# What do we need more of during this private beta?

It's 6 days into the private beta, we only have 66 questions, and Robert Cartaino has very discreetly announced:

We decided to extend this private beta another 7 days. The question count is a bit on the low side to launch and that gives you a bit more time to flesh out the scope and let the community self-moderation take a bit more hold.

I don't think we actually have any major moderation or scope issue, but I can't deny that traffic is slow, both in terms of questions and in terms of answers. We have quality but not quantity.

What do we most urgently need?

• More elementary questions?
• More advanced questions?
• More varied questions?
• More users to provide questions?
• More answers? (25% unanswered questions worries me.)
• Better answers?
• More users to provide answers?
• Something else?

I'd especially like to get some input from people who haven't been participating very actively, if at all. Is the site not what you expected? Are you afraid of not fitting (in which case, don't be, I think we all agree that this site should target all levels, whether or not we agree on the best way to reach this goal)? Is the issue the level of questions, or the topics?

• Hello people. You are very happy about the born of this community, but we have an inconvenient truth for you. Your community is not behaving well and in reality it is a suffering newborn with real chance of dying in his first days of life. You will need to work better if you want that this baby survives. Mar 12 '12 at 20:39
• @Victor Your reasoning being the same as in your main user's description (which contains blatantly wrong statements)?
– Raphael Mod
Mar 19 '12 at 7:22
• @Raphael: Take some humour sense. :) Mar 19 '12 at 18:22

## 2 Answers

More elementary questions?

I asked a question here on meta about stocking the site with frequently-asked questions on introductory material, to which we could provide very strong answers before the public beta. The answer was that we should not ask elementary questions just to stock the beta, so I have not been doing that, although I think there could be a lot of value in it.

More advanced questions?

I feel like we're getting plenty of "advanced" questions, no more are required. We are going to want to attract lots of people, lots of good answers, lots of votes, and asking purposefully difficult questions is not going to help that. It doesn't help that a lot of questions in CS have as their best answer "nobody knows".

More varied questions?

I have been trying this, and the questions seem to get a good number of votes and in some cases attract some discussion... so this might be a good way to go. Surely more variety will attract a wider user base, get more people involved, and (very importantly) help to solidify the scope of the site. I say yes to this.

More users to provide questions?

I definitely feel like this one is important. It seems obvious to me that avid users, who want to see this site turn into anything, should be asking at least as many questions as they provide answers. More good questions will attract more good users and help solidify the site's scope; good answers might keep experts away (why compete for rep when a few people are scrambling to get all the votes on a few questions) and I doubt it would attract a lot of novices (who are more likely to be drive-by posters, if CS.SE becomes anything like SO). I say yes to this.

More answers?

Asking more questions will lead to more answers as well, without (necessarily) increasing the fraction of bad answers, which would probably increase if we were to just give more (arbitrary) answers for each question. Not a fan.

Better answers?

Better questions will naturally lead to better answers. We need to focus on asking a bunch of good questions in a variety of areas, and answers will increase in quantity and (possibly) quality as well.

More users to provide answers?

See above.

Something else?

It could be useful for avid users to ask easy questions with no intention to answer each other's questions. This might attract users who aren't as interested (currently) by making it possible to get more points, since they won't have to compete with avid (and, one would assume, generally somewhat well-qualified) users. This might be counter to the whole Q&A thing that SE does, but it might be a (temporary) sacrifice worth making. An implementation that might work is to just put a hold on users above a certain rep (200? 500? 1000?) from answering any question which hasn't been around for two days.

• Maybe I was wrong to discourage seeding very easy questions. I was hoping that they would come naturally, but there haven't been many (several from Gigili, and a few others). So maybe “we” (CS graduates) do need to seed more undergraduate-level questions after all. Mar 12 '12 at 21:02
• @Gilles I think they will come astonishingly easily once the site establishes itself. Another benefit of seeding now is that we can reduce noise later; instead of providing two dozen general introductions to using the pumping lemma (some of which might be wrong!) we can just close as a duplicate. That way, we will get better, more specific questions when appropriate (users will narrow their questions) or, in any event, less redundancy. Mar 12 '12 at 21:06
• I like your last paragraph; we hyper-active people should take care to let others have a shot. Also, I think pumping lemma is a great thing to discuss!
– Raphael Mod
Mar 12 '12 at 23:43
• @Gilles We are wiser now, but your comment (towards yourself) is not fair: I think having a nice set of "non-trivial" is as important as having some undergraduate-level questions. And since it is clear (at least to me) that once the beta becomes public we will be swamped with the latter, it was a right call at that time.. Anyway, It was definitely justified to delay the public beta in several days. The burst of question people posted today is exactly what I felt was missing. Mar 13 '12 at 4:09

I wanted to offer some thoughts as a user who committed but has not posted to the site yet.

More varied questions?

When I originally committed to the proposal, it had the description in this Area 51 Discuss question:

"for computer Science practitioners, researchers, and CS students interested in topics like theoretical computer science, artificial intelligence, programming language design, digital logic, or any other area of CS not directly related to programming."

I understand why it was changed, but it seems as though that spirit of diversification has been dropped. I'm not a computer scientist, I've never even played one on TV, but I enjoy the subject, and would like to learn more. However, I see what seems to be a lot of highly abstract algorithm-related questions, and to my untrained eye, this seems to bear a lot more resemblance to CSTheory than not. I understand enough about the politics to know that many of them aren't "research level" and, therefore, are off topic there.

Perhaps that somewhat more recent incarnation of the description which included more applied questions (that I latched onto) was falling away from the original aim of the site (which was perhaps a more undergraduate friendly collective for theory -- CS is a science after all, and I have no problem with that!), and it has been brought back to that point. If that's the case, then I don't think this is the site for me. I will probably muster up a few questions (I'm following the Coursera algorithms class), but I'm also happy to stand on the sidelines and cheer, vote, edit, etc. to get this moving, as I know that it is exactly what others want.

• I don't follow your answer. The description has only been tightened, the site is still for CS “students, researchers and practitioners”. Is “a more undergraduate friendly collective for theory” what you want? (In which case the site does have questions in this vein.) Or do you want more applied questions? (We do have a few, but clearly there's a strong theory bias in the people who participate in the beta.) I feel you want something different, but I don't understand what. Mar 17 '12 at 1:05
• @Gilles Please see my edit, it's more that I thought there would be more applied questions. Sorry for the ambiguity there. Mar 17 '12 at 1:12
• You can't expect me to get my ideas out and have it make sense in one fell swoop. Mar 17 '12 at 1:13
• Thanks for the clarification. (And yes, I know the feeling.) I think I'm the only frequent poster so far who's not a CS student or professor… We'll have more applied questions if people ask them! (Hint, hint.) I, for one, know nothing about graphics or NLP or many other applied topics. Ok, and admittedly answers tend to verge towards theory; that is something we need to watch. Mar 17 '12 at 1:26
• I completely agree with the sentiment of this answer. we need many more questions that are NOT related to theory: where are the interesting DB/OS/graphics/data mining/AI questions ? To be honest, I've been trying to think of nice level-appropriate questions to answer in these areas and so far have failed :( Mar 17 '12 at 5:40
• I think non-theory questions are more than welcome here. It is just an artifact of the a number current experts on the site having come from sites like cstheory and bring the theory mindset with them. Unfortunately I have only asked and answered very theory-like questions myself, but I REALLY want this site to go towards the science part of computer science and I think everyone should be encouraged to ask more science-y questions instead of the typical math-y questions on cstheory. Mar 17 '12 at 6:38
• @ArtemKaznatcheev (and @Suresh) thanks for the feedback. What would you consider to be a 'science-y' topic? Mar 17 '12 at 6:41
• Of the fields I have come in contact with: machine learning, AI, and Bioinformatics tend to be more science-y. There is definitely a community of computer scientists that practice science as opposed to math, unfortunately I am not super familiar with them or their methods :(. Mar 17 '12 at 6:43
• Shameless self plug, but I feel these questions (originally asked on ML.SE but that got disbanded and merged into stats) are more about the science-y approach: 1, 2, 3. Unfortunately I have not had time to reflect on the answers and think of follow up questions. However, I would like to see questions like this on cs.SE. Mar 17 '12 at 6:49
• @Artem Yeah, I'd enjoy seeing some ML here for sure, the ML beta was a fiasco and everything ended up on Cross Validated. It's tricky to separate what might be good on here that's not necessarily good on SO. Obviously, more theory oriented questions would thrive here, but something that's too theoretical may not be an "actual problem." Mar 17 '12 at 6:51
• @Artem I'll have to look at those in detail. The one on neurogenesis is particularly interesting and would warrant more science. Mar 17 '12 at 6:53
• @jonsca The question on neurogenesis was meant as an engineer accompaniment to my similar question of cogsci.SE. I decided to test the waters here with a vague neural networks question, lets see if it generates any interest. Mar 17 '12 at 7:05
• @Gilles Before you think that I complained and walked, I'm still around, but I've just been on the busy side. Hehe. I have a couple of tags hanging open to get some background to try and ask a question on agent-based computing as that's something that I haven't seen here and I'd like to know more about. Apr 1 '12 at 0:29