I received this email in response to my column No Integrity: A Systemic Problem.

 

From: LGFCD

To: Editor

Date: 06 Aug 2004

 

MySQL is copyrighted by its authors in the company by the same name. It is under the GNU GPL, which grants users ample liberties, but that doesn’t mean in any way it is in the public domain.

 

Just as with closed products, free software can also be garbage. But it does offer an alternative process where the few who understand can create something with little caring for the short term which drives big businesses.

 

Witness that Required Technology has for several years failed to bring anything to light, and that Alphora has been forced to implement SQL NULLs; meanwhile various free software projects like Opus, Duro and Rel have been slowly building tools that have already their usefulness and are relational, while not yet full RDBMSs.

 

 

Fabian Pascal Responds: We revised the text to say that MySQL is Open Source, not Public Domain.

 

Any tool can “create something with little caring for the short term which drives big business”, but our point was that promoting products that lack integrity not only as “reliable DBMSs”, but as relational DBMSs, is misleading, and contributes to poor understanding of fundamentals, defeating progress. It is one thing to know the fundamentals and build gradually to a sound foundation, and quite another to be ignorant of such and dismiss it as unnecessary, or “just theory”, or other such nonsense.

 

That Alphora had to succumb to pressure and implement NULLs proves our point rather nicely: even when an attempt is made to do the right thing, foundation knowledge is so utterly lacking and the promotion of lesser technologies so powerful, that technology must be degraded to be sellable. Incidentally, this is not the first time that has happened e.g. Ingres’ data language QUEL was superior in some ways to SQL (see my 1987 article Language Redundancy and DBMS Performance in Database Programming and Design), but was not “IBM compatible”. And it probably won’t be the last time either.

 

The status of Required Technologies (RT) has little to do with the “ability to bring anything to light”, and more to do with the dark side of US business, which does not negate our comments in the column. We regret to have omitted to mention the Open Source efforts to implement true RDBMSs (TRDBMS), Duro and REL, which are accessible via this site’s Links page. They are not full-fledged products yet, and we suspect they face an uphill battle given how the industry operates. We also believe, however, that RT’s technology is necessary for the implementation of any true RDBMS (TRDBMS), to bring it to a level at which even the industry won’t be able to ignore. This is particularly true with respect to NULLs: in PRACTICAL DATABASE FOUNDATIONS paper #8, The Final NULL in the Coffin, we propose a logically correct solution to missing data without NULLs, to which RT’s Transrelational Model lends itself particularly well.

 

 

Posted 10/08/04