Re: [xml-dev] Caught napping!

[Bob Hutchison <hutch@xampl.com>]

On 01/11/09 1:16 PM, "Joe English" <jenglish@flightlab.com> wrote:

> 
> PaulT wrote:
> 
>> Yes, we know that RDB does not solve the problem.
>> Honest.
>> But I'd put it slightly different.
>> "Almost all current SQL servers - kinda suck".
>> I can elaborate why, but that's not my point.
> 
> Can *anyone* elaborate on this?  Fabian Pascal
> goes on at length about how SQL is a poor shadow
> of the true Relational Model, but I can't seem to
> find any specific criticisms on his web site.
> (Presumably he goes into detail in one of his
> books, but I'm reluctant to shell out the cash.)
> 

You are probably looking for the paper "The Third Manifesto" by Darwen and
Date (ACM 1995 I think). I've found a link to a postscript version on this
page: http://www.palslib.com/Fundamentals/The_Relational_Model.html -- the
manifesto is in the section called "Objects and the Relational Model".

There are other articles here that might be of interest.

The basic problem is that SQL does not implement the relational model, it
implements something less. If you have ever tried modelling tree structures
in SQL you will have noticed some difficulties. If you had a true relational
model this would not be difficult. If I recall, the relational model can
model arbitrary datatypes as well (and objects).

Cheers,
Bob