I was doing some reviewing earlier when I came across this question, which in my view doesn't have sufficient code to constitute a reviewable question. It essentially consists of a data structure short[3][3] and a description of how values can be stored/updated in it. Edward had already left a comment, so I voted to close and was going to move on when I noticed that the question had just been answered.

I find it hard to reconcile how you can have a code review without having any code to review (it also seems to be bad form to answer questions likely to be closed), so I downvoted, commented and flagged the question as 'Not an Answer'.

The question has indeed now been closed (as broken/code not written), however the answer flag was declined as 'a moderator reviewed your flag, but found no evidence to support it'.

On the face of it, the answer reads like it is answering a question, has attracted an upvote and it was useful to the asker (see comment on question). However, that's because it was providing an alternate solution to a design question, not because it was reviewing code.

  • Is the decline just one of those things / bad timing (I'm pretty sure it was handled at the same time as a bunch of spam flags), so I should just keep doing what I'm doing?
  • Is 'not an answer' the wrong flag to have used here, because it doesn't provide sufficient context to the mod handling the flag. Perhaps a custom flag as suggested here, with a description of why it's not an answer would have been more appropriate?
  • Am I completely off base here and I should have just left the answer alone?
  • 3
    \$\begingroup\$ To quote the Mug, "Invalid answers on off-topic questions in the process of collecting close votes should be downvoted without blinking, and hopefully removed too. If the question is edited into shape, IMO invalid answers are NAA." \$\endgroup\$
    – Peilonrayz Mod
    Commented Dec 9, 2016 at 22:41
  • \$\begingroup\$ I would argue that a data structure is code. (Not saying this question was valid, though.) You can't really have a useful program without them. \$\endgroup\$ Commented Dec 10, 2016 at 7:47
  • \$\begingroup\$ @user1118321 a data structure could be code if complete, however putting something in code quotes doesn't achieve that. On its own, short[3][3] is no more code than variableName. \$\endgroup\$
    – forsvarir
    Commented Dec 10, 2016 at 9:12

3 Answers 3


The question at hand should be (and is) closed. Off-topic questions should not be answered. It's as simple as that.

The answer should never have been written. It has been written by a new, inexperienced user. The answer will hopefully be removed and the question stays closed.

Alternate solutions are only acceptable if they support a review. In this case, there was no code. If there's no code, there can't be a review. Therefor, this question can't and shouldn't be answered.


I was the one declining the flag. My definition of "not an answer" might not match what others have. My definition is that if it is an attempt at answering the question, it is an answer.

I find it unreasonable to treat an off-topic question with "Hey, this question is off-topic, let's remove all the answers on it.

On the face of it, the answer reads like it is answering a question, has attracted an upvote and it was useful to the asker (see comment on question), however that's because it was providing an alternate solution to a design question not because it was reviewing code.

As you say, it was useful to the asker and to me that is more worth than always deleting answers because "it is not a review and cannot be a review because the question doesn't contain any code".

I honestly see very few gains with deleting answers to off-topic questions just for the sake of it. I found this answer to be a good one, so as a moderator I did not want to outright delete it. Of course the community is free to upvote and downvote the answer as it feels.

  • \$\begingroup\$ I'd say that if a question is off topic, it should be closed and deleted if it is unsaveable. Removing answers on unrecoverable questions serves no purpose, so just don't bother with them. If the question is recoverable, removing the answer depends on whether it is reasonable and/or saveable by the author. \$\endgroup\$
    – user34073
    Commented Dec 10, 2016 at 1:30
  • 1
    \$\begingroup\$ I think my main issue with this is that for me, the fundamental requirement for a code review is the presence of code. So where there is no code, there is no possibility of an on-topic question. With no question, there is nothing to answer so anything posted as an answer can't be an attempt to answer the question (because there is no question). That doesn't mean the post can't be useful / insightful, but it is a comment and so should have a similar lifespan (except that my understanding is it could be undeleted if the question is fixed in the future). \$\endgroup\$
    – forsvarir
    Commented Dec 10, 2016 at 2:36
  • 1
    \$\begingroup\$ @forsvarir It might not be a code review, but it is still an answer. The problem is the question, not the answer. Just because it's not an on-topic question doesn't make it a non-question. The question is still a question, but an off-topic one, and the answer is still an answer. I don't see a reason to force-classify this answer to be a comment. It still says what is says and it helped the OP. \$\endgroup\$ Commented Dec 10, 2016 at 10:42

As I said it before, you should refrain from answering a question that's likely to get closed.

But what if somebody did answer? What to do with that?

The "not an answer" flag is not a good fit. If the question is "how to make carrot soup", and somebody answers with the recipe, it's simply not true that "it's not an answer". Probably it is an answer, it might even be a great answer, and really help the OP. Even though we don't want such content, saying it's "not an answer" is inappropriate.

Since the real problem is the question, not the answer, I suggest to target the question. Keep in mind that there are standard moderation tools to deal with bad questions, so if you have the privilege to vote to close, downvote, vote to delete, then use that, rather than flagging.

  • \$\begingroup\$ I'm not really aiming to generalise this across all soup related questions, but doesn't the context of the site impact what constitutes an answer? In a code review meeting for a cooking site somebody might mention a great recipe they came across, but I wouldn't consider that part of the review, it is part of the commentary. There is no 'off topic' option for answers, in my mind on a code review site, if it isn't attempting to review code it fails to meet the bar for being an answer. \$\endgroup\$
    – forsvarir
    Commented Dec 10, 2016 at 9:27
  • \$\begingroup\$ What if the question was where can I download free music and people started posting their favourite sharing sites. They are all answers to the question, not spam since they are answering the question, should they be ignored? Sure, the question is the cause, but each subsequent answer exacerbates the effect and by not addressing the answers as well aren't we essentially saying it's OK to answer with off topic content? \$\endgroup\$
    – forsvarir
    Commented Dec 10, 2016 at 9:32
  • 1
    \$\begingroup\$ This seems like a perfect example of a question that should be migrated. \$\endgroup\$
    – pppery
    Commented Dec 12, 2016 at 1:33

You must log in to answer this question.

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