On 4/23/07, Dan Bron <[EMAIL PROTECTED]> wrote:
RM> I believe negative_zero=:64{.1 satisfies this constraintThen we do not agree on the definition of "a J FP array" and so this meta-discussion is moot. I'm willing to bet the meta^:N-discussion will be too.
Yeah, I ignored the "FP" part -- though in retrospect, it probably means "floating point". That said, I think J could be modified such that the existence of this bit pattern as a floating point value in a J array would result in a domain error in all J contexts which care (which effectively, would mean that the array is not a J array). Or, if you prefer, it could be given the behavior of some other bit pattern which is not a valid J array. I'm not certain I see the value in these approaches, but I also don't see how it's "not possible". -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
