[Sigia-l] Classifying the internet

[Ziya Oz]

> Why are we assuming that any given site only has one classification?

We're not assuming that, but creating classification systems *without* any
regard to usage is asking for trouble.

That's why I asked what the purpose of this classification is. If a
non-trivial 'site' maps to many types, and large ones potentially to dozens,
all multiplied by gazillion sites out there, just how meaningful the applied
results will be? (I don't know. Do you?)

Don't think of it as a website classification system, think of it more
abstractly as a cardinality matching system and you'll recognize *practical*
limitations. If you design data warehouses, for example, this is a common
architectural issue.

> Why can't amazon be both an e-commerce and an informational site?

Because essentially every site 'sells' something. Method of transaction, the
time it takes to consume that transaction or agents of transaction may be
very different, but one way or another there's 'selling' going on, whether
the user of that site recognizes it as such or not. (Hence the confusion on
the nature of 'vanity' here.)

If I slap AdSense on my blog, but don't consciously change the way I create
my informational content, have I suddenly transformed it into an ecommerce
site? At that point you jump into the abyss of semantic definitions, which
can only be usefully pruned in the context of a *purpose* for this exercise.

This is why framing a problem is the most important aspect of design.

-- 
Ziya

It depends.
If it didn't, you'd be out of a job.