Niko Matsakis

> Nit: This doc takes a bit too long, as written, to establish firmly that the ".." in the syntax `(..): ` is not a meta-notation but rather a concrete...

@rfcbot concern wait-till-july-9 I'm all in on this, but I think this is a big enough feature that I want to have time for some more discussion before we actually...

@rfcbot reviewed ...that said, I know I want to accept this feature in some form. I'm going to check my box and I'll register concerns if I see anything that...

@rfcbot resolve wait-till-july-9 I think there's been a fair bit of feedback, so I don't feel the need to "pump the brakes" anymore. I'm still reading and digesting it all...

I've not read the updated RFC, but I wanted to share a few general thoughts re: semver, opt-in, etc... First, in terms of semver, there are two mitigations 1. As...

> So, like a `PhantomNotFreeze` type? That would be the obvious way to go that matches precedent. I hope we can shift to negative impls but that's a separate effort.

I've had some use cases for this, I *think* in salsa, I have to think about whether shallow- or deep- would've been useful. I suspect shallow is the thing I...

Reasonable optin for "disable all auto traits", I suppose. On Wed, May 15, 2024, at 5:51 PM, Clar Fon wrote: > > > Doesn't `UnsafeCell` have the same effect as...

I have wanted this feature many, many times over the years. It would also allow the methods defined (rather awkwardly) as inherent methods on dyn Any to be moved to...

That said, a design meeting plus some time for folks to comment both seem good. One possible extension: final associated types, which I have also wanted. I have to go...