Re: Relational and multivalue databases

"Dawn M. Wolthuis" <dwolt_at_tincat-group.com> wrote in message news:c16gdv$vue$1_at_news.netins.net...
> I was once told that there are people who think of cars as fascinating
> machines and those who think of cars as (perhaps fascinating) machines
that
> get us from one place to another. This person then said that we call the
> first group "men" and the second group "women".

;-)

> I certainly disagree with
> that as a blanket statement, but I suspect there could be emperical data
to
> suggest something similar related to computers.

You didn't have to spoil the effect of the previous paragraph with this sentence.

> I have no fascination for technology outside of how it provides
improvements
> for people or creation. I do have a love of mathematics for mathematics
> sake, however. And if that is what is going on with discussions of
> relational theory, then continue the game, by all means. There is no need
> to have implementations of the theory, however, unless it is useful.
Surely
> implementations of relational databases are useful and have been accepted
by
> IT professionals. However, some of the reaons they are employed have to
do
> with the false notion that there is something about relational theory that
> is "more mathematical" or more orthodox or pure. Companies then employ
> these beasts thinking they are more cost-effective.

No, people who tinker with things sometimes come up with very effective solutions. Hint: L.Torvald. [Subtitles: I don't mean they should necessarily be role models, either]

> So, if we are tinkering with cars in our garage, count me out -- I just
> don't care. If we are telling people that cars are better when they use
> more gas, and if people then believe these claims and buy bigger and
> "better" cars, then I do care and I feel a need to speak up and say it
ain't
> so.

No, if people tell you they want to save the world, that is most often is b**t.

> The purpose of tinkering with our car is to have really cool tires,
bumpers,
> etc? I sure hope not. I think Codd really thought he was coming up with
a
> better theory that would translate into something good in database
> implementations. There are good things about the relational model and its
> implementations and there are failures as well. There is nothing
anointing
> the relational database model from above as being the best approach to
> managing data in a software application.
>
> So, testing out various database theories and finding the pros and cons of
> each as it relates to the actual USE of products that attempt to implement
a
> theory seems like what one might want to do with a discussion related to
> database theory.

No, you don't have to spend your lifetime at customer site in order to be useful. You can't possibly have all time in the universe to check out all the crank theories out there. Received on Sat Feb 21 2004 - 04:05:40 CET