Re: Can these constraint be implemented in an RDBMS ?

From: Jonathan Leffler <jleffler_at_earthlink.net>
Date: Tue, 02 Mar 2004 15:06:19 GMT
Message-ID: <Lp11c.17721$aT1.1256_at_newsread1.news.pas.earthlink.net>


ben brugman wrote:

> "Alfredo Novoa" <alfredo_at_ncs.es> wrote:

>>"ben brugman" <ben_at_niethier.nl> wrote:
>>>In another thread I was asked to produce an example which
>>>can not be implemented with RDBMS constraints.
>>><snip>
>>>The implementation has to be done in Oracle or SQL-server.
>>
>>There is a clear contradiction. Oracle and SQL-Server are not RDBMSs.

>
> Can you clearify this a bit more ?

In the c.d.t news group, it is generally considered that there are few, if any, actual RDBMS - Alphora is generally reckoned to be the best there is.

Most of the other DBMS mentioned are SQL DBMS - a crude and very deviant approximation to a TRDBMS (true relational database management system) - or are multi-value DBMS (typically Pick-based, though Dawn Wolthuis has another name which I forget).

> Can you name RDBMSs ?

Alphora.

> Do you have a solution for the problem if this requirement (Oracle
> or SQL-server) is dropped. The imaginairy customer still likes to
> make a choice between at least two 'RDBMSs'.

I've not checked the problem definition in detail, but I suspect both the SQL DBMS you mentioned can manage it - unless Bob was being too brusque in dismissing the problem as solvable. And most of the others, with the probable exception of MySQL (and even that might be mildly slanderous), can handle it too.

-- 
Jonathan Leffler                   #include <disclaimer.h>
Email: jleffler_at_earthlink.net, jleffler_at_us.ibm.com
Guardian of DBD::Informix v2003.04 -- http://dbi.perl.org/
Received on Tue Mar 02 2004 - 16:06:19 CET

Original text of this message