2
$\begingroup$

If I post a question on CS.SE and don't get an answer, is it OK to post the same question on TCS.SE after waiting a week or so? In other words, is cross-posting on CS.SE and TCS.SE OK (rather than migrating) if there is a delay of a week between them?

I'm asking to see if I can clarify the policy, because I am confused.

To be clear: I am not asking about simultaneous cross-posting. I'm asking about cross-posting after a delay of a week, when the question didn't get an answer on the original site. I'm not talking about migrating a question or deleting the original one and posting it on a new site; I'm talking about cross-posting, where the question is open on both CS.SE and TCS.SE.

Here is what I have been able to find:

  • The standard StackExchange policy is: it is not OK to cross-post the same question on multiple sites. If you post it on one site and after a week or two don't get an answer, you can flag it to ask the moderators to migrate it, but you should not cross-post (not even after a week or two).

    Source: Is cross-posting a question on multiple Stack Exchange sites permitted if the question is on-topic for each site?

  • Currently, the TCS policy says that it is OK to cross-post on TCS.SE and MathOverflow, if you post on one and don't get an answer even after waiting a week or so (and if you must cross-link the two). However this doesn't say anything about other SE sites other than MathOverflow, so it's not clear if this applies more broadly.

    Source: https://cstheory.meta.stackexchange.com/a/231/5038

  • There was a proposal to change that policy to allow simultaneous cross-posting on TCS.SE and MO, but that proposal was marked status-declined (it was rejected).

    Source: Changing policy on crossposting

  • Peter Shor points out that one of the close reasons says

    Our site policy prohibits simultaneous crossposting: it duplicates effort and fractures discussion. Crossposting is permitted after a week has passed without a satisfying answer elsewhere. When crossposting please summarize the relevant discussions from other sites in your question and link between the copies in both directions.

    This sure sounds like cross-posting after a delay is allowed. On the other hand, when you click on the link to the "site policy", the policy described at the link target actually only mentions MathOverflow, not other SE sites. (I hadn't noticed this previously, because I don't have enough reputation to vote to close.)

  • CS.SE does not seem to have any special policy about this, and I can't find any discussion of this on Meta.CS.SE. Therefore, I would assume that CS.SE abides by the standard StackExchange policy, by default.

    Also, given the standard StackExchange policy and the impact on both sites, I would guess that probably cross-posting on two SE sites should only be allowed if both sites opt into a policy of allowing cross-posting. CS.SE does not seem to have opted to allow cross-posting on CS.SE+TCS.SE, and since I can't find any clear evidence that TCS.SE has, either. (It appears that before adopting the current policy regarding cross-posting on TCS.SE+MO, both TCS.SE and MO were consulted; however, I can't find any evidence of any consultation with CS.SE about cross-posting on CS.SE+TCS.SE.)

My conclusion: It's not clear to me what is allowed. The information on the policy does not seem entirely self-consistent, to my eyes.

So, what is the current policy? Is cross-posting on CS.SE and TCS.SE allowed, if you wait a week?

And if this has not been previously addressed, what should the policy be?

$\endgroup$
  • 2
    $\begingroup$ Why do I have a different impression? One of our custom closing reasons says: "Our site policy prohibits simultaneous crossposting: it duplicates effort and fractures discussion. Crossposting is permitted after a week has passed without a satisfying answer elsewhere. When crossposting please summarize the relevant discussions from other sites in your question and link between the copies in both directions." $\endgroup$ – Peter Shor Jul 1 '14 at 10:00
  • $\begingroup$ @PeterShor, good find! Thank you. I've updated the question with this information -- that's significant. $\endgroup$ – D.W. Jul 1 '14 at 15:37
3
$\begingroup$

The policy is general. MO is mentioned only because it was the only site where this happened regularly with at the time of the discussion. In practice the policy has applied to all other sites including CS.SE.

The policy is about what we do on cstheory. Other sites are free to act as they see fit, e.g. they can close their copy of a cross-posted question if they want to. We on cstheory don't do so as long as the requirements in the policy are satisfied, and I don't see a reason to forbid cross-posting as long as the requirements in the policy are satisfied. Note that it is not just one week delay, the OP also accepts the responsibility to keep both copies up-to-date. The benefits are: 1. other cstheory users interested in the question who do not visit other sites will get any new information about the question, and 2. cstheory users can post new answers if they want to.

I think the policy has worked reasonably well here on cstheory where users are usually keen to share what they find out about their question with the rest of us, and any improper uses of cross-posting have been dealt with well using the current policy. Whether the same policy would work on other sites I don't know.

$\endgroup$
0
$\begingroup$

If this has not been addressed before, here is my proposed policy (to allow others to vote on it): We should not allow cross-posting on CS.SE and TCS.SE.

Of course many people might like to get more attention by posting their question on as many places as possible, but that doesn't mean it's good for the community for that to happen. It has some costs:

  1. It fragments discussion and answers.

  2. Inevitably the two will diverge. For instance, there might be a good answer on one but not the other. Even if they are cross-linked, this makes it harder than necessary for others to discover the answer to the question.

  3. It makes it harder for others who have the same question to find the question and benefit from the answers we've built up. Splitting answers across two pages dilutes the "Google juice" of both.

These problems are explained well at the StackExchange page on this issue.

Migration solves the above problems.

However, migration has its own costs:

  1. If you migrate, the question disappears from the old site. The old URL doesn't stop working -- it redirects to the location on the new site, so Google searches that turn up the old URL will still get you to the right place -- but search on the new site will no longer turn up the old question. This makes it impossible for people who have the same question to find the old question by searching on the old site. To find it, they'd have to either search on Google or search on the new site. So, if the question was squarely within scope on the old site, this makes it harder to find the information.

This means there are some pros and cons either way.

Weighing the pros and cons, I'd suggest we stick with the default StackExchange policy and prohibit cross-posting on CS.SE and TCS.SE (cross-posting shouldn't be allowed, not even if you wait a week or two for an answer and get nothing).

$\endgroup$
  • 2
    $\begingroup$ If you migrate, the question vanishes from the original site. This makes it impossible (not just harder) for people who have the same question and are just searching within cstheory.SE to discover the question. This also will lead to duplicate effort. $\endgroup$ – Peter Shor Jul 1 '14 at 10:10
  • $\begingroup$ @PeterShor, good point, thanks! (Minor point: the URL for the question on the original sites still works; it just redirects to the new location on the site. See, e.g., cstheory.stackexchange.com/questions/14820 But as you say, search on the old site won't find it.) $\endgroup$ – D.W. Jul 1 '14 at 15:27
  • $\begingroup$ Btw, on csthoery we separate the discussion from voting. To pass a policy you need to post a new question with explicit guidelines for voting after the discussion. See this. $\endgroup$ – Kaveh Jul 2 '14 at 23:11
  • $\begingroup$ Thank you, @Kaveh, for all the information! Much appreciated. $\endgroup$ – D.W. Jul 2 '14 at 23:32

You must log in to answer this question.

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