Bill Baxter wrote: > I guess that's what's missing in D, some way to write an operator > function outside of both A and B that makes the final decision. Supporting that means supporting ADL, something I want to avoid.