> On Thu, Dec 02, 2004 at 10:53:37PM -0500, Berend Tober wrote:
>
>> I learned that the unusual behavior (or at least the behavior that
>> seems weird to me) regarding relational integrity and uniquness
>> constraints as been around for a while, and some people actually think
>> is is SUPPOSED to work that way ...
>
> Who would that be? Because I've always thought that most people
> (everyone?) think of it as a bug that nobody has bothered to fix. Not
> that the fix is easy, mind you ...
Sorry, I can't name names from recollection. One of the posts I read from
a few years ago was prefaced with some comment like "I'm not sure its
really a bug...but" and some talk about "the SQL Standard". My personal
take-away was that the behavior was not going to be fixed in the near term
and that I could not count on the behavior I expected, so I abandoned the
use of inheritance and implemented as described. And don't say "fix it
yourself" because I might be inclined to try, if I had the appropriate
skills.