On Sat, Sep 8, 2018 at 5:54 AM, Wojciech S. Czarnecki <o...@fairbe.org> wrote: > On Sat, 8 Sep 2018 03:59:45 -0700 > Ian Lance Taylor <i...@golang.org> wrote: > >> Why is it useful to say "this polymorphic function may >> only be instantiated with unsigned integer types?" > > E.g. because an algorithm needs this constraint.
OK, but, in general the Go type system is not very complex. There are many algorithms we write in Go today that have constraints on their types that we can not implement. For example, it is well known and accepted that an interface type may permit a type that will not work for the algorithm. So perhaps I should say it this way: how much complexity are we willing to tolerate to permit a polymorphic function to forbid some type arguments? To expand on that, I believe that we need a mechanism to set a lower bound, if you will, on the type arguments: they have to implement certain operations. The question is whether we need a mechanism to set an upper bound on type arguments, or whether it is sufficient to rely on documentation. Ian -- 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.
