Research delayed-extension type parameters

[poteat]

There is a theory that explicit type Foo<T extends string> = T syntax as opposed to type Foo<T> = T extends string ? T : never syntax causes TS eager evaluation to happen - it could be that migrating to the latter form could improve inference performance / allow more "magically cyclic" structures.

[MajorLift]

That's super interesting, and great that there's an option for controlling lazy evaluation.

[michaelKurowski]

@poteat This is very interesting, could you provide some examples or source where did you read about it so that it would be possible to read up on the topic?