It is perfectly possible for an object to export memory in a read-only way that may still change. Another object may have a writeable view:

x = obj.readonly_view
y = obj.writable_view
obj.move_to_next_image() # changes memory under x, y

So, hashing using tobytes() doesn't make any sense at all to me. A memoryview != its contents.

You could compare memoryviews simply by comparing the Py_buffer structure, but that's going to be confusing for a lot of users. I would really prefer unhashable (+ if needed a method for comparing contents).

(FWIW, not sure how relevant this is; in NumPy, == does 

In [1]: np.array([1,2,3]) == np.array([1,3,3])
Out[1]: array([ True, False,  True], dtype=bool)

Cython will follow this behaviour for its "typed memory views", which is Cython's builting mechanism for PEP 3118 which is not quite the same as CPython "memoryview". But I understand that following this behaviour is probably out of the question for CPython.)