Extending my question. Was: The relational model and relational algebra - why did SQL become the industry standard?

"Lauri Pietarinen" <lauri.pietarinen_at_atbusiness.com> wrote in message news:e9d83568.0302101114.7fd43f21_at_posting.google.com...
> > 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?
>
> Good question!
>
> It is probably related to many things, among them
> - the supremity of IBM during 70's and 80's
> - being afraid of having to remove duplicates,
> because of (alleged) potential performance implications -
> hence SQL's pseudu-tuple-bag model
> (see http://www.dbdebunk.com/cjddtdt.htm)
> - basically the fact that the relational model
> as presented by Codd in 1970 was such a
> radical departure from the traditional way
> of thinking that a watered down version was
> (and still is!) the only version that the
> public can stomach
>
> The product mentioned in another posting in this thread is
> called Dataphor; more information can be found at
> www.alphora.com.
>
> There is a web site dedicated to The Third Manifesto
> (www.thethirdmanifesto.com). It does not have
> as much information as the book but it is a good start.
>
> I see that Hugh Darwen has made his presentation
> "The Askew Wall" available on the site.
>
> regards,
> Lauri Pietarinen

Thank you very much Mr Piertarinen, and everyone else replying to this thread.

After reading quite a lot about this very interesting subject that I didn't know was so controversial as it seems to be, I'd like to extend my question a bit further.

First and foremost it seems that CJ Date and Pascal are the most vocal of the critcs of the SQL language, and they seem to be making a whole host of valid arguments of the website Database Debunkings(www.dbdebunk.com). Are these arguments considered correct and valid by DB-theorists in general, or is this a two-man crusade type of thing? The article Double Trouble, Double Trouble Part I by Date(http://www.dbdebunk.com/cjddtdt.htm) is especially interesting, as it concerns the inclusion of duplicate tuples(rows) in SQL-based DBs. I have also heard of, although never practised, optimizing a DB with duplicates. Are there arguments FOR such a strategy? If not, as Date claims there isn't, why would people create duplicates for optimization? Ignorance or deeper-knowledge of the shortcomings of SQL?

Best regards
Amund Trovåg Received on Tue Feb 11 2003 - 11:09:20 CET