5
$\begingroup$

There is some concern that this new site may bifurcate the statistical computing community that has been extremely active over the last several months at Stackoverflow under the R tag. The thought is that some, or many, of the users asking and answering on SO may migrate here because their work is more closely related to the statistical interpretation of some R function, rather than purely programatic.

I realize that that part of this question was already adressed here; however, it would be very useful if there were some integration between the two sites. I know that at some level the sites are integrated, but in an effort to keep the community unified it would be extremely beneficial if things like search and 'Related Questions' were integrated between the two.

This is meant as more of a question for discussion, but if someone knows whether or not this is even possible I would appreciate it.

Summary: It would be nice to integrate "search" and "related questions" between this site and the [r] tag on stackoverflow.

| |
$\endgroup$
  • 1
    $\begingroup$ That would be a great feature to have, and I'm sure other sites would benefit. As I understand it, the StackExchange team monitors the feature-request tag, so let's see what they think. $\endgroup$ – Shane Jul 26 '10 at 21:26
  • $\begingroup$ I think StackOverflow meta is more general to the whole SE universe, so it should be asked there. Nevertheless, I support; when more sites will go out of Area, this will became a very serious issue. $\endgroup$ – user88 Jul 26 '10 at 21:33
  • $\begingroup$ And, if it is a discussion, it should be community wiki. $\endgroup$ – user88 Jul 26 '10 at 21:33
2
$\begingroup$

Another solution to this problem is a series of amazing applications that will be delivered along with the new Stackoverflow API. It would be trivial to mix sites using this.

| |
$\endgroup$
1
$\begingroup$

Drew, I have the same concern. I've watched the [r] tag community grow on Stack Overflow since the early first "flash mobs" where the site was seeded with good questions and answers. That effort even got mentioned as a "good thing" by Joel and Jeff in the Stack Overflow podcast. It took a long time to get enough good questions and answers on SO for it to be a useful resource. It's also beginning to bubble up good R results in Google, which is very valuable.

Being an economist I tend to look at Q/A sites as a market. For the market to clear there has to be 1) good questions 2) good answers 3) long term providers of 1 & 2. If any becomes absent then the market fails to clear. I define "clearing" loosely as "questions get reasonable answers in a reasonable amount of time for the majority of users." Some of those users won't have to type in the question because they come to the site, find the answer, maybe vote an answer and/or question up and then they go away with what they came for. That market just cleared. What has helped SO become a clearing market for [r] questions is the concentration of many good questions where lots of folks find answers before they type questions and lots of fast answers when users do type in a question.

So coming back to Drew's comment: If the set of R users (which is very small compared to Python, Java Script, PHP, etc.) gets split across multiple Stack Exchange sites I'm concerned that the R questions on all of those sites will fail to clear.

If I was "King of the Internet" (registered trademark of Al Gore) I'd make all questions that are mostly syntactical go to Stack Overflow and let the ones that are more related about an algo go to a site devoted to that algo or concept. But even as King that would be a real bitch to do in practice. In SO R questions I see lots of questions that the asker thinks are about his field or his algorithm but, in reality, he's struggling with an R syntax or data structure problem. So expecting the user to understand what level of abstraction his question resides is not reasonable. So as King I would just move the question... assuming I could determine the appropriate level of abstraction, which I probably could not.

I'm concerned that expecting this to 'self organize' may not be reasonable and we may neuter Stack Overflow as a source of good R questions and answers.

| |
$\endgroup$
  • $\begingroup$ @JD: I agree with most of this, but I have to say: I'm glad that you're not "king of the internets". :) I can only begin to imagine what kind of warped place it would be! $\endgroup$ – Shane Jul 27 '10 at 16:33
  • $\begingroup$ It may be worth coming back to add that three years down the track, what happens is that 'pure R' questions on stats.SE are nearly always moved to SO, while ones with statistical content are kept. At present there are almost 32000 R questions on SO (almost 90% with accepted answers and more than a dozen with over 100 upvotes), and there are more than 4000 questions tagged R on stats.SE. While there's likely always going to be ongoing debate about where to draw the line, it generally seems to work very well (as someone with the bronze r-tag badge on both sites I see it from both ends) ... (ctd) $\endgroup$ – Glen_b Jun 20 '13 at 23:44
  • $\begingroup$ (ctd) ... the active users on stats.SE mark to close what they think should be moved to SO and the moderators are typically astonishingly fast at moving them. Even so, r is a very popular tag on stats.SE and any question usually has an answer within minutes, gets transferred to SO soon after where it attracts a bunch more good answers, often from people in the R core group. SO is not remotely being neutered, and is clearly being kept as the main resource for 'coding' questions. In short, through the efforts of a few dozen individuals, it works pretty much as you might hope. $\endgroup$ – Glen_b Jun 20 '13 at 23:46

You must log in to answer this question.

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