IP Multicast in Differentiated Services (DS) Networks
draft-bless-diffserv-multicast-07

From: RFC Editor
Subject: authors 48 hours: RFC 3754
          NOW AVAILABLE
Date: Mon, 5 Apr 2004 16:25:03 -0700
To: bless@tm.uka.de, klaus@wehrle.com
Cc: RFC Editor , Bill Fenner
   

From Roland:
the abstract and intro, as they currently read, were suggested by the RFC editor.
However, I'll try to provide one-two sentences summary about the problem.
Should we submit another draft, or, to whom should we provide the additions?

--

As said before, the current abstract and intro are a suggestion
from the RFC editor which we adopted. However, I tried to express the
main problems in a few sentences (not an easy task...). Here is the
proposal:

Replicated multicast packets normally get the same forwarding
treatment as the original packets (because of the copied DiffServ
codepoint). Due to the dynamic creation of multicast branches, it is
difficult to control the amount of network resources that are consumed
by multicast traffic. This leads to situations where other flows (even
unicast traffic) are affected adversely by multicast traffic.

Maybe a native speaker can still improve or even shorten this.

The first paragraph of the Introduction really is poor.  It says:
 
    This document discusses the problems of IP Multicast use in
    Differentiated Services (DS) networks, expanding on the discussion
    in RFC 2475 ("An Architecture of Differentiated Services"). It also
    suggests possible solutions to these problems, describes a potential
    implementation model, and presents simulation results.

What are "these problems?"  I expected the introduction to list the problems for which it proposes solutions.

From: RFC Editor
Subject: Informational RFC to be:
        draft-bless-diffserv-multicast-07.txt
Date: Mon, 29 Sep 2003 09:50:24 -0700
To: IESG , iesg-secretary
Cc: RFC Editor , bless@tm.uka.de,
    klaus@wehrle.com

IESG,

This RFC-to-be was submitted to the RFC Editor to be published as
Informational: draft-bless-diffserv-multicast-07.txt.


Four week timeout is initiated (27 October 2003).

I've been talking with Roland about this draft since Salt Lake City; I'm happy with its content now and expect it to come in via the RFC Editor.