I recently checked a copy of REDUCE out of the REDUCE project's
repository to find information on how its pattern matching system
worked. While I was digging through its source code, I was very
surprised to discover that at some point in time a Java-based Lisp was
created for REDUCE called JLisp. Here is a copy of the Java version of
REDUCE:
http://sage.ssu.portsmouth.oh.us/tkosan/misc/jlisp.jar
Just type java -jar jlisp.jar to launch JLisp. Here is some input to
get you started
2+2;
factorize(x^105-1);
solve(log(sin(x+3))^5 = 8,x);
Here is a link to the the REDUCE manual and website:
http://reduce-algebra.sourceforge.net/manual/manual-pdf.html
http://reduce-algebra.sourceforge.net/
Anyway, REDUCE appears to be a fast, high-quality CAS and the fact
that it has a Java version has me wondering about the possibility of
combining it with MathPiper.
Markus, I am especially interested in hearing your thoughts on the
Java version of JLisp. It may be a better match for the needs of
GeoGebraCAS than MathPiper currently is.
Ted