UnevaluatedConsts tend to be extremely conceptually similar to type aliases, and it's useful to think of things like const items as behaving like type aliases. We have AliasTy/AliasTyKind, AliasTerm/AliasTermKind, and currently, UnevaluatedConst/UnevaluatedConstKind. The last should be renamed to AliasConst/AliasConstKind.
My guess would be that not all uses of the word "Unevaluated" should be blindly replaced to "Alias", though, there are probably some things where it's genuinely more useful to think of it as an unevaluated thunk.
For example, I'm unsure if MIR's duplicate of UnevaluatedConst (which currently holds a direct DefId rather than an UnevaluatedConstKind) should be renamed or not - probably it should? but idk. Feel free to reach out if you're unsure about a use!
UnevaluatedConsts tend to be extremely conceptually similar to type aliases, and it's useful to think of things like const items as behaving like type aliases. We have
AliasTy/AliasTyKind,AliasTerm/AliasTermKind, and currently,UnevaluatedConst/UnevaluatedConstKind. The last should be renamed toAliasConst/AliasConstKind.My guess would be that not all uses of the word "Unevaluated" should be blindly replaced to "Alias", though, there are probably some things where it's genuinely more useful to think of it as an unevaluated thunk.
For example, I'm unsure if MIR's duplicate of UnevaluatedConst (which currently holds a direct DefId rather than an UnevaluatedConstKind) should be renamed or not - probably it should? but idk. Feel free to reach out if you're unsure about a use!