Claus Reinke wrote:
most of us would be happy if instance contexts would be required to uniquely determine the instance to be chosen, a rather conservative extension of current practice.
I'm not so sure about the "most of us", as you note yourself the defaulting pattern is quite popular (and useful). I certainly couldn't live without it. And even that aside, I'd much rather have the type system infer a "most particular" instance than to have to specify that myself. Also IMHO, adding a new construct (type (in)equality) to the language is a lot more obtrusive than to do something meaningful of the constructs that the language already provides. So I'd have issues with "conservative" as well... Of course, this is all from the perspective of a user, not a type inference engine implementor... ;-) /Niklas