Haskell and its community

I belong to the Haskell community and have done so pretty much since I leaned the language back in 1995 (although I learned a dialect called Gofer back then). I very much like Haskell and one key thing I like about it is that it encourages high level programming and very abtract programming. It is very easy and fun to write a very general piece of software. Likewise I like the Haskell community. There are lots of really brainy people around and pretty much all of them are attracted to Haskell for the same reason; it elegance.

But these things that I like about Haskell and its community are also the biggest drawback. In most programming language programmers take pride in writing complete programs for a specific purpose. You can easily find many programs whose name have the prefix py, indicating a program written in Python. The Perl community has produced loads of useful programs. Not to mention C. What programs have the Haskell community produced? Virtually none. Well, there is GHC, but I don't count that. Then there are small programs such as darcs and lambdabot. But are there any more examples?

The problem is that Haskell programmers take too much pride in writing elegant and general libraries. It has become a kind of game to tweak the type system to get the library extremely generic and compact, providing both very high-level combinators and type safety. This game is fun and intellectually stimulating. But libraries of this sort will not give Haskell any fame or widespread use. What we need is to start writing real programs in Haskell. Programs that people use. The libraries are there. Sure, one can never get enough of libraries but we certainly have a critical mass of them. We need to start hacking for real!

No comments: