On Jan 4, 2008, at 9:35 AM, Rich Shepard wrote:
> It would be appropriate to convert NaN to NULL since the valid is
> literally unknown.
It's not unknown: it's known to be something other than a number,
which is not the same as unknown.
*considers quoting Rumsfield, but common sense prevails*
Michael Glaesemann
michael.glaesemann@myyearbook.com