ISO Good-enough Idea

Joel Spolsky's latest article is largely a rehash of one of his earliest articles, whereby he expounds on his belief (which I share) that there's no substitute for top-notch programmers. The thing that got me thinking was his "shaky claim that most people are wrong in thinking you need an idea to make a successful software company," and this chart:

Best Working Conditions => Best Programmers => Best Software => Profit!

The problem I have with this is that you still need a pretty good idea to start with, or you're never going to get the best programmers in the first place. Great programmers aren't in it for the money; they want to work on stuff that interests them. Cool stuff. Sexy stuff. Having a great working environment with lots of other great programmers around is a big draw too, but how are you going to achieve that without the great idea in the first place?

This hits a chord with me because I know a lot of great programmers with whom I would love to start a company, but I've been waiting for the perfect idea to come along and it hasn't. But perhaps the trick is to just come up with a good-enough idea -- good enough to get one or two other great programmers to dive into it with you -- and just take it from there.

So...who's got a good-enough idea?