scottmcm

> Also since this is not about stable layout guarantees, I wouldn't bother with the `repr`. I agree with Ralf on this. I think you probably want to look just...

> We maintain some trivial validity invariants (such as alignment and address space limits) that a user cannot define I, at least, fully expect us to *eventually* have *some* way...

> Just to follow up on some of the discussion, it wasn't immediately clear to me that types _similar_ to `Box`, like `Vec` and `Arc`, genuinely don't have these semantics...

Thinking more and seeing updates, I think I'm fine to check a box here, subject to a note that I think this would be a really good thing to split...

This seems entirely reasonable to me -- it's basically like every trait has the supertrait ε and you can go from `dyn Foo + ε` to just `dyn ε`. @rfcbot...

I haven't dug into the example specifically, but whenever GVN re-uses something it removes all the storage markers for that value and changes all the uses to `copy` rather than...

I've wanted something like this for [ages](https://rust-lang.zulipchat.com/#narrow/stream/213817-t-lang/topic/Revisiting.20default.20struct.20field.20values/near/214260348). I took a fresh read through here after the updates, and other than what looks like a [vestigial section](https://github.com/rust-lang/rfcs/pull/3681#discussion_r1755045300), it looks good to...

@tmandry I [agree](https://github.com/rust-lang/rfcs/pull/3681#discussion_r1728060796) I'd like to move to something like option 3 (in the future, not here), though I've been spelling it differently in terms of how to opt-in, since...

@RalfJung I'd guess that's referring to https://github.com/rust-lang/rfcs/pull/3681#discussion_r1728060796

+1 to linting -- I think we can also do things like "hey, you put defaults for all the fields, so you should probably `derive(Default)` too" (at least in clippy/ra,...