From: David Aldridge

Date: 15 Jul 2005

 

I enjoyed your latest Database Magazine column in which you point out the folly of referencing a site such as whatis.com, but I was wondering whether you have time to share your thoughts on the Wikipedia entries for Relational Database, Model, and Database Management Systems?

 

http://en.wikipedia.org/wiki/Relational_database

http://en.wikipedia.org/wiki/Relational_model

http://en.wikipedia.org/wiki/Database_management_system

 

Thanks for your Database Debunkings website in general, by the way. As an Oracle data warehousing practitioner I really do cringe inside whenever I denormalise, or chose to not enforce foreign keys, in the name of performance.

 

 

From: Hugh Darwen

 

Yes, I was aware of the existence of this stuff.  Thanks for drawing it to my attention again.  I have made drastic changes to the entry for relational database (including the correction of the French entry for Ted Codd, where his first name was given as Edouard!). 

 

As for the entry for relational model, I think it needs to be changed to say that the first implementation was PRTV (ISBL), not the Multics thing. Otherwise it looks okay in principle, though there are no doubt several minor changes I would like to make if I can find time (I haven't looked at the detail closely).

 

 

From: Fabian Pascal

 

Hugh Darwen has made some changes to

 

http://en.wikipedia.org/wiki/Relational_database

 

Note that there are some differences between Date and Darwen's interpretation of Codd's relational model, and David McGoveran's and mine. One such difference is explained in PRACTICAL DATABASE FOUNDATIONS paper #1, Truly Relational: What It Really Means.

 

I hasten to point out that the two interpretations are not necessarily inconsistent.

 

 

From: David Aldridge

 

Wikipedia attracts its fair share of criticism, but when you know that folks of Hugh’s caliber have contributed to it, then it is a definite reassurance to theoretical novitiates like myself.

 

 

From: Fabian Pascal

 

1. Hugh does not contribute to it, but when I referred your query to him, he decided to make some revisions.

 

2. I just received back the Wikipedia printouts I sent Chris for review, chockfull of scribbled comments on various flaws and the following general assessments:

 

·   Relational model: The writer's heart is in the right place. Clearly influenced by Date and Darwen. But too many flaws, not all minor. And even if those flaws were fixed, I don't think the article would help a reader very much if he/she didn't already know what the relational model was!!!!

 

·         Database Management System: Overall I don't find this very impressive.

 

·         Relational database: Marked with a check, meaning OK.

 

 

Posted 9/16/05