The issue for me here is runtime compilation, which is essential if the
synthesizer is to be something usable for serious composition and
experimentation. Canned patches are never really enough,and a pure
interpreter (which is what SAOLC, the reference implementation, is) is too
slow. That's what Csound is, and one of the reasons we use it: it's a
runtime compiler, a true compiler, not an interpreter. (I'm willing to bet
that's also why the syntax is so simple-minded: that made it easier by far
to write the compiler.)

-----Original Message-----
From: Matt J. Ingalls 
To: Michael Gogins 
Cc: Richard Dobson ; csound@maths.ex.ac.uk

Date: Wednesday, February 24, 1999 1:08 PM
Subject: Re: if...


>On Wed, 24 Feb 1999, Michael Gogins wrote:
>> back to are (a) an efficient, public-domain version of the MPEG-4
Structured
>> Audio Orchestra Language, which is an updated, C syntax version of
Csound's
>
>  i totally agree with you.  just think if all the timedeveloping
>(fixing)Csound the past 4 or 5 years had been spent creating a new
>SAOL-like version....
>
>> something like Phil Burt's JSyn, but with runtime loading of unit
generators
>
> (Phil Burk)
>
>and i dont see any reason why a SAOL implementation couldnt be using
>Phil's C library (perhaps he is doing this already..)
>
>-matt
>