After seeing cases like this where an obviously too minor edit is approved (not incorrect, mind you, just too minor) I think we need a "short edit" warning in the suggested edits queue. For example, any edit bellow n characters changed shows a warning, or, a dialog box asking if their sure it isn't too minor when pressing the approve button. This would only be for short edits. What do you think?
What is meta?
- Any Stack Exchange user can ask a question
- The community provides support, votes on ideas, and reports bugs
- Your voice helps shape the way Stack Exchange operates
"Too minor" is a matter of effect. The issue is not that too few characters are changed--tiny edit suggestions are already rejected automatically. An edit can potentially change a huge amount of text but still be too minor in that the change accomplished very little.
(And in this situation, too minor would be the best rejection reason to choose when reviewing, since such an edit often does not radically change what the post is saying.)
Arguably, the system could recognize sentences beginning with actual English words that are not capitalized, and misspelled words, and could warn that the edit is possibly too minor if all it does is to correct one or two of them, while leaving a comparably large number of other similar errors. Similarly, a warning could be shown for edits that just add one tag. I'm not sure it would be worthwhile to display a warning to reviewers in these situations, but that would potentially be productive, and would help reviewers keep in mind some of the ways an edit can not be very good.
In those situations, it's typically better to improve the edit than to click Reject (fixing the rest of the errors, or copy-editing the body of the post as well as editing tags). So if a special message were shown, it should suggest improving the edit to the user.
Ultimately, I suspect such messages would be shown in inappropriate situations often enough that they wouldn't be a good idea. And I know they would not be shown for many edits that don't improve the post in numerous obvious (to a human) ways. Still, it might be something worth testing out.
The specific criterion you've suggested, though--the number of characters added, removed, or changed--that should not be the basis for any automated message to reviewers. This is not sufficiently related to what's really too minor, to be worthwhile. And more seriously, it would probably have the effect of training reviewers into an incorrect mindset about what edits are appropriate and what edits are not. An edit that capitalizes the beginning of six sentences is very likely appropriate and not too minor. An edit that changes thirty instances of one correct spelling into another correct spelling is virtually always inappropriate.
"Too minor" is a subjective call. You say this edit was "obviously too minor", but I disagree. Any attempt to programatically limit the possibility of these would be counter productive. A fix in the title of a question just a few minutes old is something I would almost always approve.
Instead of trying to limit those, the system should find a way to encourage more in depth constructive edits. Perhaps for short edits, the "warning" could be a friendly reminder that they should attempt to find and fix not just one but all problems with the post.
There are times when a single character can be a productive edit. A single character in code can fix a broken answer. A glaring grammar mistake in a question title might be worth fixing even if the post is otherwise perfect. In this case, I would have approved the edit because the title was quite bothersome before. Ideally the user should have stopped to fix additional problems with the post as well, but they should not have been blocked from making that edit and it would have been worth approving. Maybe some of the reviewers should have chosen to "improve" it rather than just pass on, but the end result was a clear improvement, and being a post just a few minutes old, editing it isn't going to spam the front page or otherwise hurt anybody.