6
$\begingroup$

This is a follow-up to some comments at this question

If the new cs-theory-lite site is created (it is still in pre-beta), a tricky issue is which questions here might be migrated there. I want to ask the moderators to show restraint, especially at first, in migrating questions.

One area where this is particularly relevant is recursion theory, my research area. This area has clear applications to computer science, but certain topics are studied much more deeply as part of "mathematical logic", and really should be asked here. It would be truly unfortunate if questions in this area were moved from this site, where they are completely on-topic, to the new cs-theory-lite site, only because they happen to have the word "computable" in them. It's worth pointing out that many undergraduate CS curricula today have completely eliminated recursion theory, or have reduced it to part of a single semester class that also covers subrecursive hierarchies.

On the other hand, there are other topics such as "computational complexity theory" which may appear very related to someone outside the area (such as a moderator who is not an expert in the field). These could probably be safely migrated to cs-theory-lite. But distinguishing these from the ones that belong here is very delicate. When I look at the "computability" tag, I see some questions that belong here, some that should be migrated, and some that are unclear.

I do not want to wade into a political discussion about which areas are "really" computer science and which are "really" recursion theory; it just churns water. A less dramatic way to handle things, at least at first, might be for the moderators to assume that if something was on-topic last week, it will still be on-topic once the new site is created. If a question gets flagged for moving, by all means they could still move it.

$\endgroup$
9
  • 7
    $\begingroup$ You seem to be misinformed about the new site's scope as you keep using the phrase "cs-theory-lite". This is not what the new site is going to be. It will be for the whole field of computer science (and, by necessity, stuff from overlaps with other areas), of which TCS is only a (small?) part. Therefore, the overlap between math.SE and the new site is going to be so small that I think a panic is not warranted. $\endgroup$
    – Raphael
    Feb 11, 2012 at 1:49
  • 3
    $\begingroup$ If you are only afraid that questions that interest you might wander off to another site, be advised that using multiple sites on StackExchange is very easy. I know that mathematicians often sneer at computer scientists, but we don't bite and, believe it or not, we sometimes know our stuff. ;) $\endgroup$
    – Raphael
    Feb 11, 2012 at 1:52
  • 3
    $\begingroup$ Because there is already a "theoretical computer science" site for research-level questions, and because of comments I see in discussions about cs-lite, it seems like the main motivation for the new site is to handle questions that are not at the right level for TCS. Of course the name "cs-lite" is a joke, like "mathunderflow", just to distinguish the new site from the existing TCS site. $\endgroup$ Feb 11, 2012 at 1:59
  • 2
    $\begingroup$ I do agree there is relatively little overlap with the proposed questions; the main one that would be perfectly fine here is "Is it possible to enumerate all propositions entailed by a set of formulas in higher-order logic?". I simply want to avert any mass-migration. $\endgroup$ Feb 11, 2012 at 2:02
  • 9
    $\begingroup$ For some people that might have been the main motivation, but there are whole (sub)areas of CS that have no home at all, for instance networks, databases and programming languages in their academic flavors. We had to do a lot of (frustrating) writing on area51 to convince people that there is much more to CS than programming and TCS, which is (or should be) a trivial fact for every computer scientist. Therefore, I think it casts a wrong light on the proposal to call it the little brother of cstheory.SE (even in jest). $\endgroup$
    – Raphael
    Feb 11, 2012 at 2:13
  • 2
    $\begingroup$ I apologize if the term caused offense; none was intended. $\endgroup$ Feb 11, 2012 at 3:57
  • $\begingroup$ None taken; just wanted to clarify. $\endgroup$
    – Raphael
    Feb 11, 2012 at 13:00
  • $\begingroup$ I share your concern about mass-migration, but this is more about SE admins and Mathematics mods issue than about the new site. Some areas are shared between Math and CS (mostly TCS like computability, combinatorics, logic, ...) but there is also a much much larger non-overlapping part as Raphael mentioned: Systems (computer networks, operating systems, software engineering, ...), AI (computational linguistics, vision, machine learning, knowledge representation and reasoning, ...), computer/digital graphics, ... $\endgroup$
    – Kaveh
    Feb 13, 2012 at 20:08
  • $\begingroup$ As of March 20th, cs.SE is in public beta! $\endgroup$
    – Raphael
    Mar 20, 2012 at 22:22

4 Answers 4

9
$\begingroup$

I for one do not plan to move any existing questions to the new site... and, as usual, only move new ones when sufficient flagging occurs.

$\endgroup$
8
$\begingroup$

Here's my opinion on the issue.

  • Granting that there are overlaps between the purpose of the two sites, if a question is otherwise on-topic here at Math.SE before the other site materialised, it should still be on-topic afterwards (unless what is on-topic is re-defined).
  • If a question is otherwise off-topic here at Math.SE, it is still off-topic; but with the other site coming into being, instead of closing it flat out as off topic, we can redirect the question to a better home.
  • But just because a question is on-topic doesn't mean that it must stay on Math.SE. There are a few scenarios:

    • The OP realises that he can get a better answer at the other site, in which case we migrate on demand.
    • The question stays open and unanswered for a while, remains a marginal fit for on-topic here, but possibly much more on-topic at the new site. In which case the moderators have to make a judgement call.
  • When the new site matures and acquires a clearly defined expertise (like, in the case of Cross Validated), we as a community can decide whether to re-define our mission statement.

  • In the growth phase of the new website, the moderators here will have to rely largely on the expertise of the moderators of the new website in helping to make the final call on whether a question is migrated. (Between the description given by Carl and Raphael, I am getting the sense that the exact scope of the new site has not been completely settled?)

Also remember the progression of a new website coming into being: there is first a private beta phase, which graduates to a public beta phase, and which eventually to a full site.

  • In the private beta phase there should plainly be no migration there at all. (For one thing, the OP may not even be one who is privy to the private beta.)
  • In the public beta phase, questions should not be migrated because it is "a better fit there". A question should only be migrated when either
    • It is off topic here.
    • The OP requests it. (Perhaps after being prompted in the comments about the existence of the other site.)
$\endgroup$
17
  • 1
    $\begingroup$ "In the public beta phase, questions should not be migrated because it is 'a better fit there'." - on the other hand there was the precedent of a cryptography question being migrated to crypto.SE even if it is defensibly on-topic here... $\endgroup$ Feb 11, 2012 at 16:12
  • $\begingroup$ @J.M. jog my memory please? In the abstract I'd say that is a bad move. But I don't know if it is specifically off-topic or whether the OP requested it to be moved... $\endgroup$ Feb 11, 2012 at 17:02
  • 1
    $\begingroup$ I had this in mind... at least the OP managed to find the migrated question. $\endgroup$ Feb 11, 2012 at 17:51
  • $\begingroup$ @J.M. Just want to clarify that that particular instance was not one due to any of the Math.SE community mods. I hope it is quite clear that the above represent my opinion only; I wouldn't even claim that they represent precisely what Qiaochu, Zev, and Mariano (who has indicated his opinion elsewhere on this thread) think (though I suspect that they would not disagree with me by much). What the other ♦s do are certainly outside of our control... $\endgroup$ Feb 13, 2012 at 8:55
  • $\begingroup$ But does that question have mathematical content? The actual question seems to be asking about the philosophy about designs of crypto-systems, not about the mathematics behind them. I feel like it is one of those borderline on-topic cases where it depends on answerers to give a mathematical reading of the question itself. $\endgroup$ Feb 13, 2012 at 8:57
  • 1
    $\begingroup$ I know it wasn't a move by any of you guys (if it seemed I was pointing at you guys, I didn't intend to make it sound that way, and I apologize). I suppose it's because I certainly was reading the question with my mathematical glasses. Anyway, it's a done deal, and I just wanted to point out that moves not requested by the OP have certainly happened here before; I'm pretty sure digging deep enough should yield more examples. $\endgroup$ Feb 13, 2012 at 9:18
  • $\begingroup$ @J.M.: yes we do regularly migrate to physics, stats, and stackoverflow. But those are established sites, which are different from public beta sites. (Also, I felt that the clarification is needed because a lot of the readers of this thread [with < 10K rep] will not be able to see the original copy of the question you linked to.) $\endgroup$ Feb 13, 2012 at 14:58
  • $\begingroup$ If you want exceptions to the no migration to beta w/o OP request rule: this one was migrated because the OP asked both here and DSP, and received answer there. This one is just off-topic: it is specifically about how to do something in Scientific Python. This one is about mathematica... $\endgroup$ Feb 13, 2012 at 15:01
  • $\begingroup$ I mostly agree with your answer, expect this part: "But just because a question is on-topic doesn't mean that it must stay on Math.SE ...". It should be the OP's call in this case. The OP can cross-post on the other site if she decides to in these situations. Moderators should not migrate the questions which are on-topic here. $\endgroup$
    – Kaveh
    Feb 13, 2012 at 20:11
  • $\begingroup$ @Kaveh I thought users were generally discouraged from cross-posting? $\endgroup$
    – Raphael
    Feb 18, 2012 at 23:16
  • $\begingroup$ @Raphael, I don't see any problem as long as they are not simultaneous and the user keeps all of them up-to-date. $\endgroup$
    – Kaveh
    Feb 18, 2012 at 23:20
  • 1
    $\begingroup$ @Kaveh: it is, however, a general StackExchange policy to discourage cross-posting. $\endgroup$ Feb 20, 2012 at 10:05
  • $\begingroup$ @Willie, they also say that the sites are not run by them but by the communities. There are other SE general policies that you don't follow on math.se. There are also general policies that they don't follow themselves from time to time. It seems to me that these general policies are more like general guidelines based on experience and not formal rules to follow blindly at all times. $\endgroup$
    – Kaveh
    Feb 20, 2012 at 15:56
  • $\begingroup$ @Kaveh: to draw an analogue to the US constitution, things having to do with question migration and cross-posting is similar to interstate commerce. Since those situations will necessarily involve possibly inconveniencing other sites (as oppose to policies that apply only to our own happy little community), you have to really think twice before deciding to willfully ignore other people/community's opinion. $\endgroup$ Feb 20, 2012 at 16:04
  • $\begingroup$ @Willie, I agree about the general idea you stated (to some extent). But if you are not migrating then that interaction doesn't occur. And as a moderator on math.se you can't do much about a cross-post on another site. If they don't like cross-posts they can close it and explain it to the user. What I am suggesting doesn't mean that we encourage cross-posting on other sites, but only that it is fine to do it from math.se's perspective as long as it follows some reasonable expectations (links in both directions, waited a reasonable time without answer, keeps all copies up to date). $\endgroup$
    – Kaveh
    Feb 20, 2012 at 16:11
5
$\begingroup$

As I stated elsewhere, the advent of a new site does not change the scope of math.SE per se.

Regarding topics: what is part of CS and what is part of math changes from university to university and country to country, so I think if a questions chooses either site (for a question suitable to both), it is fine.

I would suggest to migrate questions that are ontopic here only on questioners' request (which might happen if the answers are unsatisfying). It might be a good idea to inform users of the new site, though, if it seems they do not get what they need.

$\endgroup$
1
  • 3
    $\begingroup$ Migrating only on request is also a nicely conservative option. I wanted to raise this now so the moderators have a chance to think about it before the new site goes live and they have to make actual migration decisions. $\endgroup$ Feb 11, 2012 at 2:00
0
$\begingroup$

I would like for even new questions not to be moved.

The trouble is, once five votes to exile a question have accumulated, it is impossible (or at least extremely difficult) for even a large majority of the community to take it back.

$\endgroup$
13
  • 1
    $\begingroup$ I think a question should end up where the questioner's needs can be best served. I have seen basic TCS questions answered correctly on math.SE but in a way that would (in my experience) not be helpful to a practically oriented CS student. The same may be true the other way round (mathematician with CS major asks on cs.SE) in the future. Jealousy among SE sites is not helpful, imho; we should take good care of questions and not much else. $\endgroup$
    – Raphael
    Feb 11, 2012 at 13:05
  • $\begingroup$ I don't agree with the second paragraph. Options for migration (which I think is what meant by exile) is not by default open to all other sites. Right now the only ones available to the masses are the targets Meta.Math, Stats.SE, and Physics.SE. There's currently no way for non-moderators to "exile" questions to, say, TCS, or the proposed Computer Science SE. And from our records, the questions which are voted to be migrated are all generally very clear cut cases where another site is much more suitable. I am not even aware of a single instance of someone wanting a question back. $\endgroup$ Feb 11, 2012 at 13:26
  • $\begingroup$ I therefore feel that your second paragraph is just a strawman (under the circumstances). And while I appreciate your opinion in your first paragraph, can you perhaps given an explanation why? $\endgroup$ Feb 11, 2012 at 13:30
  • 2
    $\begingroup$ @Willie, I was assuming (given the large overlap between math and CS) that the new site would be added to the list of migration targets. $\endgroup$ Feb 11, 2012 at 14:12
  • 2
    $\begingroup$ @Willie, as for "an explanation why", I thought that was obvious. It would make this site significantly less fun if the computer science questions were taken away from here. (And there's no way I'm going to add another site, with separate question lists, favorites page, rep counter, etc. to my procrastination program). $\endgroup$ Feb 11, 2012 at 14:22
  • 1
    $\begingroup$ @Henning: sites are not automatically added to migration targets. Moderators have to specifically request that from the Stack Team, and the request have to be backed up by a demonstration that there are a lot of off topic questions posted here by mistake, and that our site would benefit from giving users the ability to choose to migrate questions there. For example, it was only fairly recently that Physics.SE was added to the list of targets by my request. $\endgroup$ Feb 11, 2012 at 16:56
  • 1
    $\begingroup$ And I asked you for an explanation precisely because it was not obvious why you phrased a categorical objection to migrating questions to another site. If your reason is that you think "those questions belong here" because they made the site more fun etc, perhaps you should also clarify what you mean by "those questions". Like Raphael mentioned, there is no universally agreed line between computer science and mathematics, and I certainly cannot guess which are the questions you'd like to see stay here. $\endgroup$ Feb 11, 2012 at 17:00
  • $\begingroup$ @Willie: I don't mean anything by "those questions", since I didn't say that -- in fact your comment above is the first and only time the word "those" is used on this page. I wrote "the computer science questions" -- what part of that do you think needs clarification? It means questions asked here that are about computer science. Exactly because there is not bright line betweeen computer science and mathematics, I think it would be wrong to tell any asker of computer science questions that his subject now has its own ghetto and he is not welcome with it here. $\endgroup$ Feb 11, 2012 at 17:38
  • 3
    $\begingroup$ @Henning: I think you misunderstand. I am asking for clarification on what you want to be welcome here. For example, to quote Raphael, "networks, databases, and programming languages in their academic flavors" are subsets of CS. But I would not consider them to be on-topic questions here on Math.SE, regardless of whether there is a CS StackExchange site. I want to know what you consider as Computer Science that overlaps with Mathematics, and especially about those edge cases so we don't close/migrate questions that you think have mathematical content. $\endgroup$ Feb 11, 2012 at 17:57
  • $\begingroup$ In particular, I am asking you whether by "computer science questions" you mean "the subset which is about mathematics" and/or "like those ones which have now an established presence on Math.SE", or do you intend for the scope to be even broader (or narrower, take your pick). $\endgroup$ Feb 11, 2012 at 18:00
  • 4
    $\begingroup$ @Willie, I'm a lumper: I'd welcome anything that has any arguable connection with mathematical reasoning. Algorithmics, complexity theory, recursion theory, formal languages, etc. all have a solid presence here, and I would really like to see some questions about programming language theory too. There could be a gap in network and database theory, where some questions would be too abstract for SO yet don't feel like mathematics. But I wouldn't want questions about queueing theory or relational algebra to be rejected just because they are network or database motivated. $\endgroup$ Feb 11, 2012 at 18:20
  • $\begingroup$ Thanks for the clarification. That was helpful. $\endgroup$ Feb 13, 2012 at 8:47
  • $\begingroup$ It is always possible for OP to cross-posting his questions on another site, so migrating questions which are already on-topic on here is not needed at all IMO. If the question doesn't get a satisfying answer we can suggest another site, but the question should remain open here as long as it is on topic on this site and the OP wants to keep it here (as Raphael wrote). $\endgroup$
    – Kaveh
    Feb 13, 2012 at 20:19

You must log in to answer this question.

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