[
Lists Home |
Date Index |
Thread Index
]
That's ok Jonathan as long as one only publishes to the web.
But that isn't the same as system independent reusable information.
That's why I tried to make it clear what On The Web and Off The Web
mean... and it all comes down to the use of a system identifier
for a public name (read published).
That conflates the need to disambiguate with
the need to use the URI. In other words, all the web definitions
do is introduce tautologies to lynch the information to the web,
and the web is neither the first or last system that will host
this information. You can't get out of it: URIs are system IDs.
All you can do is talk around it until everyone accepts that
by specification, they should only publish to the web.
And it isn't unique. Using a mapping to conflate and deflate these
is precisely how earlier SGML hypertext systems worked too. The
unique bit in the Web is the use of DNS names and even before that,
host names were used. We dropped IADS and IDEAS on top of older
networks (Novell as I recall), and they *just worked*.
Because of the use of DNS and by dint of making that a name
for an authority, over time and practice, the system ID signature
becomes the assertion of ownership. That is a disaster unless
we get that on good legal footing. Not doing so is just the
orignal "let's ignore it until we solve it" attitude that got
us to PICS.
len
-----Original Message-----
From: jborden@attbi.com [mailto:jborden@attbi.com]
Violently agree. Many folks complain that this view conflates "name" and "locator", yet indeed this so-called conflation is exactly what makes the Web interesting, and why it uniquely does what it does.
This is the Web's most defining characteristic(IMNSHO).
|
