On 03/02/2014 02:03 PM, Marko Rauhamaa wrote: > Michael Torrie <torr...@gmail.com>: > >> I don't see why == wouldn't continue to work if os.POSIX_FADV_RANDOM >> became an object of a different type. > > It probably would. > > If one were begging for trouble, one *could* define: > > class ABC: > A = 1 > B = 1.0 > C = 1+0j
And one could also set A=1 and B=1 if he was trying to be stupid. That would fail the equality test and the identity test (in CPython). Seems like this argument is getting a bit on the absurd side. The normal idiom is to use equality checks to test state variables' *values*. I don't know of any developer that would purposely try to break that when defining a new module or class. If Mark H wants to use an idiom that isn't conventional, or isn't widely used, he is free to do so; I can't see much harm in it. But certainly it's not the "normal" way that it's done in Python from what I can see. -- https://mail.python.org/mailman/listinfo/python-list
