Should we automatically close old zombies?

Question

There are a few views of "unanswered" questions on Code Review:

• https://codereview.stackexchange.com/questions?tab=Unanswered ("Questions" on left sidebar, "unanswered" tag.) This is ordered by vote count.
• https://codereview.meta.stackexchange.com/unanswered ("Unanswered" on left sidebar). This is ordered by a custom choice which is saved, vote count by default.

When ordering by vote count, the top posts are from the 2010s. Personally, if I wanted a code review of a 1-page code snippet in 2014, I would not still want a code review, so I also don't review these as a reviewer. I understand that there's instead a push to review all of these, but I admit I don't really understand why.

Would it perhaps make sense to close very old questions automatically on this particular exchange?

Answer 1

No

Personally

Personally if someone reviews one of my old questions and teaches me something new I would be as grateful as if the question were one day old. Again the age of the code or question doesn't matter, I could have been making the same mistake since 2014! Now hopefully I've not been so silly to have made a mistake for so long, but how can I know that if my question has been closed?

My opinion on the matter is very different than yours:

Personally, if I wanted a code review of a 1-page code snippet in 2014, I would not still want a code review, so I also don't review these as a reviewer.

Close Reasons

Our close reasons exist to fix problems.

• Our no. 1 reason, Code Not Working As Intended, is a key close reason and stops our site from become Code Debuggers or Stack Overflow+.

• Our Author or Maintainer close reason helps prevent our answerers from legal and moral quandaries. If someone illegally posts code under the wrong license we'd rather our answerers not to get in trouble. Additionally, we had someone come to our site asking for a review of some code to determine if someone should be fired! We as a site decided to not condone this.

But your suggestion is to employ symptomatic fixes, not root cause fixes.

• the reason I care is that I use the unanswered questions list as a "TO DO" of things to review. Arguably, I should do something else

  Yes the issue is SE search or your workflow, not close reasons.

• I understand that there's instead a push to review all of these, but I admit I don't really understand why.

  The issue then should be why we have zombies and how to stop questions from becoming zombies, not how to kill them once they've become zombies. Or to question our push to review zombies.

Stack Exchange

Stack Exchange is intended to be a searchable resource for everyone. Have a question? Google it and read the Stack Overflow/etc. answers to fix your problem. I've only ever posted a handful of questions on Stack Overflow, but Stack Overflow has answered thousands of my questions.

Now you could argue Code Review isn't like Stack Overflow. However I've had answers be upvoted years after the question was active because of this. So my answer was helpful for someone who wasn't the OP.

As such your reasoning doesn't take into account people who aren't the OP. And moves away from the Stack Exchange model.

Overall

I appreciate your anecdotal evidence is different than mine. However your anecdotal reasoning doesn't match the Stack Exchange model or the reasoning behind our existing close reasons.

Additionally your suggestion will impact users (like me) and cause them to have a negative experience of the site. For what benefit?

Answer 2

No. Automatically closing old, unanswered questions is not a good idea.

There are a variety of reasons for a question to be unanswered. This includes a poor question (in which case it will likely get downvoted and deleted). Or it could be some good, well-written code that many (re)viewers are unable to find any suggestions for improvements. Or maybe there is nobody else (other than the OP) that understands the code well enough to be able to comment.

Since these are old questions, and have received several (or even many) upvotes, the questions are good. Since there are no answers, the code in the question likely is reasonably well written to accomplish whatever it is supposed to. Therefore, leaving them on the site can show to others how to solve that problem, or what is apparently some well-written code can look like.

This can make one wonder: "If I don't see any problems with the code, should I leave an answer that says so?" Probably not, because even if you don't see a problem doesn't mean someone else won't. (And the fastest way to have that happen is to leave an answer that says there are not problems with the code.)