[
Lists Home |
Date Index |
Thread Index
]
And that's a good thing. Insistence on using
http: in namespace declarations is going to make
these choices a lot harder than they have to be.
While it has been said that those who don't
use resolvable URIs can live to regret it, it
is possible to have resolvable URIs that don't
use http. One can have REST without HTTP.
Consistent interfaces can be a feature of any
protocol. I'm not sure one can have a very
reliable Internet service system without
consistent identifiers.
HTTP like RDF has to thrive on its own merits and they
are, as Fielding says, consistent interfaces.
Even URIs are just data. If that defines the
system boundaries of the Web, good. That is a
coherent boundary. BEEP enabled systems can
share data with it, share languages with it,
and so on, but it is a different system. I don't
know anything about BEEP? Does it use the
spec'd URI identification system?
Is BEEP-bound SOAP considered to be "on
the Web" or "on the Internet" as "BEEP-enabled
services"? Web services don't have to kill
HTTP. In fact, they can't. They can make
it less important to those who need or want
services based on a different architecture
that emphasizes different features.
len
-----Original Message-----
From: bob mcwhirter [mailto:bob@werken.com]
Been lightly following this thread, but...
SOAP at least (not that it's the end-all-be-all of web services)
has recently been bound to BEEP. HTTP isn't the only thing out
there...
|
