Niko Matsakis
Niko Matsakis
Sorry @ldm0 ! I'll try to give useful feedback ASAP
I'm reviewing this now.
Apologies for the very long delay.
Sorry @ldm0, got underwater again, but coming out of it :)
Thanks. I'll take a look at the tests over the next few days! I'm trying to allocate some time every day for reviewing and notifications...
@aturon huh, that's annoying!
@eddyb > Should we infer unspecified associated types in expressions, just like type parameters? Hmm, probably, yes. However, that alone wouldn't be good enough. We'd also need the "GC" strategy...
@eddyb I was just going on the example, where `T` was not needed; but yeah perhaps in real cases, `T` would be used.
The one form that is not applicable across _all_ type objects is `T(buffer, offset)`, though one can write `T.Array(buffer, offset, 1)` (or whatever) to create an array of length one.
The branch 2022-08-coherence includes some preliminary work here: It introduces an overlap check and checks that the various impls that are in scope do not have types that successfully unify....