On Tue, Jun 15, 2010 at 11:54 AM, Chris Browne <cbbrowne@acm.org> wrote:
> I'd like that feature, and I don't think it takes too much arguing to
> get to the point that a declarative "IMMUTABLE" control is rather less
>
Not only that, but if you were to, say, make the PK field IMMUTABLE
you could then optimize out the locks involved when updating rows with
FK relations to the PK (or any other FK field, I suppose)