Wikipedia talk:VisualEditor/Why

A comment that had no heading

You've gotta be kidding me. You do realize that you have thousands of editors, right? I could refresh the recent changes after a split second and see at least ten new edits. Look at us on the other hand. We have maybe ten active editors, including staff. You aren't going to fix your "problem" with a WYSIWYG editor, trust me. People want WYSIWYG editors? Fine, tell them to use wikEd. I don't want to hear a bunch of newbies coming to my wiki and demanding a rich text editor. If you can't take the time to learn wikicode, you're an idiot. It is by no means complicated, and this babying of editors will only serve to lower the general knowledge of the wiki editing community. Point is that everything you do affects all other wikis, because everyone assumes that everything should be just like you. With this in mind, I'm telling you that this is a terrible idea, and that I greatly dislike not only it, but your flawed logic. Good day.173.78.149.187 (talk) 23:35, 29 June 2013 (UTC)

The word idiot has a technical meaning that doesn't apply to people who are too lazy to learn a markup language. Idiots are people who cannot function independently due to profound mental deficiency, and require lifetime custodial care. People who cannot learn a markup language, or do not want to learn one, are, in contrast, normal. Would encouraging larger numbers of normal people to edit on a wiki improve it? To answer that question one only needs to spend a few hours reading sites that normal people can use, such as Twitter, and the user comments below videos on YouTube. For better or worse, building a wiki is not something everybody is cut out to do. Maybe only a few percent of people are cognitively equipped to contribute productively here - the type of people who read encyclopedias when they were children. All of those people can easily master a markup language, before they move on to the vastly more complex rules for content. --Teratornis (talk) 23:11, 6 November 2014 (UTC)

Updated chart?

The chart on the top of this page only goes to Sept 2009. Is there more current data that could be posted here? Thanks! GoingBatty (talk) 03:23, 4 July 2013 (UTC)

I'm actually working on a script for updating number of active editors, at least (the Italians wanted the data). I can look at at least generating an updated version of that for enwiki; newbie activity, I'm not sure what metric they used (and neither is the uploader, it seems.) Okeyes (WMF) (talk) 11:24, 4 July 2013 (UTC)

Just worry about the number of active editors at this point. The graph on the page is so out of date that it's irrelevant and actually harms the presentation. I'm rather concerned that the supporting information showing editor decline isn't being included in a logical fashion. Eric Zachte, I believe, has been doing editor numbers for years, and I believe this is publicly released information. Maybe just crib his notes? Risker (talk) 04:06, 8 July 2013 (UTC)

Can you explain what you would consider a logical fashion? Okeyes (WMF) (talk) 08:33, 8 July 2013 (UTC)

Well, the feature is being unveiled in 2013. The statistical information that is being presented to justify its existence is 4 years old. It is logical to assume that the data supporting the creation of this extension has continued to be collected and reviewed, and that the data is available at least internally, if not publicly. Thus, it should not be particularly difficult to produce this chart with extension into 2012 at least, and probably well into 2013. Pretty much everyone who is working on this extension throughout its history is familiar with the prickliness of online communities when dealing with fundamental changes to the manner in which they work; providing the best, and most current, evidence of the need for the change gives those who support the change one more tool to promote it. Please keep in mind that there is actually a fair bit of support within the community for this extension, but it makes it really difficult to promote something when the evidence of potential value is not being presented. In other words, current stats will make it easier for others, including me, to defend this course of action. Risker (talk) 16:25, 8 July 2013 (UTC)

Okay. So, as I initially said to GoingBatty, I'm generating this data. Okeyes (WMF) (talk) 07:26, 9 July 2013 (UTC)

Fair enough, Okeyes. :) I know that you're fairly new to the team for communicating on this change, so please note this isn't directed at you. If one of the selling points of a feature is that it will fix A and B, and the feature will affect every user from that point forward and cost a huge chunk of the budget, there really needs to be good, continuing evidence that A and B are problems that need to be fixed. (There should also be evidence that the feature will actually fix A and B, which is completely lacking in this case, as far as I can tell - since the feature wasn't actually usable, even in a test environment, until about 24 hours before it was deployed as the default on this project.) As I say...these are things that need to be addressed, but not by you personally. Risker (talk) 17:53, 9 July 2013 (UTC)

What do you mean by usable? And: I'm as new to this team as anyone else working the change management side :). Okeyes (WMF) (talk) 17:56, 9 July 2013 (UTC)

Until just before deployment, VE could not do references (mandatory for editing) or handle templates in any useful way. And yes, you've only been working on this change for weeks, not years. I noted the reference to 'change management' on another WMF site and found it rather amusing, but that may simply be an inside-baseball joke. Risker (talk) 18:07, 9 July 2013 (UTC)

Crisis? What crisis?

The silly claim on the project page that "Wikipedia is in the middle of a crisis regarding editor numbers" is exasperating. Any project with almost 40,000 volunteers working on it regularly is a runaway success. When so many of the basic subjects have already been covered in articles created to date, a saturation level is naturally arrived at. Wikipedia continues to be a wonderful success. — O'Dea (talk) 06:14, 6 July 2013 (UTC)

Can you point to the data for the 40,000 editors figure? It's not a statistic I've encountered before. Okeyes (WMF) (talk) 23:59, 7 July 2013 (UTC)

@Okeyes (WMF): - The blue line on the graph at the top of Wikipedia:VisualEditor/Why shows about 40,000 active editors for September 2009. GoingBatty (talk) 03:04, 8 July 2013 (UTC)

Ahh. Well, yeah, in 2009. I am working on getting updated data tomorrow :). Okeyes (WMF) (talk) 08:32, 8 July 2013 (UTC)

@Okeyes (WMF): - any news on the updated data? Thanks! GoingBatty (talk) 20:02, 20 July 2013 (UTC)

@Okeyes (WMF): - any news on the updated data? Thanks! Pointillist (talk) 21:54, 6 August 2013 (UTC)

Ack, sorry :(. It's a very slow-running process I have to maintain a connection to make, so I'm going to try and tie it to a cron job instead. Okeyes (WMF) (talk) 17:27, 7 August 2013 (UTC)

@Okeyes (WMF): - any news on the updated data? Thanks! GoingBatty (talk) 03:11, 31 October 2013 (UTC)

Bah, sorry; I swear I'd posted this! Now, where did I put it... Okeyes (WMF) (talk) 16:57, 31 October 2013 (UTC)

the analytics slave keeps timing out on the request. I'll give it one more shot, and then just use the publicly-available data (which isn't as extensive in time, unfortunately). Okeyes (WMF) (talk) 23:16, 3 November 2013 (UTC)

Finally succeeded. Okeyes (WMF) (talk) 21:39, 14 November 2013 (UTC)

Probably unfair but...

Figure 1. Article edits. The geometric mean number of article revisions is plotted with standard error bars by condition (Wikimarkup-only left, VE right).

Figure 2. Productive edits. The geometric mean number of productive edits is plotted with standard error bars by condition (Wikimarkup-only left, VE right).

Figure 3. Hours spent editing. The geometric mean hours spent editing per user is plotted with standard error bars by condition (Wikimarkup-only left, VE right).

Figure 4: Users who make at least one edit Proporition of new users who manage to make a single edit (Wikimarkup-only left, VE right).

These are the results for the VisualEditor we had at time of launch, and, it must be said, it's been improved significantly since then. However, as no new tests are planned so far as I know, we need to extrapolate from this data. Besides the bugs, VisualEditor is very weak on features. Templates cause problems with it, referencing is only slowly being brought up to scratch, images are difficult. Is it really necessary for everyone to edit under this phase, while development continues? Probably not. This data wasn't available at time of launch. However, I would argue it really should have been. One should never presume something is a good idea before the data comes in. Adam Cuerden ^(talk) 19:54, 20 July 2013 (UTC)

The dumbass effect

It's common knowledge that if you make something easy for dumbasses to use then dumbasses will use it, so skewing all your beautiful but useless statistics. The lovely graph in the article is being interpreted the wrong way. The PROPORTION of new editors staying past a year is NOT the issue. The absolute number is the issue, which after a peak is declining slowly, which is not bad considering the attention span of the average teenager these days. DON'T PANIC! In the words of Douglas Adams, there is always a restaurant at the end of the Universe.--Petebutt (talk) 01:45, 27 October 2013 (UTC)

This is a ridiculous comment; not knowing how to use Wikipedia's mark-up language does NOT make you a dumbass and certainly does NOT mean you have nothing to contribute. To assert such a thing would be elitist and this sort of attitude will ensure that Wikipedia will never have a diverse range of editors. One should be tolerant of others.

thedoctar (talk) 15:20, 12 June 2014 (UTC)

Wikipedia's markup language (how to edit) represents a trivial amount of complexity in relation to Wikipedia's rules for content (what to edit). This is obvious by comparing:

• The topics in WP:EIW (the Editor's Index to Wikipedia) relating to markup. Only a handful of topics relate specifically to markup.
• The topics in WP:EIW relating to rules for everything else: style, content, editor behavior, dispute resolution, etc.

Wikipedia frustrates new editors not just because the markup language appears jarring to people conditioned to touch-screen dumbed-down simplicity (optimized for content consumption rather than creation), but because almost everything a new editor might assume about what he or she can contribute is probably somewhere from partially to completely wrong. Wikipedia loses editors because the Wikipedia community functions as an industrial-scale content deletion machine. Vast subsets of human knowledge are unwelcome here, such as many things that are useful, procedural, or insufficiently "notable". I have no complaints about these rules, because Wikipedia cannot be all things to all people and must therefore limit its content. But the new editor does not understand the minefield of assumption-defying complexity he or she is about to wade into after clicking an edit link.

Before a new editor can edit productively here, he or she must master the vast body of editing rules. (By "productively", I mean making edits that stick.) The most efficient way to do this is to read the friendly manuals. A visual editor, among other things, is an attempt to relieve the user of the need to read those manuals which are necessary to master a markup language. A visual editor would likely open Wikipedia editing to more people who have no interest in reading manuals, and who therefore will have a low probability of navigating Wikipedia's far more complex rules for content. Thus a visual editor might be like building a more efficient ramp for encouraging more cattle to wander into a slaughterhouse.

The Cargo cult fallacy occurs when "ill-considered effort and ceremony takes place but goes unrewarded due to a flawed model of causation." A visual editor will stop the editor loss only if the markup language is really the factor driving editors away. A more likely explanation, I suspect, is that editors leave when their hard work gets deleted for violating some obscure rules in the voluminous manuals they were too lazy to read.

To disparage Petebutt's comments as "elitist" makes little sense, because Wikipedia's daunting skill demands are a monument to elitism. To edit productively here requires a user to invest hundreds or even thousands of hours in either reading the friendly manuals, or in painfully rediscovering their content through trial and error experience. Petebutt is not the one telling people they have nothing to contribute - that would instead be the editors who delete their contributions. It is misleading to adhere to politeness in our communication while getting bare-knuckle with people's work.

What Wikipedia needs is not (just) a visual editor, but an intelligent agent that can see and understand everything a user is trying to do, and advise the user just as Wikipedia's most expert human editors would do, in real time, if they were looking over that user's shoulder. However, by the time such expertise can be embedded in software, Wikipedia will no longer need human editors, as the software agents could do all the content editing themselves. That is, I believe, the ultimate solution to all of Wikipedia's editor retention problems - intelligent software that will do all the content aggregation from reliable sources and article editing automatically. Such software might be only a few decades away, thus making human editor retention only a temporary concern. Even sooner we might have software agents that can massively increase the productivity of our skilled editors by doing an increasing amount of grunt work, thus allowing a shrinking pool of editors to maintain whatever level of output we need to refrain from panic. --Teratornis (talk) 22:53, 6 November 2014 (UTC)