Robert Griesemer comments

Results 89 comments of


                                            Robert Griesemer

go/types, types2: inconsistent behavior with recursive generic types that produces deadlock, invalid recursive, unknown field error

We won't get to this for 1.24. Moving along.

spec: type inference can't infer types for a generic union of functions

Putting on hold because we also don't support calling the function in the generic code.

runtime/race: `reference to `WakeByAddressSingle'` when compiling with race on windows/amd64 1.21.11

@golang/windows

go/printer: Invalid reformatting of slice type with a /* comment.

Fixing this will require additional position information in the `go/ast.ArrayType` node (position of the `]`, trivial but requires API change) or better "guessing" of the position of "]" based on...

go/types, types2: simplify type-checking of method receivers with type parameters

Too late again for 1.21. Moving to 1.22.

go/types, types2: simplify type-checking of method receivers with type parameters

Moving to 1.22. We didn't get to this.

go/types, types2: simplify type-checking of method receivers with type parameters

Moving to 1.24. We didn't get to this.

types2, go/types: the scope of a type parameter identifier in a method of a generic type is too large

Hm. We're treating the receiver declaration like any other parameter list (including type parameter lists) for simplicity. I suppose [we could say](https://tip.golang.org/ref/spec#Declarations_and_scope): > The scope of an identifier denoting a...

types2, go/types: the scope of a type parameter identifier in a method of a generic type is too large

@go101 The suggestion for the new scope rule [I gave](https://github.com/golang/go/issues/51503#issuecomment-1059822040) seems good enough: the scope of ordinary parameters (including the receiver) is the function body and type parameters are also...

types2, go/types: the scope of a type parameter identifier in a method of a generic type is too large

The corresponding proposal has been accepted. This issue now tracks the implementation.