Re: Poll: Expert user vs. Internals Expert

"dawn" <dawnwolthuis_at_gmail.com> wrote in message news:1148645874.367179.161900_at_i39g2000cwa.googlegroups.com...

[snip big poke at RM fundamentalism]

> Fortunately, the world has encountered this type of thinking before,
> among religious fundamentalists. Unfortunately, society has not
> figured out how to address it. Such fundamentalism in any religion is
> scary, the RM being no exception.
>
> But, of course, I do like logic and set theory, employing such while
> working with non-relational databases. Cheers! --dawn
>

Here's my take on it: the only response that passes the sanity test for me is to appeal to the eighty-twenty rule.

Here's how it plays out for me:

The RDM gives me 80% of what I want in a data model. Certain things, like trees can be expressed more elegantly in the hierarchical model than in the relational model. But there are workarounds, as the nested sets technique shows. On thother hand, things like many-to-many relationships are handled more elegantly in RDM than in HDM, and the workarounds are much clumsier. So I have a bias towards staying within the RDM and doing what I have to within that bound.

Sometimes, I have to step outside that bound. So be it.

SQL gives me 80% of what I want in an interface language between a programming environment and a relational database. Much has been made of the defects of SQL from an RDM purist perspective. Those defects are real, and should be addressed on the next iteration. But the defects are not so fatal as to render SQL either useless or harmful for implementing an approximation to a relational design. You just have to know what your doing, and steer between the potholes.

Oracle RDBMS gives me 80% of what I want in an SQL based implementation of a relational DBMS. Oracle got some things wrong, and some of those things persist from one revision to the next. I'm not ready to abandon Oracle until I find something better. And, yes, I do use marketshare as one indicator, among many, of what might be worthwhile learning.

The above recipe is a good pragma for people who want to design and build within the current framework. It's a recipe for stagnation among those pioneers who want to invent or discover something better. There is room for both ways of operating in the real world. I believe there is room for multiple ways of thinking even in this newsgroup. However, there are participants who evidently disagree. Received on Fri May 26 2006 - 15:11:38 CEST