[baguasquirrel]

I've actually been trying to prove people like this guy wrong for years. My takeaway is that Haskell is not the problem. The real problem is that solving people's problems is hard. Designing and iterating on a product or service is hard. Finding and involving the ambitious folks with halfway well-baked ideas is hard.

Now I'm going to get a lot of flack here for saying this but I simply just don't feel that most of the stuff we make today is of particularly high craftsmanship. We need one of two things to happen.

The first is that the demand for thoroughly well-crafted software has to increase. But that much is out of our hands as it is conditional on the public at large. This will happen eventually. New tech starts life as a toy. At some point, people will want those toys to be reliable.

Going the other way, we could make it more rewarding for people to build scalable, maintainable systems in Haskell. But we are already well on the way to doing this. For example, Snap was much easier to use than Happstack (IMHO), and Happstack itself has already much improved over the years. The libraries for Couch and Mongo are as good as any other. Tools aside then, Haskell's main impediment is that can be a bit intimidating. But a friendly and openminded developer community can go a long way. We just have to keep this in mind, and keep fostering projects.

[szany]

cf. http://conal.net/blog/posts/can-functional-programming-be-li...