| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
On Apr 22 2012, Vincent Lefevre wrote:
On 2012-04-20 16:11:45 -0400, Ian McIntosh wrote: However in a math expression, it is not possible to differentiate these two zero representations, and for P1788 they would be regarded as a single zero. And it didn't have signed infinities in the integer types (let's recall that Nick suggested signed zeros for signed infinities).
Er, no, I didn't. I am not sure which remark you are referring to. Signed infinities are both useful and mathematically well-defined in themselves, but signed zeroes are mathematically well-defined only if they are distinguished from true zero. Regards, Nick Maclaren.