Re: The relational model and relational algebra - why did SQL become the industry standard?

From: Alfredo Novoa <alfredo_at_ncs.es>
Date: 10 Feb 2003 05:44:18 -0800
Message-ID: <e4330f45.0302100544.1a1965a_at_posting.google.com>


"Amund Trov g" <amund_at_texassibir.com> wrote in message news:<b208pe$27s0$1_at_toralf.uib.no>...
> Hi all DB-theorists! :)
>
> I am currently trying to work out why SQL and not relational algebra became
> the industry standard in 1986(I think it was). Is this because of the more
> "technical" language of rel. algebra and the difficulty in creating
> universal queries, or is it something else?

I don't think it was due to strong technical reasons.

SQL was designed by IBM labs, and IBM had an have a lot of power on the IT industry.

SQL is the VHS of the database languages. Quel could be the Betamax (better but dead).

I recomend you this little book:

THE DATABASE RELATIONAL MODEL:
A RETROSPECTIVE REVIEW AND ANALYSIS
(ISBN: 0201612941) Codd prefered the relational calculus as the basis of relational languages instead of the relational algebra.

This could be one of the causes because most of the first relational languages were based on the relational calculus.

By the way, Date has doubts about the strength of Codd's arguments.

Another reason could be that SQL designers thougth that relational algebra and calculus look too mathematical for the average practicioner, and they tried to make something different. But when they saw that the results were not very good they introduced again some features from the relational calculus.

SQL started as a simplified lab prototype, and its very poor scalar type support was probably due to that.

> I am a student at a master's program in information science, and I am trying
> to see if there is a reason for doing my dissertation on something that
> involves making relational algebra more usable, by e.g. making new
> operators. Right now I am trying to figure out why relational algebra is a
> dead language.

As Finarfin said, the example language of The Third Manifesto is based on the relational algebra.

www.thethirdmanifesto.com

IMO the relational algebra is more or less as usable as the relational calculus. Some expressions are easier using the relational algebra and some others are easier with the relational calculus.

> Thanks in advance for any help and pointers.

Good luck.

Regards
  Alfredo Received on Mon Feb 10 2003 - 14:44:18 CET

Original text of this message