On Fri, May 15, 2026 at 6:44 PM Nicolas Grekas <[email protected]>
wrote:
>
> not so much - and only transitory for the libs that care about this - vs
> no similar path in the proposed RFC as is:
>
> class Foo#<T>
> implements Baz#<U>
> {
> public do(
> #<T> // <- this could be a way to address your concern Zebulan -
> for Larry, that's already what we do for attributes on args
> DateTimeInterface
> $bar,
> ) : List#<string>
> {
> // [...]
> }
> }
>
Would it be an option to add both variations, where #<...> functions purely
as forward compatibility?
