re: The Rise Of Functional Programming: F#/Scala/Haskell and the failing of Lisp

int19h — Tue, 16 Sep 2008 15:23:16 GMT

I'd say that another major Lisp failure was the complete and utter lack of static typing. Yes, there were type hints in CL that could be used to improve performance, but I don't really care about performance these days - I care more about correctness. Neither CL nor Scheme nor any other popular Lisp dialect had that, while most functional languages have traditionally enjoyed very expressive type systems with advanced inference throughout, and further development in FP seems to be type-related as well (effect typing in particular is potentially very useful in the context of implicit concurrency).

The only part that I find sad about CL demise is that the ideas of CLOS have never found popularity. I really miss the generalized concept of multimethods, and loathe ugly and hackish workarounds such as the visitor pattern. A pity... but even so, give me explicit static parametric typing over that any day.

Erlang

DanielWeinreb — Thu, 18 Sep 2008 17:16:38 GMT

Having put so much emphasis on lack of side effects and on concurrency, I was surprised that you did not mention Erlang. Those are its cornerstones, and the Erlang community claims great real-world successes.

I agree completely with you about the extraordinarily high quality of Lambda The Ultimate. I worry, though, that the people who participate in those discussions may be projecting their technical enthusiasm for functional languages into optimism that those languages will catch on in the marketplace, optimism which might go somewhat overboard.

It's historically sort of hard for a new language to catch on and get "traction". Although, lately, I must admit, we seem to be in the middle of a golden age of language experimentation and proliferation.

Don't get me wrong: I am very interested in these languages. I did a blog article about Scala, which seems potentially very interesting. On the other hand, a lot of people trying new languages are trying somewhat

more conventional ones such as Ruby and Groovy. It's hard to anticipate the future direction of technology in the software world; the older I get, the less likely I am to take strong stands about what's around the corner.

Your article above is valuable for, among other things, bringing peoples' attention to these languages and explaining their benefits. Getting that word out is surely the first step toward "traction".