[Haskell-cafe] Limitations of generalization of type constructors

[Tom Ellis]

On Mon, Sep 18, 2023 at 09:23:19AM -0400, Ryan Scott wrote:
> Short answer: This is expected behavior, and it's unlikely that behavior
> will change any time soon.
> 
> Long answer: The key phrase in your original post is "the quantifier
> moved".
[...]
> The reason why this restriction exists is because there is no type-level
> counterpart to `\ (@z)` in Core, which would be required for
> regeneralization at the kind level. 

Aha, thank you, Ryan, for the helpful explanation.  In that case I
will continue to write data types specialised for each use case.
That's not too painful in my particular application.

Whenever I get stuck on this stuff I marvel at how wonderfully
everything works out at the value level!

Tom