User talk:Arthur Rubin

Test signature Arthur Rubin^en (talk^en 04:26, 29 June 2013 (UTC)

Test signature 2 Arthur Rubin ^en.Wiki (talk^en.Wiki) 11:10, 29 June 2013 (UTC)

Colors

In case you'd like a little more detail: It's theoretically possible to change the color in the visual editor, but you have to either use a template (and presumably enwiki doesn't have any black-on-black text-coloring templates) or trick it into accepting the wikitext code for setting the colors, at which point the fact that you happened to use the visual editor rather than one of the wikitext editors is superfluous. In terms of the effort required to do this in the visual mode, I use the ‎<mark>...‎</mark> highlighting codes fairly often, and the (non-)support level for those tags is the same. While I do know how to force those through the visual mode, it's a lot easier to switch to the wikitext mode. It's the kind of thing you'd probably do only if someone (e.g., some dev) was complaining that the visual editor didn't support something, and you wanted to make a diff to tease them. Whatamidoing (WMF) (talk) 16:02, 29 May 2019 (UTC)

User:Whatamidoing (WMF); I don't think my comment was about VE specifically, I was trying to say that some vandalism is difficult to detect without looking at the Wikitext. On the other hand, it is probably easier to intentionally create subtle damage pages using the Wikitext editor than by VE. (I say, intentionally, because a good fraction (probably 10%) of VE edits to articles I watch have bad links, and I suspect they are unintentional.) Arthur Rubin ^en.Wiki (talk^en.Wiki) 23:17, 29 May 2019 (UTC)

The specific example was black-on-black text: we can't have rich text if we want to avoid vandalism, because we'll end up with silly things like black-on-black text. I think that's testable: we have rich text right now, and we can check for vandalism. On the specific example, we don't seem to have an outbreak of black-on-black text, or even any single example of that happening. On the general question, back in the day, the research showed a slightly lower risk of reversions marked as vandalism for the visual editor than for the older wikitext editors. Now – that was then, and that kind of behavior does change. It could also have been user perception (e.g., I associate some removals in the WTE with vandalism, but these other removals in the visual editor must be a good-faith decision).

Why do you think that rich text is inherently a bigger risk for vandalism than a markup language? It seems like a fairly irrelevant consideration to me. Whatamidoing (WMF) (talk) 16:03, 30 May 2019 (UTC)

Rich text allows for adding "Easter Eggs" (including vandalism) which cannot be seen by the naked eye, but can be seen when highlighting or applying other user-side modifications. In the markup language, you can see them, or at least see there is something suspicious.

I think most experienced editors on en.Wikipedia assume that most of the errors in Visual Editor edits may be undetectable to (or at least undetected by) the editor, and hence may not be vandalism, even though they would clearly be vandalism if intentional. Many of the bad links I have seen in Visual Editor edits are of that form. Arthur Rubin ^en.Wiki (talk^en.Wiki) 19:59, 30 May 2019 (UTC)

Formatting

You can put as much raw wikitext into any Flow post as you want. Just wrap it in <pre> tags. All you have to do (for this and many other things) is start typing the wikitext code, and it'll take care of it for you.

(Parsoid isn't a format. Parsoid is a method of converting wikitext (what's stored in the database) into HTML5 (what's displayed by your web browser when you read a Wikipedia article), and vice versa. Parsoid is replacing the old (unidirectional, HTML4) parser, and incidentally fixing a few bugs and bringing it into line with modern HTML standards along the way. Naturally, that kind of change isn't always appreciated by everyone [1] but overall it doesn't really affect non-technical editors much.) Whatamidoing (WMF) (talk) 21:30, 27 November 2019 (UTC)

Whatamidoing (WMF) I really don't object to Parsoid, as long as changes in Wikicode made by the round-trip process may be reported to the editor. (It's more complicated with templates which create incomplete HTML, something needs to be done before the old parser is eliminated, but I'm not sure exactly what it is.) Sorry if I didn't make that clear in my comments. Should I clarify? Arthur Rubin ^en.Wiki (talk^en.Wiki) 21:49, 27 November 2019 (UTC)

Hi Arthur Rubin, for this particular project, I doubt that it matters. Parsoid only transforms wikitext if it needs to (e.g., not when it's given wikitext for a page that's being stored in wikitext [Flow posts are unique for not being stored in wikitext]), and I don't think people will usually be adding {{hat}} tags in a quick reply, so unbalanced code shoudn't be a typical concern. It's probably not worth clarifying. Whatamidoing (WMF) (talk) 20:39, 2 December 2019 (UTC)

Do me a favor?

Could you test the latest version of the Reply tool on the Beta Cluster? The instructions are at Talk pages project/replying/prototype testing#Reply tool version 2.0. I think it would take just as long to read the instructions as to do it. Click on https://w.wiki/RZf to post your feedback. (That link is a little slow for me, but it loads the list of questions.) Thanks, Whatamidoing (WMF) (talk) 02:06, 1 June 2020 (UTC)