How far is too far when it comes to editing questions?

Question

I was considering heavily editing this question when Josh then edited and answered the question himself. The edits are significantly better than how it was originally phrased but in the end the question could boil down to something along the lines of how to create sprites that allow for user customization without having to create every single combination ahead of time which is what I was considering editing the question down to. This would remove all the fluff of it being a browser based 2d MMORPG. Or at least I consider these details fluff.

So I provide the above question as an example of the occasional question I'd like to edit and the types of edits I would make, but generally don't make these edits in fear of completely modifying the question to a point where it may upset the asker. Do I take the risk and make the edit, or is there some line that I should be aware of and make an effort not to cross when it comes to editing questions?

Answer 1

In general, I advocate that you edit with assertiveness and conviction. Respect the original asker, but don't be shy. The asker (or anybody else) can always undo your edit.

It's especially important, I feel, to take this tack with questions that are otherwise really bad or close-worthy (the specific question that spurred this discussion was not such, but they exist).

Most of the time, I find that bold edits result in better answers.

For that specific question, being an MMO is potentially important (I briefly mention why toward the end of my answer). I agree the browser-based bit is not, but since it neither adds value nor removes it, I chose to keep that aspect of the question.

One could argue that we should distil all questions to their purest forms, where possible, and this is definitely a question that does apply to just about any 2D game with many character sprite possibilities.

However, I'd argue we should also keep search ability in mind: a less experienced developer will often assume, occasionally incorrectly, that certain aspects of their specific project (browser-based, MMO, fantasy genre, uses the Foo API, et cetera) are relevant to their problem when they are (or do the opposite). They will thus search for those terms, and if we edit all our questions into their pure form they may not find that question in the search results.

If we leave the "extra bits" and details when they do no harm, askers may have a better chance of finding a pre-existing question that solves their problem. If they don't or don't realize that, and ask a similar question (replacing "fantasy genre" with "sci-fi genre" or whatever), well, that's why we have the concept of marking questions as duplicates (rather than deleting said duplicates): they serve as breadcrumbs back to the canonical answer.