[GiNaC-devel] About a (poor) parity in expressions.

[Alexei Sheplyakov]

On Sun, Oct 17, 2010 at 11:40 PM, Sergei Steshenko <sergstesh at yahoo.com> wrote:

> To me it looks like two missing rewriting rules.

By design GiNaC does not apply any rewriting rules on its own. GiNaC is
designed to process large expressions (~10^7 -- 10^9 terms), therefore
implicit evaluation does only "simple" (that is, O(N^2)) transformations,
everything else (normal(), evalm(), etc) is done only upon the user request.