I am certain that much of the speed problem with saolc is the parser. You
should be aware that Ross Bencina is writing sarun, a SAOL decoder that will
be faster than saolc and not as fast as sfront. There are also some
commercial companies creating SAOL decoders, and I think they include Create
Labs (Sound Blaster) and the people who make SSeyo Koan Pro.
You can schedule notes from inside an instrument in SAOL using the "instr"
statement. In other words, you can write self-contained algorithmic
composition programs in SAOL.
-----Original Message-----
From: Koen Dejonghe [mailto:koen_dejonghe@hotmail.com]
Sent: Tuesday, November 09, 1999 8:19 PM
To: saol-dev@media.mit.edu
Subject: saolc compiler
Hi all,
Is anyone working on the improvement of the saolc compiler as far as speed
is concerned ? I've been reading on compilers and playing with lex (flex)
and yacc (bison) for the last couple of weeks now and I can almost
understand it. Which in turn gives me enough confidence to allow this urge
to come up that tells me dive into the saolc code. And make it a faster
compiler (or interpreter, as you like.) But maybe the speed has nothing to
do with the way the input is parsed, and the problem is elsewhere ... ?
Another reason why I'd like to go into it is that I still don't understand
why the score needs to be written in a separate file. Is there one good
reason to do so ? How can I generate endless variations in the score using
recursive and/or fractal algorithms ? Files are per definition limited
(unless I use a named pipe, but that smells like a kludge because then I
would need another program written in again another language that runs at
the same time dumping its output to the pipe etc. etc.) So in brief I can't
write the output of my never ending routine to a file. If on the other hand
I would be able to generate the score inside the saolc program, then there
would be no need for such a file.
Best regards to you all,
Koen Dejonghe.
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
This archive was generated by hypermail 2b29 : Wed May 10 2000 - 12:15:45 EDT