[OasisG]

Depends on what you mean by "stealth".

If stealth means spending over a year and a boatload of cash developing a product without ever getting any feedback from any potential customers, then yeah, that's likely a dead end.

If stealth, however, means meeting quietly with a small (but representative) group of potential customers while you work the kinks out of your project before launch, then I think that is a smart strategy. There's nothing wrong with giving yourself as much of a head start as possible.

As a general rule, I don't talk about my ideas/project with other developers. I especially don't discuss it with industry people who may be my future competitors. What I might discuss are things like the overarching problem I'm attempting to solve. I may even describe some of the very basic mechanisms, or ask questions that hint at the philosophy behind my approach.

I know HN can be fiercely against the notion that ideas matter at all, but most successful businesses have their "secret sauce" and it's not smart to just give it away before you're primed to leverage it to your benefit. I mean, front page today is an article about Coke's secret recipe; sometimes, the what really does matter as much as the how.