Timon Gehr:

IIRC Don said this shouldn't work since the context is not actually known at compile time.

I remember Don saying similar things on some other usages of 'enum'. If all this needs to become accepts-invalid errors, then dmd 2.065 seems a good moment to fix this messy situation. Suggestions from Walter/Andrei are welcome.

Bye,
bearophile

Reply via email to