I've just been considering that I found would find it very useful to be able to view certain details about the number and types of users online at any time. In particular, the following statistics I would find quite helpful for the given reasons:

  1. Number of users online - generally useful to determine how people could potentially answer your question, or whether you might find a question you can ask.
  2. Breakdown of number of users online by a) reputation, b) favourite tags, c) number of questions asked, d) number of answers made. All these could possibly be used in a user's judgement of whether it's worth them idling around StackOverflow at the current time.
  3. Current/recent traffic rate (page hits per second?) - gives you similar info to point 1, except it largely ignores "inactive" users that are online.
  4. List of moderators online - perhaps most useful for other moderators so that they know when someone is watching the questions.
  5. List of 'famous' users online - mainly for curiosity/fun.

What are people's thoughts on these ideas? Has it been considered previously? Certainly I see no technological obstacle to implementing such a feature.

4 Answers 4


This idea has come up before, although with not as much specificity.

I'll be paraphrasing from my earlier answer.

Basically: It does not matter how many users are online. Moreover, we do not want users to care.

It basically amounts to a way to 'game' the system. You only ask questions when the average rep is high enough, or when there are enough users online. The trouble with this pattern is that it is self-enforcing, it will lead to higher and higher server loads at certain times, and many fewer users at the other times.

The main reason why information like that would be useful is if you really need your question answered. However, there have been requests that were related to making 'urgent' or 'time-sensitive' questions, and they all get rejected.

SO is not about farming out work. SO should be a resource for you to bring your toughest problems and it should be an extra tool in addition to your continued attempts to solve the problem. Because that is the case, it should not matter when you put your question up. If it is not answered in 3 (?) days, you can add a bounty to it, but before then, you can only wait. Or continually add edits as you gain information.

If it is about getting people to answer your question, there are ways to do that without this information. This information would change the usage habits and it would almost definitely be for the worse.

  • This is at least a reasonably well-thought out response, though I'm not inclined to agree with all the points. Certainly, there is an element of self-reinforcement - yet there is both positive and negative. When the site gets too active at certain times, questions will be swamped, and users will ask questionsat less active times; the same applies conversely. Now, whether this simply readjusts the equilibrium and thus does no good, that is quite possible. I still believe the matter could do with more consideration. Finally, I'm not quite sure what you're saying about "famring out" work...
    – Noldorin
    Commented Sep 22, 2009 at 1:15
  • I agree with your comments there, though I was not intending to imply that SO was for farming out work, in any way. I'm curious: the down-vote was yours, or not? It's intriguing that the other question was up-voted so highly yet mine was down-voted, yet I should accept the arbitrary nature of voting.
    – Noldorin
    Commented Sep 22, 2009 at 1:16
  • I did not downvote, I would have commented. That's a personal choice by the way, not something necessary. The farming out of work is basically a comment on the fact that people who need something urgently should go elsewhere. SO is not for urgent work. Any feature (such as yours, I was implying) that is designed to encourage strategic timing of questioning, should be discouraged.
    – devinb
    Commented Sep 22, 2009 at 12:07
  • @devinb: Fair enough. I'm not sure about the exact nature of the effects, as I said - an equilibrium would certainly be re-established - where, we cannot be sure. It has been an observation that the quick response time of StackOverflow makes it akin to IRC in many ways, which was previously the source for "urgent" help. Wouldn't seem to be going against the trend, in my view; nonetheless, I see your point.
    – Noldorin
    Commented Sep 22, 2009 at 16:51

I'm going to quote Cletus as I've done before.

Repeat after me: SO is not a social networking site.

  • 1
    SO is not a social networking site. Commented Sep 21, 2009 at 19:54
  • This neither answers my question nor contains any reason behind it. Forums display online users - they are hardly for social networking.
    – Noldorin
    Commented Sep 22, 2009 at 1:06
  • 1
    Forums are more about social networking than SO. SO is about getting your questions answered. Period. Commented Sep 22, 2009 at 10:38

To be honest, I'm kind of curious as to what the traffic patterns for Stack Overflow are like; however, I also think that seeing that information real-time is a bad idea. There have been a couple blog entries (e.g. "Us Versus Hyphen" and "Where In The World Do Stack Overflow Users Come From?") that address the topic of traffic patterns and I think that ultimately, that would be the best place for such information.

  • Yeah, you make a good point. What do you think about having the statistics updated weekly and published on StackOverflow in a user-friendly way (charts, small tables, etc.). I would at least be curious to see that, and it could at least yield some interesting insights, no?
    – Noldorin
    Commented Sep 22, 2009 at 16:53

It would certainly be possible to code, but it might incur a performance hit.

Running aggregate statistics like that is not a cheap task.

  • Solution: SO@home
    – Mark C
    Commented Apr 12, 2010 at 21:53

You must log in to answer this question.

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