Waldemar Biernacki wrote:
I have to admit that I think that as shorter source code as better. Therefore the question to those who know Perl: is the source code of a client-server application shorter in Haskel to compare to Perl one? How much shorter (more or less)?
My only hard data point here is this: I once converted (just for fun) a short (two screenful) Perl script to Haskell and found that both are about the same length (in LOC). (I find the Haskell version is more readable, but that may just be a personal preference; Perl's syntax is a bit too noisy for my taste.) Anyway, don't let yourself be misled by small trivial examples. For some idioms (e.g. line oriented I/O) Perl has special syntactic sugar that leads to very short code. This makes it quite useful for one-liners that you can hack directly into the command line. However, as programs get larger and more complex, other factors dominate the equation. I'd say the most important aspect is how easy it is to factor common code into functions/subroutines. I think it is very hard to beat Haskell here because of it's excellent support for higher order functions. Yes, you can do higher order functions in Perl (I often do it, my Perl programming is heavily influenced by Haskell ;) but without a type system it can become quite a challenge. Cheers Ben