[HTTPSubstrate-16] Propoposed criticsm of RFC3205

The TAG feels that the tone of the RDF is likely to promote 


Here is the gist/fodder of the proposed  comment on RFC 3205.  You were 
going to write a comment

Whereas tunneling of application protocols over application protocols 
is generally messy,
the reuse of HTTP  in keeping with its design principles is a good 
Indeed, the use of HTTP to create web of information should not be
limited to "conventional web browsing".

A very large number of applications
for data and commerce will hopefully publish all kinds of data for
use by all kinds of ways. To attempt to partition this space of data,
even though it contains many applications, is bad, because
interoperability across applications is extremely valuable.
Also, human-readable "traditional" web pages and
machine-readable resources are often closely connected,
and must often be, in practice, run on the same server on the same port.

It is quite reasonable for a web server server, on port 80
to serve a great variety,

There are many new applications in which there may be a temptation to
invent a new protocol, but a much better way is to reuse HTTP.
This cuts down code size, and reuses features such as proxy
caching and all the tools and support which exists for HTTP.

(Maybe draw attention to the relevant TAG findings about HTTP get, etc)

Tim bl

Received on Thursday, 6 February 2003 18:10:40 UTC