> The alt_eigen solution recommended earlier in this list can't pass its own tests on 5.30.
The errors stem from a problem computing the nullspace of a zero matrix when ratmx is true:
(%i7) nullspace(matrix([0,0],[0,0]));
Maxima encountered a Lisp error: GREAT: internal error: unexpected MRAT argument
A workaround might be to apply ratdisrep to the output of triangularize in nullspace, , but I'm not sure this is the
correct fix. Maybe this bug manifested because great was changed to give an error for CRE expressions?
--Barton