I was not careful enough when I prepared the report, because I actually
wanted to show this one.
After applying the patch, the errors disappear.
Obviously, the compiler delays instantiation if the type arguments are
not constant and then forgets to instantiate the default implementation
of the ca
QETARKA -->
Warnings:
[1] findRelation: not known that (UnivariateLaurentSeriesCategory C) is of
mode (CATEGORY domain (SIGNATURE coerce ($ (Variable (QUOTE q (SIGNATURE
differentiate ($ $ (Variable (QUOTE q (IF (has CX (Algebra (Fraction
(Integer (SIGNATURE integrate ($ $ (Var
I struggle with a problem that I believe is due to a compiler bug.
With the attached program and
)co foo.spad
findIdentity()$QEtaRamanujanKolberg(Fraction Integer)
I get
(1) -> findIdentity()$QEtaRamanujanKolberg(Fraction Integer)
[:> , foo, 1]
Function: laurent : (Integer, Stream(Fraction(
Ralf Hemmecke wrote:
>
> I would have expected that the following program is rejected by the
> compiler, since instead of
>
> dbgprintx("all", all::OutputForm)
>
> I have only written
>
> dbgprintx("all", all:OutputForm)
>
> Needless to say that the program does not implement c
I would have expected that the following program is rejected by the
compiler, since instead of
dbgprintx("all", all::OutputForm)
I have only written
dbgprintx("all", all:OutputForm)
Needless to say that the program does not implement coerce:%->OuputForm
and should be rejected an
why doesn't this compile?
)abb domain COMP Composition
Composition(): Join(Monoid, RetractableTo List Integer) with
nearConcatenation: (%, %) -> %
== add
Rep := List PositiveInteger
per(a: Rep): % == a pretend %
rep(a: %): Rep == a pretend Rep