Crystal Durham

cc @dragostis since you've been talking about an IR framework to use with pest

Very unscientific test: ```sh cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ git.exe checkout master Switched to branch 'master' Your branch is up to date with 'upstream/master'. cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ cargo.exe clean cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ time cargo.exe build --release Compiling proc-macro2...

Whoops, just realized I missed a `cargo clean` there, so here it is with a clean: ``` cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ cargo.exe clean cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ time cargo.exe build --release Compiling unicode-xid v0.2.0 Compiling proc-macro2...

Second test (with `procmacro2_semver_exempt`) shows an increase, so it's probably all within noise without controlling better. ``` cad97@DESKTOP-HIBC3H1:/mnt/d/usr/Documents/Code/Rust/proc-macro2$ git checkout master Switched to branch 'master' Your branch is up to...

> This combined bound is expected to be used more often than just `?Sized` outside of FFI, right? An owning pointer use will generally want `?Sized + MetaSized`, and a...

I got curious and ended up writing [a partial implementation of `?MetaSized`](https://github.com/rust-lang/rust/compare/master...CAD97:rust:metasized)[^impl] for the current edition (i.e. `?Sized` still implies `MetaSized`, but ignoring anything beyond the trait itself). My goal...

It's easiest to explain by example: ```rust // Assuming some trait associated type, e.g. trait Deref { type Target: ?MetaSized; // ... } // when declaring the function fn f()...

(off-topic meta note: wow, you can see that this review took me all day, because the timestamp of the root comment of the review is "now" but the timestamps of...

This specific reply wanders pretty far off topic, but I don't think the resolve overhead for `&dyn Allocator` (with `trait Allocator: Store + ...`) is _strictly_ necessary, and it could...

> it's not possible to have a `const Store` implementation for any `Allocator`, but it is possible to have a `const StoreDangling`, I seem to recall intent to remove `~const`...