Should we improve the code to be reviewed in an edit?

Question

I noticed this suggest edit:

This suggested edit does two things:

• Replace using namespace std;, which is considered bad practice in C++, with std::.

• Fix a divide-by-zero bug by special-casing zero.

This edit was rejected by πάντα ῥεῖ, and then approved by Mathieu Guindon ♦.

I disagree with this edit. We do not generally allow editing someone else's code in a question because editing the post is not constructive and it is much more helpful to leave a comment or answer that points out the shortcomings. Code Review is a question and answer site for seeking peer review of your code. An edit with the summary "updated code" is not a review in my opinion. Therefore, this edit kinda defeats the purpose of code reviewing. Therefore, I rolled back the edit as soon as I saw it.

However, I am not 100% sure my action was appropriate. This edit was approved by a moderator, so there has to be a reason. Was my rollback correct? If not, why not?

Answer 1

This specific case is a bit unusual in that the edit was "suggested" by a user with the same username as the original author. Note also that the original author's account is unregistered, which unfortunately makes it really easy to lose access to the account.

Unregistered users are basically identified through their cookies making it impossible to use that login across different devices. I assume that's what made Mat take a closer look at the users.

With moderator-only information I'm confident that these two users are in fact the same person, so the suggested edit should be treated as an edit by the OP (which implies that the edit is perfectly fine, so long as it doesn't invalidate an answer). At the time of writing no answer has been given, so I've rolled back the rollback.