> Yes, Spad compiler should properly track meaning of '%'. But
> in this case using F is clearer both for compiler and readers
> of the code.
Well, I do not know the internal datastructure for storing the
information, but it seems that when the compiler sees the line
if F has with qetaGrade: % -> ZZ then (*)
it should attach the property to % that it actually means F (and only to
the % at the type of qetaGrade).
OK, maybe I am too naive here but perhaps it could be a small change to
let the compiler not misinterpret the % when it comes to
n: PP := qetaGrade(t)::PP
where t : F.
Maybe I should add it to the github issues section.
Ralf