R: [Haskell-cafe] Hacking Haskell in Nightclubs?

[Jerzy Karczmarczuk]

Santoemma Enrico wrote:
> Doing music is a hard low level problem, a concurrency problem and is also an interesting problem for intelligent and behavioural computation.
> 
> If it's true that Haskell can do its best on the second and third aspect, the undertaking seems to wrap with foreing interface one of the many good C libraries for low level audio/midi management.
> 
> Btw, about the third aspect I liked your approach to model animations through behaviour in your book School of expressions.
> 
> So now the questions seem to be: how difficult is today to wrap a C library? Which is the right tool to do it? C->Haskell?

About low-level...
Using C libraries, or using Haskell to produce CSound scripts, or SAOL, etc. is
always possible, but not too ambitious, for a genuine Functional Sectarian.

If you permit a shameless plug, I produced some purely functional programs which
generated/simulated basic musical instruments using the waveguide models. It has
been don not in Haskell but in Clean, since Clean arrays and unboxed lists were
much more efficient than the equivalent structures in Haskell *at the time when
I worked on that*.

If you wish to have a look on the paper:

http://users.info.unicaen.fr/~karczma/arpap/cleasyn.pdf

Of course I can send the Clean stuff to the <<volenti>> as well, but I am aware
that this is a 5th-column attitude in a Haskell list... Still, you can try to
translate the models into Haskell, which should be conceptually trivial.

The best.

Jerzy Karczmarczuk