[Haskell-cafe] ordNub

[Tom Ellis]

On Thu, Jan 01, 2015 at 02:14:16PM +0000, Tom Ellis wrote:
> So let me clarify that I want (<=) to be a total order in the sense that it
> is
> 
> * reflexive
> * symmetric
> * transitive
> 
> and `(x <= y) && (y <= x)` implies that `x == y`. (*)

Correction: I mean "antisymmetric" not "symmetric".  Anti-symmetry is
exactly this condition (*).

Futhermore I want (==) to be at least as fine-grained as extensional
equivalence.

Tom