George,

I'm not sure to what part of Fabio's message you're referring to, however I agree with his assesment that the "avoid nulls at all costs" seems very strange.

the article specifies that the problem is not so much the internals of SQL server (I really don't care), but more in the complexity of handling three value logics. I agree that Handling three value logic is sometimes difficult to handle, but there are lots of cases and examples where you cannot arround it. How do you store empty dates in SQL server? Well I know that whole tribes are using NULLs because SQL server does not have an equivalent to an empty date.

Also when storing scientifical data, which might be missing in stead of 0 ? What value do you store ? a Default value ? If so, how do you handle calculation of statistical values without NULLs ? Using a filter to exclude the default value ?

Well, this Ron might be the project manager of SQL server 7.0, but does not make him a good database developer?

Also not that we might talk about different versions of SQL server, with possibly different technical implementations.

Walter,