Re: NULLs: theoretical problems?

On Aug 26, 8:57 pm, "V.J. Kumar" <vjkm..._at_gmail.com> wrote:
> Jon Heggland wrote:
> > . You still have
> > the excluded middle problem that V. J. Kumar complains about.
>
> No longer, it appears, there is a problem because Jan modified his
> definitions ! In the updated version, def(x) :(x or not x ) is now
> true for undefined x ! Likewise, def(x) : (x or true) is also true
> for undefined x, so your surmise about the two SQL statements being
> different may be totallly off mark !
>
> It quite possible I misunderstood Jan's 'def(x) :true'. Who knows ?
> He may still be willing to present a coherent account of his DEF logic
> instead of feeding me with his punchy and petulant one-liners.

That didn't seem right to me, but since I haven't been following the thread too closely I reviewed the thread to be sure. I found no evidence that there has been any difference between any of the definitions Jan has given of his DEF construct. (Although the syntax has changed slightly from the original.) It has always been that def(x): f(x) evaluates to false if x is undefined, and evaluates to f(x) otherwise, where f(x) is a boolean formula in which x appears free.

Nor did I find any petulance. Instead, I was again struck by Jan's impressive patience in the face having to explain a simple concept over and over.

Marshall Received on Mon Aug 27 2007 - 08:53:49 CEST