regression by commit 991f8347

[Qian Yun]

I found this regression when I was reviewing the reference book.

In chapter 1.16, or 1 page before chapter 2, there's:

f(0) == 1; g(0) == 1
f(n) == e/2*f(n-1) - x*g(n-1)
g(n) == -x*f(n-1) + d/3*g(n-1)
f(3)

Internal Error
The function * with signature (%, Integer) -> % is missing from
domain Polynomial(Fraction (Integer))



If we write "f(n) == f(n-1)*e/2 - x*g(n-1)",
"g(n) == -x*f(n-1) + g(n-1)*d/3", then there's no problem.

Git bisect (with the help of fricas0-repo) points to commit
991f8347 in 2022-Jun-26.

Is this a bug or shall we update this example?

- Qian

[Waldek Hebisch]

This is a big meta-bug, namely there are functions which are declared
but which are implemented nowhere. In particular, when R is
LinearlyExplicitOver(Integer), then FullyLinearlyExplicitOver(R)
implies RightModule(Integer) and we should implement corresponding
multiplication.

Good fix should implement multiplication in appriopriate place
(or some small number of critical places) so that it is inherited
in other places. However, this has potential for breaking other
things, so I really prefer to delay this after release.

--
Waldek Hebisch