[m-dev.] discussion about the implementation of compact type representations
zoltan.somogyi at runbox.com
Tue Oct 31 03:28:06 AEDT 2017
On Tue, 31 Oct 2017 03:24:48 +1100, Mark Brown <mark at mercurylang.org> wrote:
> >> - If the argument of the function symbol has an existential type,
> >> that should not be an error, but should make the type NOT a notag type.
> How about if it's an existentially typed dummy type, such as
> :- type foo ---> some [T] foo(unit(T)).
> If there's just a type_info in there, that could also be considered
> notag. Is it worth including this case?
Have you ever seen such a type in real life? I haven't, and I would be
surprised if you have. I think that makes it pretty clear that optimizing
the representation of such types is not a priority.
More information about the developers