Proposed Issue: RFC3205

The IETF has recently published RFC3205, "On the use of HTTP as a 
Substrate" [1] as Best Current Practice.

This document makes a number of recommendations regarding the use of 
HTTP. Some are reasonable, such as guidelines about what kinds of 
scenarios the HTTP is most useful in, how to use media types and methods 
to extend the HTTP, etc. However, it also bases a number of 
recommendations on a fuzzily-defined concept of 'traditional use' of the 
HTTP. These directives may seriously limit the future potential of the 
Web, effectively freezing its capability to common practice in 2001.

During its review, issues were raised from the perspective of the XML 
Protocol WG, but were represented as individual, rather than 
representing the WG or the Consortium. The IESG notes[2]:

   The W3C have discussed the document, and the response from W3C can be
   found as http://lists.w3.org/Archives/Public/xml-dist-
app/2000Dec/0061.html.
   The IESG notices the input from the W3C, but, find the points on lack 
of
   more specific information on what is appropriate or not not being 
something
   that makes the document clearer. The IESG finds the goal of this 
document is
   to raise issues, and not specifically answer all questions.

I request that the TAG review RFC3205 to determine whether its 
recommendations conform to the current conception of the Web 
Architecture and/or the design goals of the HTTP, as well as to examine 
the issues it raises.

If the TAG determines that it is detrimental to the Web, I know of are 
two potential courses of action;
   A. request that the document be retired, as per RFC2026 [3], Section 
6.3.
   B. submit a document and request that it replace RFC3205.

In the event that the TAG chooses to author a replacement, I'd be happy 
to help, availability permitting.

Note that a process appeal is AFAIK not possible; the window for appeals 
is two months from the protocol action. There may also be 
inter-organisational coordination issues involved; however, I imagine 
they are out of scope for the TAG.


1. http://www.ietf.org/rfc/rfc3205.txt
2. http://www.ietf.org/IESG/Announcements/draft-moore-using-http.ann
3. http://www.ietf.org/rfc/rfc2026.txt

--
Mark Nottingham
http://www.mnot.net/

Received on Sunday, 24 March 2002 20:50:34 UTC