Exceptions are actually part of floats, they're called "signalling nans".
So technically Python is correct when it decided that 0.0/0.0 should raise an exception instead of just quietly returning NaN. Raising an exception is a standards-conforming option.
In practice, I've found signalling NaNs to be completely unworkable and gave up on them. The trouble is they eagerly convert to quiet NaNs, too eagerly.
I am firmly in the belief that sNaNs were a mistake in IEEE 754, and all they really serve to do is to create hard trivia questions for compiler writers.
technically I guess it should return sNAN (so app can check for it if it want to handle it differently) and raise exception if sNaN is used in (non-comparison) operation
