I have always assumed that extensionality is a paradigmatic example of a property of mathematical objects (sets) which is essential to those objects--- if your set theory doesn't obey extensionality, it isn't set theory.
Given the existence of alternative set theories, such as non-well-founded set theories, though, it occurred to me that maybe I shouldn't be so certain of this.
Which brings me to my question: have any set theories been developed which don't include the axiom of extensionality, or at least which include set-like objects which are not extensional?
This question is certainly related to my own, but doesn't consider theories which have been developed, only ways that extensionality might fail by manipulating models of ZFC.