Re: Declarative constraints in practical terms

From: David Cressey <dcressey_at_verizon.net>
Date: Sun, 26 Feb 2006 20:35:21 GMT
Message-ID: <dioMf.1109$SJ2.554@trndny01>

"Bob Hairgrove" <invalid_at_bigfoot.com> wrote in message news:l20402l2h5079vm6qbqf7tt2kng3gclkqg_at_4ax.com...
> On Sun, 26 Feb 2006 19:13:45 GMT, "David Cressey"
> <dcressey_at_verizon.net> wrote:
>
> >It seems to me that "CREATE TABLE ..." is imperative. When the
imperative
> >is carried out, two things happen. First the table is created. Second,
the
> >table's description is included in the metadata.
> >
> >It's not hard to come up with an engine that will read the metadata, and
> >reconstruct the original create scripts, or something logically
equivalent.
>
> Most RDBMS such as DB2 and Oracle have had this capability for a long
> time.

You are right. The RDBMS that I know the best, DEC Rdb/VMS had htis capability built into RMU/EXTRACT as far back as 1986. I keep coming back to Rdb when I want to make a point about the internals, but that reference is growing more and more meaningless to most readers as time goes by. Received on Sun Feb 26 2006 - 14:35:21 CST