32

So, apparently, there is a feature that edits on posts that are on hold only push the post into the reopen queue if the author of the edit has not flagged the post. As in the linked question on ELL.meta, it is known that this can lead to strange behaviour when the author of the edit also "flags" the post for reopening in the queue, invalidating the review.

I think I have discovered a new variation of this bug that is more concerning as it is much easier to trigger: This review on physics.SE has been invalidated. There are no pending reopen votes on the post, so the situation described earlier has not happened. The edit pushing the post into the reopen queue is by the owner of the question themselves. From looking at the (moderator) timeline, I surmise that the following happened:

  • The owner edits the post, initiating reopen review.
  • The owner flags a comment by another user (now deleted)
  • The system considers the comment flag a "flag against the post", meaning the edit does no longer qualify the question for review, invalidating the review

There are no other actions in the timeline between the edit and the review being invalidated, which is how I am reasonably sure that this interpretation is correct.

Given that, in the best case, a post that is put on hold has comments on it asking for clarification and a good edit should make those comments no longer needed, it seems in fact likely that a conscientious editor might also flag comments on the edited post for removal. If this in fact invalidates the reopen review initiated by this edit, this seems like a bug that should be fixed, especially since there is no indication at all to the users involved that this happened - the reopen review simply silently fails.

17

Ok, this is fixed - all comment flags will be henceforth ignored for the purpose of this check, along with any "in need of moderator intervention" flags raised on the post itself (that exception already existed).

Thanks for pointing this out; by my estimation this was probably responsible for around 700 invalidated reopen reviews on Stack Overflow alone, with several hundred more across the network; not a crippling number, but far more than is reasonable.

You must log in to answer this question.

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