[linux-audio-dev] OSC vs MIDI

[Albert Graef]

Erik de Castro Lopo wrote:
> Has anyone else looked at O'Caml? :
> 
>     http://www.ocaml.org/
> 
> I've been coding in it for about 6 weeks and I'm REALLY enjoying
> it. Its the most fun I've had coding in ages.

Yes, those modern functional languages really bring back the joy to 
hacking. :) The trouble with ML and Haskell is, though, while they 
already have some facilities for graphics programming, at this time they 
mostly lack other multimedia libraries. (And I'm wondering how suitable 
Haskell would be for realtime stuff, given its lazy evaluation strategy. 
ML certainly is, in fact it's been reported that in some cases compiled 
ML is even faster than C ...)

But the world doesn't end there. E.g., there's Faust for audio 
programming (developed at Grame, must admit that I haven't tried it yet, 
but I certainly will), and (shameless plug) Q (http://q-lang.sf.net), 
which already has interfaces to PortAudio, libsndfile (of course ;-), 
MIDI, OSC, SuperCollider and whatnot. Q takes the route to represent 
MIDI and OSC messages as high-level (algebraic) data types, so most 
transformations (such as mapping MIDI input to SuperCollider control 
messages) are almost trivial to implement. Q is also embeddable in C, so 
you can use it inside a C/C++ program if you need that.

> If anyone has the FFTW sources on their machines, they already
> have some O'Caml source :-). The FFTW authors wrote a program 
> in O'Caml to generate the C code for the codelets.

Yes, that's a nice twist. :) Apparently plugging together the different 
FFTW algorithms becomes so complicated that they use OCaml to do 
algebraic simplifications on the resulting graphs.

Albert

-- 
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email:  Dr.Graef at t-online.de, ag at muwiinfa.geschichte.uni-mainz.de
WWW:    http://www.musikwissenschaft.uni-mainz.de/~ag