Maybe he should have said the design is done for go1.

Regardless, issues are still filed under the tag of `go2`, even by Rob Pike. There's no reason to believe that go2 won't eventually happen.

On the other hand, there's every reason to believe the "Go [a]Way" prohibition on certain language features will leave Go 2 insufficiently attractive vs Go 1 to overcome the switching costs, dragging out any migration period. See also: Python 3.