Jacob Pratt

Alright…there's 28 instances of an `OffsetDateTime` being output there. Needless to say that's a lot. However, I can't say that having an `OffsetDateTime` in a `Vec` is an unusual situation....

Is there an RFC for something along the lines of what @KrishnaSannasi said? This could absolutely allow the compiler to optimize space, and could halve the size of certain structs....

@KrishnaSannasi One of the key advantages of range bounded integers is the ability for smaller representation, which isn't possible without rustc integration.

Yes, most of the API would be possible to implement as a crate, though treating it as a primitive would not be. Ideally, I'd be able to do `let x:...

Just for anyone that may be thinking of doing the same thing — I intend to begin work on a basic implementation of ranged integers on nightly. With min_const_generics looking...

@newpavlov Neither `time` nor `once_cell` violated the MSRV contract. I have already explained this to you for `time`. Please keep the discussion in time-rs/time#484, where it has been taking place...

> consider MSRV bump a breaking change Even setting aside that bumping MSRV is largely accepted as a minor version bump, not major, there are enormous consequences to a change...

With regard to base 10, I've found that using the approach followed [here](https://commaok.xyz/post/lookup_tables/) is approximately 10% faster on `u32` using a crude benchmark. Presumably this holds for larger integers, but...

Of course cache is huge for something like that. Just throwing it out there; it's plenty fast either way.

Agreed 100%. If a pseudoclass isn't feasible (and I agree it isn't for the reason Tab stated), there should still be some way of knowing.