Closed questions that remain closed are meant to be deleted eventually. Closure (for a reason other than exact duplicate) indicates that the question is not suitable in its present form. Ideally, the question should be improved to make it suitable and be reopened. If the improvement doesn't happen, the question should be deleted.

Users with at least 2000 reputation can vote to delete closed questions, but the tools for that aren't great, so moderators typically end up doing it. As long as we have a low volume of questions, we can sustain a community review for closed questions.

So here's the deal: this is the list of questions that have been closed for at least 30 days, for a reason other than exact duplicate or migration. Please review these questions. If you think one of these questions should be reopened, answer here in defense of the question. If the question requires improvement, please edit it! This especially goes for poor questions with a good answer — don't hesitate to make a heroic edit to save the thread.

This review will go on for a week, until some time Monday 27 August. If you defend a question, please vote to reopen in addition to posting here, if you have enough reputation. Moderators may help with the reopening to ensure that a question doesn't stay forgotten merely because it couldn't get enough reopen votes. Questions that remain undefended after the review period will be deleted.

A week's review period and an extra week later, 4 questions remain contested: “someone should edit them”, but nobody's doing any editing… Edit these questions into something suitable for the site, or be silent forever.


4 Answers 4


After the announced week plus one more week, I went and deleted all the questions that nobody defended.

Two questions were edited since this review began:

Congratulations to A Schulz and Kaveh for getting into the spirit of the thing! However, in both cases, the question was unanswered, the original asker hasn't returned to the site since soon after posting the question, and the question is not of great value (one is a homework exercise, the other is a reference request). I think these questions would get better treatment if someone who cared posted them in their own words, so I'm deleting the existing questions.

(The morale of this, I think, is edit early. We'll try to do these close reviews more often, but they can't replace improving the question as soon as it is asked, when the asker still cares.)

Of the questions that were defended as “someone should edit this”, I'm deleting:

I'm leaving the following questions open for now, because they have answers that are worth salvaging. But they can't stay forever: either argue that they should be reopened in their present state, or edit them to make them suitable.

I reopened one question:

  • $\begingroup$ Given our lack of a policy, "List of intro TCS books" can not be edited to be reopened, imho. I would still like to keep it around until we have such a policy, if only to not break links to it. $\endgroup$
    – Raphael
    Sep 2, 2012 at 15:15
  • $\begingroup$ I edited Resources about theory of computation... to focus more on the quest for understanding rather than resources. I think it's good enough to keep, even if it is not a pure CS question. $\endgroup$
    – Raphael
    Sep 2, 2012 at 15:26
  • $\begingroup$ @Raphael Ok, thanks, I reopened Strategies for becoming unstuck in understanding TCS . Still not a great question but better than it was. $\endgroup$ Sep 2, 2012 at 15:30
  • $\begingroup$ I also edited How does shotgun hill climbing differ from normal hill climbing?. I think it's a good question now (without all the gutter about normal hill climbing that was easily answered by Wikipedia). There is most likely not a good answer to the question (it is a heuristic) but an honestly asking OP might not know that. An answer should be able to shed some light on the issue either way. $\endgroup$
    – Raphael
    Sep 2, 2012 at 15:41

I think those can be saved:

And I'd like to keep List of intro TCS books for those who don't know much about TCS around for the day we have a clear policy on list questions.

  • $\begingroup$ I think all three would require a heroic edit. I have no idea how to save them. As for the list questions, Stack Exchange does have a policy, even if we aren't applying it. $\endgroup$ Aug 19, 2012 at 23:19
  • 2
    $\begingroup$ @Gilles: I'll try and edit the three next week. About the list question, we have linked to it several times since it's closing so it arguably has its use even if it remains closed forever. $\endgroup$
    – Raphael
    Aug 20, 2012 at 5:55
  • 1
    $\begingroup$ Given that the Critera question was replaced by a better question asking the same thing, it probably should go. $\endgroup$ Aug 21, 2012 at 11:03
  • $\begingroup$ @DaveClarke: Can you please link to the one you mean? Would it be feasible/worthwhile to merge the answers there? $\endgroup$
    – Raphael
    Aug 26, 2012 at 19:54
  • 1
    $\begingroup$ cs.stackexchange.com/questions/1954/…. No need to merge answers, as both were copied across by their authors. $\endgroup$ Aug 26, 2012 at 20:50
  • $\begingroup$ I've left the “can be saved” questions undeleted for now, but someone needs to edit them into shape. (Not me because I don't see a way to salvage them.) We do have a policy on list questions (it's SE-wide), even if we aren't applying it. $\endgroup$ Sep 2, 2012 at 14:15

I will use these symbols:

  • V: vague
  • B: too broad
  • NR: no-reply from the OP yet
  • G: should be generalized
  • W: wait, unclear policy
  • D: duplicate
  • OT: off-topic

Can be saved by others:

Vote to delete:


But it seems that we already have a more general question regarding how not to solve P vs. NP problems. So no need to keep this one.

This is also replaced by a better question.

  • $\begingroup$ IMO: NARQ/broad and NARQ/vague questions can be saved by asking the OP to be more specific, if the OP doesn't reply in a reasonable time and there are not any upvoted answers then it is fine to delete. $\endgroup$
    – Kaveh
    Aug 20, 2012 at 10:00
  • $\begingroup$ Other NARQ questions which are specific enough to get an answer can be saved by a little bit of generalization. $\endgroup$
    – Kaveh
    Aug 20, 2012 at 10:02
  • $\begingroup$ The clearly OT questions (undisputed) can be deleted. $\endgroup$
    – Kaveh
    Aug 20, 2012 at 10:03
  • $\begingroup$ The disputed OT questions seems to be a result of clarity of policies and should not be deleted (particularly if they have upvoted answers). $\endgroup$
    – Kaveh
    Aug 20, 2012 at 10:04
  • $\begingroup$ Duplicates can be deleted or merged with their copies on other sites. $\endgroup$
    – Kaveh
    Aug 20, 2012 at 10:05
  • $\begingroup$ I changed the numbers back to links, otherwise it's difficult to figure out what question you're talking about. Are you planning to edit the “can be saved by others”? As for the “can be saved by OP”, they've already had a months to do it, so by now there's no hope. $\endgroup$ Aug 20, 2012 at 21:09
  • $\begingroup$ @Gilles, thanks. I can try to fix a few over the next week. I agree if the OP has not replied in a month it is likely they won't so it is fine to delete them. $\endgroup$
    – Kaveh
    Aug 21, 2012 at 14:25
  • 1
    $\begingroup$ Please see my answer. I left the “can be saved by others” questions that had potentially valuable answers undeleted, but someone needs to go and edit them (I, for one, don't see a way to salvage them, which is why I'm not doing it). $\endgroup$ Sep 2, 2012 at 14:13

I don't imagine this will be popular, but I wouldn't mind keeping this one around: Variant of TSP in P

It seems like it is clearly within the scope of CS.SE to ask about a purported proof of P vs NP. The author isn't trying to argue anything, just asking if there is an easily-observable flaw in the proof which renders its conclusion worthless (which there is, as pointed out in a highly-voted answer). Other people who are wondering about other purported proofs could run across this question in their searches and be helped by this information (does the proof presuppose the conclusion? is a lower bound applicable to all solutions or just one which the authors explain?).

It looks like a fairly harmless "check this proof" question to me. I suppose that if we are not taking questions of that kind, it has no business here... but I don't remember whether that is the actual policy we want to have.

  • $\begingroup$ There was a meta thread about that question which ended up validating the closure $\endgroup$ Aug 22, 2012 at 22:15
  • 1
    $\begingroup$ @Gilles Ah, fair enough. I'm not sure I'm in love with the idea of introducing another subjective criterion by virtue of which we can close (and delete?) questions, but if that's what we're doing, good enough. $\endgroup$
    – Patrick87
    Aug 22, 2012 at 22:25
  • 1
    $\begingroup$ Deleted since the meta thread shows that the community doesn't want this question. $\endgroup$ Sep 2, 2012 at 13:45

You must log in to answer this question.

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