There are lots of places where the Poisson
representation isn't handled properly, and a few where
Rat isn't handled properly.
I'm not sure if anyone cares about Poisson
representation, to tell the truth, but it would be nice if
specrepcheck were used more consistently.... Of
course, it would be even better if representation-specific
routines were called, but we can start with
diff(rat(x),rat(x)) => 1 OK
diff(x,rat(x)) => 0 BAD
diff(intopois(U),U) => fatal error
taylor(intopois(sin(U),U,0,3) => fatal error
ratsimp(intopois(sin(U))) => internal error
Log in to post a comment.