Re: Demo: Modelling Cost of Travel Paths Between Towns

"Neo" <neo55592_at_hotmail.com> wrote in message news:4b45d3ad.0411191557.273fb700_at_posting.google.com...

>

That page doesn't seem to be loading right now ...

But anyway, part of the difficulty in the conversation is that you're not actually trying to solve the same problem that databases exist to solve, which is data management: structure, integrity, manipulation. You seem to be focused on representation, (with an emphasis on dynamic structures) and querying, but no integrity control.

Indeed, in some ways, dynamic structure and integrity control are *incompatible.* You can't have both. Either you have a static (but updatable) schema, and you get the benefits and pay the costs thereof, or you don't, and benefits costs etc.

If one is quite concerned with data integrity, then a static schema is a very important tool. Since you're not concerned with integrity (so much) you don't appreciate the benefits a static schema produces; it's not what you're after.

Since many of the people here with a database background are quite concerned with integrity, and not concerned so much with dynamic structure, (since it doesn't help them with the problems they have) your quest seems misguided.

(There are also some disagreements about fundamentals, such as redundancy.)

So, I could come up with some pizza topping example schema, but you'd evaluate it based on how well it would handle conversion to a sandwich shop, (flexibility) and many of the others would be evaluating it based on how well it *prevented* you from ordering sandwiches with it.

There is a question of priorities.

Marshall Received on Sat Nov 20 2004 - 01:36:24 CET