[Sigia-l] visual thinking, rambling about the past

[Listera]

Ed makes two significant observations:

> Space is no longer a problem ...

This is important for more than financial or technological reasons. In the
mid-90s, you'd pay over $1,000,000 for a terabyte of storage, hanging off a
very expensive SGI or Sun server. Today you can put together a terabyte RAID
for about $1,000 to work with a $1,000 PC. That progress has been made in
far less than a decade.

This has enormous architectural as well as political ramifications. For one
thing, we can collect and keep far more information. But we no longer have
to keep everything in one central, expensive server space. Information can
now be distributed to many 'clients' that can have access to terabyte-size
storage. This further makes possible the virtualization of data/info as well
as the ability to segment massive amounts of data, encrypt them and store
them in chunks globally, to be retrieved transparently. (The latter forms
the basis of many of the P2P and 'file sharing' systems).

This is an impending tsunami and IAs ought to get ready for its
implications.

> I slapped my forehead and said, How could
> I have the foresight to store that changing
> number, how would I know anyone would ask
> that?

You can't and that's why the notion of having one gigantic furniture site
'that has everything' is so utterly nutty. It's also the reason why we're
moving towards web services and distributed systems. Instead of looking
through inert/static web 'pages' we are going to be querying dynamic sources
of data/info. So if the data/info has been captured/stored or can be
generated than it'll be available as the result of a dynamic query as
opposed to a pre-rendered static page.

This is also fundamentally important for IAs to get used to.

After all, the strength of the Internet (our most extensive information
network) lies in its *distributed* nature and its ability to link servers
and consumers of information dynamically.

Ziya
Nullius in Verba