>   I was wondering whether it would be possible to alleviate
>   some of the shortcomings of the Csound language by
>   building instruments in a more friendly and ``higher level''
>   programming language, then using a ``translator'' of some sort
>   to change the given orchestra into csound format.  Thus, the
>   .orc file would be a ``low-level'' intermediate stage, much like
>   the object files created by C compilers.

This is very important.  Using the C preprocessor
on orchestras can lend great power to the language.
This was an integral part of CMUSIC as I remember.
One can #include definitions of various scales with
various tunings, draw from a library of functions
and instruments, and implement #define macros which
reduce the variable naming problems.

Under Unix this can be mixed with 'sed', 'awk' and 'make'
to reduce the entire process to one command.  This
is why I hope that Csound will remain as a powerful
synth engine with a simple text interface.  The question
of interface needs to be addressed by each individual
as she develops a working method.

Cscore is my answer to the score generating side.
Try generating a random burst of notes with an
exponentially decreasing density within a given
moving range of pitch and tone color by using a
MIDI sequencer.  Even if it could be done, could
it be created again and again with simple commands,
using different exponents for decay, and different
pitch ranges etc...?  

Toby

	-There otta be a law-