3 ms·

I really like in-shell floating point operations. But this: > / 1 0 +Inf Is not the correct answer - even as a limit - since the limit of 1 / x (when

I really like in-shell floating point operations. But this: > / 1 0 +Inf Is not the correct answer - even as a limit - since the limit of 1 / x (when x -> 0) can be +Inf or -Inf.

xiaq 12y ago
The 0 is a positive one. > / 1 -0 -Inf BTW this is just IEEE floating point.
- bnegreve 12y ago
  Ah yes indeed [1], I didn't know. [1] http://en.wikipedia.org/wiki/IEEE_floating_point#Exception_handling http://en.wikipedia.org/wiki/IEEE_floating_point#Exception_h...