This question comes off the back of a question I asked on Stack Overflow Meta about the grace periods of answers. Specifically it was my confusion that I, personally, cannot submit answers on questions that are closed when I was already posting my answer, but that I continue to see that others can do so. This behaviour seems pretty odd to me, as why should some be able on a question that was closed while others (like myself) can't even submit the answer they had spent typing when the question is closed seconds before they click the submit button.

More interesting, it seems the users that can post questions after the question is closed don't even need to have the answer "in flight". For the linked question the user who's answer I used as an example stated the below in the comments:

(2) I did not start composing the answer before it was closed; I usually type my answer in a text editor before copy/pasting it to the site, which is what I did here.

So this means that for myself, who could be writing an answer before the closure but tried to submit after it, I would be denied the submission, but for others that didn't even start to enter their answer (on [so] at least) they still could. (Note that in the example in the linked question I closed as a dupe, so I obviously wasn't trying to answer that question.)

That isn't to say I "mind" that the answers were submitted, however, the lack of consistency seems odd here. I suspect that this has something to do with Javascript or something else being enabled/disabled (I'm not a web dev so I have no idea how those technologies work). Personally, however, I would expect those that are already answering should still be able to submit answers if others can so easily circumvent it.

Should the "answer in flight" feature that is discussed in this old answer no longer in play? If so, why was it the decision of Stack Exchange to have such different behaviour for different users? Should it not be more consistent for everyone? Although I appreciate that if the client side of things is blocking it, should the client and server side validation be made to "agree"? To quote one of the comments from the linked question as well:

Seems to me this is a pretty poor design. Especially now that the site continuously saves your work (so it's easy to detect that an answer is being edited), the policy should be changed: if you start writing an answer before a question is closed, and continue working on it through its closure, and post afterwards (for up to 4 hours) it should simply be accepted and posted. It's frustrating as hell to work for an hour on writing an answer, and then have it thrown away because 5 other people didn't understand the question.

I couldn't agree more with that sentiment; it is incredibly frustrating to put effort into answering a question only to be denied your efforts by mere seconds or minutes; especially when the vote the reopen process can be incredibly slow or when someone else posts an answer that may well show a significant less amount of effort you know you put in yourself. If the answer is being continually saved, then surely the server and client sides both know that the answer was started before the closure and should be accepted, no? If the answer truly isn't helpful/useful then others can still downvote it, just like any other answer; closure doesn't stop that.

Feature Request

The behaviour for users who try to submit an answer to a closed question should be similarly consistent for all users. Those that have client side validation enabled shouldn't be able to not post answers at all vs those that don't have it enabled. At best, those with it enabled but have an answer "in flight" should still be permitted a "grace period" to allow them similar behaviour if not the same.

Or perhaps, answers should just be blocked for all users, regardless of if they have client-side validation. Either is "fine", my point is that the experience should be consistent.

Note: In case some are wondering that I (others) shouldn't be trying posting an answer on a question that is going to be/is being closed I disagree with this. The fact that I am trying to post means that I likely implicitly disagree with the closure reason. The only exception to this would be a duplicate where I didn't find the candidate in my own searching. Though I know some users on Stack Overflow that are yet to figure out what the "Close as Duplicate" feature is, I am not one of those and that isn't what this question is about.


You must log in to answer this question.

Browse other questions tagged .