On Mon, 5 Nov 2018 at 10:21, Axel Wagner <axel.wagner...@googlemail.com> wrote:
> Even though all we did was save some references to values, the code has >> radically changed (along with its runtime implications). >> > > you'd choose a suitable small subset based on the type-constructors used > in the signature. > For the record, the signature is sufficient - what happens inside the body of the function is important too, as evidenced by the fact that the set of required operations changed significantly when the body of the function changed, even though the signature remained the same. I was trying to achieve consensus, not to implement generics I'm on board with that goal. I've just been trying to understand how your suggestions would work, because without workable suggestions, there's unlikely to be useful consensus. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
