Date: Fri, 18 Jun 1993 16:13:32 -0700
From: Bob Braden <braden@isi.edu>
To: jam@iti.org
Subject: The Internet Standards Process
Cc: vcerf@CNRI.Reston.VA.US, braden@isi.edu

Jonathan Morrell,

Vint Cerf forwarded to me the information on your Standards Development
workshop.  I write as Executive Director and long-time member of the
Internet Architecture Board (IAB).  Over the past 10 years, the IAB has
initiated, and helped to shape, the development of the current
practices for setting standards for the Internet.  Many of the people
deeply involved in setting Internet standards have been known to
express deep conviction about the effectiveness of this process, and
there is some commercial success to support their belief.

The Internet standards process is outlined in our document RFC-1310,
"The Internet Standards Process".  I can email you a copy if it is not
readily available to you.  However, the formal rules are only a small
part of the story. The rules were developed,and they operate, in the
context of a rather distinctive social culture. This is the culture of
the Internet community in general and the Internet Engineering Task
Force (IETF) in particular.  One cannot understand how Internet
standards are created without some idea of how IETF meetings are run,
and how the IETF attendees view themselves and what they are doing.
For this purpose, I recommend RFC-1391, "The Tao of IETF: A Guide for
New Attendees of the Internet Engineering Task Force".  I can
also supply a copy of this to you.

Standards setting cannot avoid politics or the profit motive.  The
Internet standardization rules and the culture of the IETF recognize
these inevitables, but attempt to subordinate them to considerations
of technical excellence.  One feature of the process that may be unique
is its reliance upon consensus rather than formal voting.  Indeed,
building a technical consensus is the bottom-line objective of the
process.  It does not always succeed, but there is a genuine attempt to
make it work.

Another feature of the IETF process that is not apparent from the
formal rules is its heavy reliance upon the Internet itself, for email
and file transfer.  One of the first steps in moving into a new work
item or topic is to create a mail exploder for discussions of the
subject.  Collaborations start easily and move efficiently using email
and FTP for rapid interchange of working documents.  Recent experiments
have shown the feasibility and utility of broadcasting IETF meetings,
using audio and compressed video, across the world-wide Internet.  I
expect that there will be a rapid increase in the use of
teleconferencing and other collaboration technology to carry on the
work of the IETF.  One cannot overcome the tyranny of timezones, but we
are attempting to use our own technology to overcome some distance
barriers.  (I neglected to mention earlier that the Internet standards
activity is increasingly international).  On the other hand, face-to-face
meetings still appear to be necessary to make decisions.

As I have observed, an essential component of the Internet way to make
standards is the particular culture of the Internet technical
community.  The specifics of the Internet approach probably will not be
applicable to a different community with a different culture.  However,
by looking closely at the non-traditional Internet approach, other
groups may be able to synthesize innovative approaches to
standardization that are suited to their own communities.

Bob Braden