What problem does the Community Wiki feature actually solve?
Pandering for rep: asking questions that will get what the community considers an unfair amount of interest (like "What is your favorite programming X?"). Community Wiki allows some of those to be tolerated, as long as the asker agrees (by making it CW) that he won't accept any rep from it. This also disincentivises those types of questions.
It has the potential to solve a number of problems:
- Opening up collaborative editing and composition to more users.
- Removing some of the disincentive for collaborating on a single, comprehensive answer vs. posting a new answer.
- Allowing open-ended questions to be asked, answered, and voted on without skewing the reputation system away from questions with concrete answers.
Whether or not any or all of these benefits are ever seen depends on the question and those contributing. The sofaq questions are some of the best examples of how this can work well...
I have noticed as a newb, that community wiki can be used as a club by experienced people on people that ask subjective like questions when they are starting out, especially when you are expecting forum like behaviour.
In stead of explaining why, yesterday I got barraged by a volley of "this should be community wiki" comments on a SU question until the question changed over by itself.
I was off line and came back to research why I was being beaten up. It was not a pleasant feeling.
- allows an answerer to state an unpopular opinion without fear of massive rep hit, and
- provides a functionality for asking questions where there isn't really a correct answer.
I think a question like Most wanted features for C# 4.0 is a good usage of community-wiki. I think it is a question about, relevant to and interesting to programmers. However, it certainly has no correct answer, nor does it really add to the knowledge base of SO. So, I think it belongs, but I don't think people should gain rep from it; which is what CW accomplishes.
Additionally, for the SO-team, I believe the automated features that turn questions CW save them a lot of headaches. Jon Skeet facts maybe belongs, maybe doesn't, but automatically after 30 answers (I believe) a question like this will become CW. Which results in generally discouraging this type of behavior. And keeps people from gaining an insane amount of rep from something that is not contributing to the knowledge base of SO.
Also, before the addition of meta, I believe CW was necessary for the few meta questions people just could not avoid and that were necessary for the system to grow.
The lack of free, concise, detailed and discussed properties about some general questions straight from
experts the community. It's actually a good feature, however, unfortunately, the site's rendering engine makes it hard to see it as a complete and well organized information source. You need to read through answers similarly to a regular question. Wikipedia, in other hand, is more like a single-paged in-place edited source.
CW allows fun/poll questions to exist without them giving away huge amounts of rep so I guess it "solves" that problem. Well, that's one way to look at it. Personally I think it causes that problem. A lot of these questions wouldn't exist otherwise.
In addition it causes a number of other problems:
- It may help avoid pandering for rep but does nothing to stop pandering for badges making some of them (eg Great Answer) pretty meaningless as they're hardly ever awarded for programming related answers;
- CW is used as a dumping ground for dubious questions. Often something which is nothing more than a discussion will get comments like "should be CW" but really the problem is that it's not a real question; and
- People use tagging something as CW as carte blanche for any kind of question. When its voted to close you'll sometimes get comments like "but I marked it CW".