[Sigia-l] RE: what would porphyry do? (was apple and pears)

[Listera]

"John O'Donovan" wrote:

> To model this sort of stuff it helps to know something about data modelling
> and object modelling and a range of LIS issues.

Or you could take an entirely different approach. Rather than hard coding
any sort of classification (relational or object), you can rely on automatic
indexing/relating/rating. The pros and cons of this has been previously
discussed a bit here, when going over Google vs hand-crafted classification.
I don't want to rehash it here, but I do want to point out the scale and
speed of development in this front is probably considerably greater than
human-powered methods. It's thus enormously attractive to enterprises where
content/data is so spread out and rival camps are so entrenched. This
approach brings together hitherto unrelated/unrelatable resource under a
common umbrella/interface, mostly in the background, without much
disruption, far sooner than otherwise and often at considerably lower costs.

Ziya
Nullius in Verba