I don't think I would ever depend on something like that for a library crate. There's enough syn+proc_macro2 pollution in the ecosystem already.

It does seem like maybe a crate with convenience macros so you can get a `&'static [Foo; N/size(Foo)]` rather than `&'static [u8; N]` and maybe even a delicious compile time "Hey jerk, that's not a valid Foo, you screwed up" if appropriate from your pre-baked data files, would be nice regardless of having more constant eval.