢereSpielChequers 18:25, 31 July 2013 (UTC)

Closure

It has been over a month since anyone commented on this RFC. It would be good to have a RFC close summarise the issues involved, even if they are no longer an issue due to subsequent config changes (like VE being turned off for anons).

I've quickly reviewed it. Here is my first take for consideration:

“ There is overwhelming support for a visual editor, however the current product is unsuitable for editing a large proportion of English Wikipedia pages, and the current product is too bugging to be deployed for widespread usage on English Wikipedia. It should be easy to disable for users who do not want to use it, and should not be the default when there are significant bugs which can cause page corruption.

The majority of the community feels the deployment was too early and too aggressive, but a large minority can appreciate the WMF tried to inform the community of the deployment, and the WMF felt wider deployment was beneficial to their development processes.

Wikitext editing is a core requirement of Wikipedia. While not the focus of this RFC, there was a very strong consensus of a smaller subset of participants that a future product 'Flow' must support wikitext editing.

”

John Vandenberg ^(chat) 01:02, 21 October 2013 (UTC)

I have closed it. Feel free to overturn if you think I shouldn't have done it or if my closure didn't match the RfC. Fram (talk) 09:50, 24 October 2013 (UTC)