I am not one to whinge and I'll admit it when I'm wrong; I apologize if I come across that way, But I don't understand the rejection of my suggested edits here:


My edit was to change two occurrences of 'RGB', in the title and prominently in the question, to 'ARGB'. This was because everything in the question, including all of its citations, actually concerned themselves with ARGB and not RGB.

Moreover, this question was specifically talking about how (A)RGB values might be packed or unpacked; in this context, an RGB value will behave very differently than an ARGB value... if readers don't catch this detail they will have been given incorrect information.

I wouldn't have complained if the edit hadn't been branded a 'comment' by 3/5ths of the reviewing staff who weighed in -- if the rejection had been more heterogenous or explanatory, or had the rejection votes come in over an interval of time more suggestive of reviewer circumspection, I'd leave it be.

But so. May I ask, what is the deal with that?

  • 3
    One problem I see is that you left the title in an inconsistent state - ARGB tuple (R,G,B). Would have been better for the reviewers to improve it I guess. But hey, they're people (and not staff, reviewers are regular users).
    – Mat
    Dec 9, 2012 at 16:21
  • I think the rejects were because of "This edit is incorrect", not the alternative "or an attempt to reply to or comment on the existing post". I'm not sure whether it should have been rejected (not enough domain knowledge). Dec 9, 2012 at 16:22
  • Beyond the missed (r,g,b) in the title, I can only assume this falls under the header "sh*t happens".
    – Bart
    Dec 9, 2012 at 16:24
  • @Mat that is a fine point, and I would have been happy to correct my mistake there, given the opportunity; I would have thought the "commit note" afforded by the field in which one explains their edits would have helped in this case.
    – fish2000
    Dec 9, 2012 at 16:25
  • @fish2000: there's no way for reviewers to "send the edit back to editor". The only options are accept, reject or improve. Improve doesn't send any sort of notification to the editor - it's as-if the edit had been approved (unless the editor actively looks at his pending/just passed edits).
    – Mat
    Dec 9, 2012 at 16:30
  • @Mat aha good to know, thanks. Now that StackExchange has this newly polished and inclusive "review" process in place, is it worth discussing new possibilites around the edit approval process? (Or, maybe that is overreaching on my part to suggest something like that...)
    – fish2000
    Dec 9, 2012 at 16:33
  • ... although it appears the other side of the coin is well-represented, so to speak: meta.stackexchange.com/questions/148167/… ... there is already quite a lively discussion there.
    – fish2000
    Dec 9, 2012 at 16:35
  • 1
    If you have a constructive idea to improve the review process (for suggested edits or more generally), feel free to suggest that (you could edit your question for that since no-one's bothered with a proper answer yet :) ). Anyone can make feature requests here, the community reacts, and sometimes stuff happens. Sometimes not. No harm in trying.
    – Mat
    Dec 9, 2012 at 16:39
  • @Mat thanks for suggesting that -- I will do so despite the presence of a bonafide answer at the time of writing, indeed.
    – fish2000
    Dec 9, 2012 at 16:49
  • 1
    @fish2000: you should not change your question once it's been answered, that's plain impolite. Open a new question, it's not like they cost anything, and focus on your feature request.
    – Mat
    Dec 9, 2012 at 16:51
  • @Mat rgr that, coming up...
    – fish2000
    Dec 9, 2012 at 16:52
  • Reverted, apologies for the breach of etiquette.
    – fish2000
    Dec 9, 2012 at 16:54

2 Answers 2


Your edit was a good one: it made the title consistent with the body and with the answers, and added some relevant tags. I would have approved it.

The “invalid edit” rejection reason lumps two different kinds of edits: factually incorrect edits, that introduce mistakes, and invalid use of the Stack Exchange engine, such as posting follow-up questions or editorial remarks inside posts. Splitting those two (I think the right thing would be to lump factually incorrect with radical change) is a long-standing request (back from when rejection reasons were introduced), but since it hasn't happened after over a year, I doubt it ever will.

  • @Giles hey thanks for saying so.
    – fish2000
    Dec 9, 2012 at 20:03

Edits are not supposed to correct mistakes (errors of thought) in questions or answers. Spelling, grammar, and formatting are all fair game. Rewriting titles to better match the question content is also fine.

When I see an edit that is inspired by "this answer is wrong" I reject it.

In your case, if the person typed ARGB everywhere and slipped once and typed RGB, correcting that would be fine. But what has actually happened is they've asked one question in the title and another in the body. I don't see how you can tell which is right. This is a case for a comment asking the user to explain, followed by a pause for the user to do the edit themselves, and possibly an edit incorporating their answer if they answer in a comment instead of by editing their question. (And yes, that particular poster probably has moved on from that post: that's a good reason not to get involved, not a reason why you get to use different rules.)

It's not relevant whether you are right in your opinion that the original post had an error of thought. Suggested edits are not for correcting those kind of mistakes.

  • 1
    The further content of the question as well as the comments by the OP suggest that ARGB is indeed what was intended. But you might indeed be right that this caused the rejections. (Though that would be a surprisingly strict view of the reviewers...are things improving? ;) )
    – Bart
    Dec 9, 2012 at 16:44
  • I do believe that most if not all of the references in the question and the answer are to ARGB -- all of the code provided in the answers I saw was ARGB. If you try to pack or unpack RGB values as ARGB, or vice versa, it won't come out right. That's why I thought the two low-key yet prominently-placed RGB references were "wrong" -- it looked like someone had made an honest mistake of the sort that you describe in your first graf as fair game, is all.
    – fish2000
    Dec 9, 2012 at 16:46
  • 3
    Probably in this case the edit summary is key. Reviewers can't see the answers or comments. If you simply asserted that you can tell from OPs comments they mean ARGB, it would probably have been accepted. Dec 9, 2012 at 16:52
  • I'll keep that in mind, going forward. Thanks.
    – fish2000
    Dec 9, 2012 at 17:14
  • 2
    Nonsense. An edit that corrects minor factual mistakes is a good thing. An edit that makes the post title consistent with the body and with all the answers (including an accepted one, so you know it's what the asker meant) is a good thing. Dec 9, 2012 at 18:34
  • @Gilles you are welcome to your opinion but users without the edit privilege are told not to suggest edits that fix factual errors and reviewers are told to reject edits that should be comments Dec 9, 2012 at 19:25
  • 1
    @KateGregory That isn't just my opinion. “Correct minor mistakes” is in the “How to Edit” instructions that are shown to every editor (whether the edit requires review or not, that doesn't make a difference). It's also common sense: it's better not to have those mistakes in the post. Why on earth would you want to reject a change that makes the post better? Dec 9, 2012 at 19:43
  • 1
    The fact the post is old is not relevant, it comes up in a web search (presumably this is why the OP felt impelled to edit it!) and so ought to be clear and consistent.
    – hayd
    Dec 9, 2012 at 22:21

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .