Every time I hear the phrase "Web of Data", I'm reminded of this quote by Henry Baker:
Because "algebraic relation theory" was much more respectible than "data processing", database theoreticians could now get tenure at respectible schools whose names did not sound like the "Control Data Institute".


massimo-a-dentico said...

This is, to my knowledge, Henry Baker's most stupid comment, by a usually intelligent, acute and wise person.

I myself have cited that correspondence in the past, when I was completely ignorant about relational data model (RDM) and real problems in data management (despite I wrote the contrary: I was Unskilled and Unaware of It [.pdf] as any other person spouting out bullshit about RDM theory).

Now I know better: his rage is completely off track - it is evident that relational data model was/is often misunderstood. Lack of performance by part of RDBMes was due to lack of understanding by part of implementors, not for inherent constraints imposed by RDM on implementations (see my answer, in particular under "edit:" heading, to this question on stackoverflow.com - Database Structure Advice Needed - where I cite some recent promising implementations, at least concerning performance; some are based on SQL, that is really crappy anyway).

Besides, he was right on transitive closure, but in a trivial way. That inconvenience was easy to overcome: a transitive closure of a graph can be computed incrementally with the help of auxiliary relations (see this my other answer on stackoverflow.com: query language for graph sets: data modeling question, in particular Incremental SQL Queries link).

Recently someone (Paul C.) on USENET newsgroup comp.databases.theory, suggested that "..various non-technical IBMers tried to obstruct or even sabotage Codd, not just attacking his ideas but him personally..". If confirmed it's hardly surprising that knowledge of practitioners in the field is in a so pathetic state then.

What scare me most is that too much people don't care to question their own (lack of) knowledge with respect to relational theory (as instead I have done years ago).

Manuel J. Simoni said...

Thanks for the interesting pointers!