MathOverflow works very well for mathematicians, in part because it is run by mathematicians, who understand how mathematicians interact. One clever thing they do to make the site useful for the research community is to redirect some questions to the math.SE site. That seems to work very well. (When they have spare time, they do visit math.SE and answer questions.)

The proposal to create a similar place to redirect basic TCS questions was quickly dismissed on area51, because people there disagree with having two sites on the same topic with different lower bounds. On the other hand, they have nothing against the MathOverflow/math.SE model because they view MathOverflow as being outside SE. (Which, frankly, mixes technical issues --- who runs the site -- with social issues --- who participates in the site.)

It seems to me that both SE people and TCS researchers might be happy with migrating the TCS researcher community outside the SE network and have it run by TCS researchers. What do you think?

PS: Just to be clear, I do not feel that we should move, but I think it's an option that is worth discussing.

  • 4
    $\begingroup$ We rely very heavily on the SE framework, so moving away seems unlikely. Though, to be honest, I do not know what it would mean to move CStheory outside of the SE network. $\endgroup$ Commented Feb 11, 2011 at 16:33
  • 3
    $\begingroup$ I actually think we could allow slightly more basic questions on CStheory than we do now. The problem is that we want to avoid being overrun with elementary questions in the way that has happened to the physics stackexchange site, say. $\endgroup$ Commented Feb 11, 2011 at 16:37
  • 3
    $\begingroup$ Keep in mind that an MO-like independent site isn't a technical option anymore. They are no longer licensing the software, so if you want a site on the SE engine, you need to be a part of the network. $\endgroup$ Commented Feb 11, 2011 at 16:41
  • $\begingroup$ @Dave: It doesn't have to be short-term. @Kaveh: I said "they", not "SE". I'll fix to be more specific. @Peter: I do not argue for keeping the same lower bound we have now. I'm only saying that having another site on the same topic with a lower lower bound seems to (1) work well for MO and (2) be unwelcome on area51.SE. $\endgroup$ Commented Feb 11, 2011 at 16:43
  • 1
    $\begingroup$ There are decent clones out there, but I do not know wether they are on par technically. Still, I think establishing a general computer science SE and advertising cstheory as place for research-level TCS would be the best way to go for everybody. We probably should not take prejudices voiced by people that have apparently absolutely no connection to academia to heart too much. $\endgroup$
    – Raphael
    Commented Feb 11, 2011 at 19:38
  • $\begingroup$ After several aborted proposals, there is now a computer science site proposal that's on the right track. $\endgroup$ Commented Jan 28, 2012 at 17:12

2 Answers 2


This whole 'throw the bums out' discussion keeps cropping up. I would like to encourage those feeling this way to think carefully about what it means:

  • someone has to get the software installed (not hard if we used OSQA)
  • someone has to maintain the software and add in patches as and when users make requests (this takes more time, much more than moderation, and also costs money)
  • we have to move a large number of users over, and we're bound to lose a lot of them in the process. theoryCS folks are not the most tech-savvy around, you know :)

In addition, we should think carefully about what real objections we have to the SE system. In any partnership of this kind, there is always strain caused by the one-size-fits-all model. I personally have occasionally found some things onerous (the various thresholds for capabilities being one of them), but I've been able to either work around it, or wait for problems to disappear. Given that the cost of moving is high, the pain induced by staying should be equally high.

My personal opinion (admittedly biased by the time I've already sunk in): I think that overall SE has been a very reasonable home for us. While there's always chatter, the SE admins cannot be held responsible for this: the whole point of the community drive voting process is that it's community driven. So I'm not interested in supporting calls to move unless the arguments for moving outweigh the costs.

  • 6
    $\begingroup$ I agree. The benefits of staying far outweigh the costs of going at this point. $\endgroup$ Commented Feb 12, 2011 at 1:25
  • 1
    $\begingroup$ Good points on the CP-ratio and community-driving. Given the fact that we are still slow-growing even with all the bottomlines and constrains we have, I think it would be a good idea to stay here for a while. I somehow like the way that we can have disscussion with other SE people. Communication benefits both of us. $\endgroup$ Commented Feb 12, 2011 at 6:08

(Moved from comments)

SE has not made it clear. It is only some users who have stated their own opinion, they might have high reputation on Area 51 but in no way they are representatives of SE. The discussion about the two level sites is going on Area 51 and I think so far it has been a constructive discussion. I think we can learn from the points they are making and to be honest some of the points seems valid to me to some extend and we can address them and that would improve cstheory (addressing them does not mean dropping research level requirement we have).

I see migrating from SE to our own site as a possibility, Suresh had a post about what software we should use on his Blog before we decided to use SE, and since the content of cstheory (questions/answers/...) is copyrighted under CC we can copy it (using the data dump or data.stackexchange.com API with some technical difficulties) but at this point I don't see any reason for doing it. The people who really represent SE (like Jeff, Robert, Joel, ...) have been quite welcoming and respectful to us, and are intelligent people and their software is quite good (and improving all the time), so I like being part of SE network and I think that the relation between us and SE is respectful and beneficial to both sides. So I think that at this point there is no need for a discussion on moving from SE network. We can discuss it in future if the situation changes.

  • $\begingroup$ Thanks for the answer. Alas, the discussion on Area51 (about two sites on the same topic at different levels) does not seem very constructive to me. I see no point in ever moving the data. $\endgroup$ Commented Feb 11, 2011 at 17:11
  • $\begingroup$ @Radu GRIGore, I think it is relatively constructive though a little bit heated. It seems to me that some of them fear that it is like a Pandora's box, if they allow it once everyone would like to do the same, and honestly I am not sure if a two level model would make sense for things like cooking. On the other hand, I think we have good reasons to think that the two level model makes sense for theoretical scientific research topics like TCS and Math (and probably Physics). $\endgroup$
    – Kaveh
    Commented Feb 11, 2011 at 18:12
  • 3
    $\begingroup$ @kaveh "honestly I am not sure if a two level model would make sense for things like cooking" in fact this part of your comment shows the real point to me. People outside our community does not see why we prefer 2 sites instead of one. To relate to cooking, if you talk to a chef, he will maybe have the same idea as we have, but for a cooking Q&A site. $\endgroup$ Commented Feb 12, 2011 at 16:12
  • $\begingroup$ @Sylvain Peyronnet, it might be the case, I am not a chef, but IMHO there is a clear distinction between the two cases. I haven't spent much time on the cooking site but my impression is that the distinction that I have in mind between research level questions and non-research level questions does not exist there. Note that the distinction I am making is not a general distinction between professional/expert site vs. newbie/amateur site, $\endgroup$
    – Kaveh
    Commented Feb 12, 2011 at 19:05
  • $\begingroup$ [continued] i.e. it is not a distinction based solely on the skill level, it is between active research questions vs. non-research/general level questions. Many good questions on this site (which make the site interesting and valuable in the first place) are so specialized that only a handful of people would know the answer to them. The number of people who regularly attend say FOCS or STOC is way below 500 and we want to have as many of them as we can on the site. And even between them the topics are highly specialized, $\endgroup$
    – Kaveh
    Commented Feb 12, 2011 at 19:10
  • $\begingroup$ [continued] a top expert in approximation algorithms might not understand talks on Crypto and and vice versa. Gaining expertise in even smaller more specialized research topics in TCS needs spending years on that small single topic (i.e. PhD). This distinction would make sense for cooking only if there was a similar amount of highly specialized active scientific research going on in cooking, in which case it would make sense to start a site for cooking researchers (and not for chefs!). $\endgroup$
    – Kaveh
    Commented Feb 12, 2011 at 19:18
  • 2
    $\begingroup$ @Kaveh, I strongly disagree: great chefs are cooking researchers. :) (I really do believe this. I know of one chef collaborating with chemistry researchers in an uni.) $\endgroup$ Commented Feb 12, 2011 at 21:21
  • $\begingroup$ @kaveh Somehow I agree with you, but my point is that if one is not in a community, how can he knows for sure what's good for the community ? The closing of the beta TCS is a good example : it was closed long before it is possible (even for us) to know whether or not it can preclude the development of cstheory... $\endgroup$ Commented Feb 12, 2011 at 21:25
  • $\begingroup$ @Sylvain Peyronnet: I think I now see your point: it might be better to leave the the decision about these issues to those who are familiar with the topic, an outsider like me might misunderstand the situation. $\endgroup$
    – Kaveh
    Commented Feb 14, 2011 at 15:16
  • $\begingroup$ @Kaveh : that's exactly my point. $\endgroup$ Commented Feb 14, 2011 at 17:27

You must log in to answer this question.

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