Re: Do Data Models Need to built on a Mathematical Concept?
Date: 7 May 2003 10:26:03 -0700
> > It is the rdb model which requires each
> > tuple to have the same degree which does not match reality.
> > In reality, things need not have the same number of attributes.
> Well, I'm not clear on how we decide whether a mathematical
> model represents reality or not.
It is not a matter of whether a model represents reality perfectly or not. It is not a black or white distinction. It is a matter of degree or grayness. We should judge a model by it degree of accuracy to represent reality, keeping in mind that our models will probably never be perfect, and need not be for many applications.
Depends on your definitions of things, especially "see" and "attributes", and level of abstraction :)
> When people try to give me examples in programming (especially
> OOP) and the classes are called things Cow or Plant or whatever,
> (and maybe they even point to their desk plant) I stop them right
> there. "That plant is not a class or instance; it's a plant," I say.
> "Let's talk about LinkedList or RGBColor or some such."
The plant on the desk could be modeled as a class/set/domain, if it can or does create other plants. The plant could be modeled as an instance/element/value, since it came from another plant like it. Many things can be abstracted/represented in our minds and in computers using linked-lists, RGB, etc.
> I don't think programming really has anything interesting to say
> about reality.
This is an interesting observation since, I see you as a very complex program and you are saying interesting things about reality.
> Programming is about math, and most of what's
> interesting in math has nothing to do with reality.
I would strongly disagree. Math is mostly interesting when it models reality accurately enough for our purpose.
> But if that's our criteria for deciding what should go into math, well,
> we'd better get rid of negative numbers then. I certainly
> never had three oranges and had someone take away five
> of them. I never encountered anything that had an irrational
> number in it.
The reason we don't get rid of negative and irrational numbers is because in some applications they do model reality fairly accurately and therefore are useful. Received on Wed May 07 2003 - 19:26:03 CEST