In my experience here, things are turned a bit upside down from other sites. Because making a good puzzle has turned out to be harder than solving it.
That is because--if your puzzle is going to be any good--you have to provide sample inputs and outputs. Which means most of the time you really should solve your puzzle before you post it. So you think one up, then solve it, then lay out the parameters...it's basically working as a CS professor for free. It is very unlike the other Q&A sites where the questioner doesn't do a whole lot of work relative to the answerers.
The maligned code-trolling tag, I think, is disliked because specifically it doesn't require that expected effort on the part of the asker.
It just seems way more fun to answer than to ask. Given the time to compose a creative question, you start getting disappointed with only a few answers, especially if people seem to be just solving your problem and not golfing it. Also by its very nature, there is no timeline for accepting answers. If you've invested a lot in defining a puzzle then why declare a winner if it closes your puzzle?
So there is an issue here about the fact that grafting puzzle/solution onto the SE model has a bit of a mismatch. I don't think it's correctible, which isn't a condemnation of the site in any way. It just means that it's always going to be kind of weird and policy needs to be laid out.
Thus I think it's still in beta until there's a larger amount of precedent set in policy space. But hopefully the ads will get runthe ads will get run to drive a bit more traffic and awareness. This one cracks me up:
