On: Why Lisp is a Big Hack (And Haskell is Doomed to S...

[dexen]

This is the first time I feel PG's `The Hundred-Year Language' may have answer different than Lisp. On the other hand, `a witty blogpost proves nothing', to misquote somebody.

[thebooktocome → dexen]

"A witty quote proves nothing" -- Voltaire

[jamesbritt → thebooktocome]

I wonder if Voltaire expected people to be quoting this ...

[pwpwp → dexen]

The great thing is: if and when Haskell ever evolves as far as described in the post, we'll have Lisp again - inside "Haskell". ;)

[Locke1689 → pwpwp]

Looking at it that way is ignoring all the benefits that Haskell can offer over Lisp. If all you can think in is Lisp then you're doomed to write Lisp in any language, to turn a phrase.

[andreyf → pwpwp]

No. Haskell has type safety which was never intended to be included in Lisp. On the other hand, Lisp has the s-expression syntax which allows for usable compile-time macros. Neither language is a subset of the other.

How useful type checking and macros are in creating real-world software is not a question I've seen answered well. Smalltalk-style languages (Python, Ruby) don't really practice either.

[pmarin → andreyf]

"Haskell has type safety which was never intended to be included in Lisp."

Racket has a dialect with static types. http://docs.racket-lang.org/ts-guide/

[Mathnerd314 → andreyf]

Haskell has Template Haskell, which is (almost) as usable as Lisp macros.

[hvs → dexen]

I think that without macros, any language claiming to "replace lisp" is only doing so superficially.

EDIT: Just to clarify, I don't think the Haskell designers are claiming that it "replaces lisp," just that this blog entry does.

[jsmcgd → hvs]

I used to think so too until I learnt about Haskell's very lazy evaluation. It's worth reading up.