Hi Helen,

I think Martijn has been mumbling with this idea for more time than I. If I understood
him correctly the problem could be divided in:

1) On DDL statements analize dependencies.
2) If necessary flag objects as "invalid"
3) Allow construct "table.column%type" for extracting underlaying datatype of existing
object.

Mmmm... I don't percieve the action of invalidating an object because of a DDL change
a lack of integrity Helen, au contraire!

Regards,
Daniel Berstein.

On 8 Aug 2003 at 9:23, Helen Borrie wrote:

> At 04:19 PM 7/08/2003 -0400, you wrote:
> >Is there any (good) reason why a domain can't be used as the datatype of
> >stored
> >procedure parameters?
> >
> >It makes little sense to me to use a domain for columns definition but
> >have to refer to
> >the underlaying datatype when coding stored procedures.
>
> PSQL modules are compiled on the basis of the structures of variables at
> compile time - they are not interpreted at run-time. What happens to that
> compiled code if the domain definition changes? When it comes to a choice
> between integrity and convenience for programmers, integrity always wins.
>
> heLen
>
>
>
>
> To unsubscribe from this group, send an email to:
> firebird-support-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>