[
Lists Home |
Date Index |
Thread Index
]
"Simon St.Laurent" wrote:
>
> > Simon St.Laurent scripsit:
> > Proxy them, same as we do with news: URLs (which likewise have no
> > locator-nature).
>
> Two questions:
>
> 1) There are no proxies for most URNs; can we bar them?
Just ignore other URNs. They are rare today because they are near
useless toda. When/if the semantic web comes about they will become
plentiful in proportion to their usefulness.
> 2) What if anything does this joyful baggage do for namespaces?
Let me suggest the following party line, which I believe to be
understandable to newbies:
* Namespace names are globally unique identifiers like UUIDs.
* But they use URI (URL) syntax rather than UUID syntax.
* Software applications do not typically dereference the URI and do not
care what information is there. You don't have to put anything there if
you do not wish to.
* The first thing you (as a human) want when you encounter a new
namespace name is documentation. Therefore, by convention the URI points
to human-readable documentation. The use of HTTP URIs (URLs) is *polite*
but not required.
* Namespace names are HTTP URLs *not* because computers will
dereference them, but because humans will. (at least today)
* In the future, it may become clear how to put machine-processable
information at a URI (semantic web stuff). Then those URIs will host
both human-readable and machine-readable data. RDDL is an early
experiment in that direction.
* If you use URNs you will NOT have any easy way to host either
documentation or machine-readable descriptions. So don't do that.
--
"When I walk on the floor for the final execution, I'll wear a denim
suit. I'll walk in there like Willie Nelson, John Wayne, Will Smith
-- Men in Black -- James Brown. Maybe do a Michael Jackson moonwalk."
Congressman James Traficant.
|
