FRANCAIS CONFERENCE OVERVIEW PRE-CONFERENCE EVENTS SPONSORSHIP + EXHIBITION MEDIA/PRESS REGISTRATION, HOUSING, TRAVEL HELP PROMOTE INET'98
|
ReportsThese reports, written by volunteers,
summarise information for people not able to attend the sessions. Their
comprehensiveness and accuracy are not guaranteed. For more information,
please contact the presenters directly. Their e-mail addresses are available
at http://www.isoc.org/inet98/program.shtml Track 1: New Applications Panel: Can standards survive the success of the Internet? By Julian Albert Kilker, 22 July 1998 This review summarises the panel's responses to the questions of the moderator Ken
Klingenstein. Members of the panel were Fred Baker of Cisco Systems, Brian Carpenter of
IBM (replacing Erik Huizer, who was not able to attend), Scott Bradner of Harvard
University, and Robert Shaw of the ITU. Ken began by asking for a brief history of the Internet standards making process. Panel
participants noted that standards are created by consensus; there must be enough consensus
to build, test, and iterate through a standard. The process has changed only slightly
since the beginning of the net. At present, it is more difficult to get to the RFC stage
when a proposed standard is published for discussion. This is both because standards are
much more complex (the example of router standards was given), and because there are many
more players in the process, and thus more "squeaky wheels" to accommodate.
While the standard processes have been generally successful, members of the panel raised
concerns about the broadening attention given to standards making, and that developers
might end up "trying to satisfy constituencies that [they] shouldn't be trying to
satisfy." In particular, Brian noted that standards organisations do not appreciate
being handed "ready-made" standards by outside interests. When asked whether the speed of standards development has changed, panel members argued
that it is about the same; the differentiated services process suggests that the process
can be quite rapid. The panel noted several problems currently facing standards making: the heavy workloads
of people involved in the process, which particularly affects the evaluation of complex
standards requiring detailed analysis (such as security standards) and reduces people's
participation in the many stages of the process. There have been cases, for example, where
it would have been helpful if people had attended IETF meetings before commenting on its
mailing lists. Introducing new people involved in the standards process to the 25-year
history and social norms of net standards is difficult. Additional problems included
keeping in touch with the many standards groups now involved at some level (or wishing to
become involved) in standards relevant to the Internet, bringing companies productively
into standards development (panel members noted that Microsoft has recently understood the benefits of collaborative approaches to standards
development), the increasing politicisation of standards development because of the
Internet's success, and the associated concerns that technical standards will increasingly
be seen as a way to restrict rather than support internet working. The overall point was that the nature of the Internet has influenced the standards
development processes. Fred noted that standards development must be collaborative because
"no one vendor supplies everything, and no one has all the best ideas". Scott
noted that, with respect to standards development, the "power of the Internet is that
you can experiment", that the apparent chaos of the net is its greatest advantage. |
About the Internet Society | Conference Overview | Pre-Conference Events | Sponsorship + Exhibition |
Registration, Housing, Travel | Media/Press | FAQs | Home
All graphics and content copyright ©1998
Internet Society.
The INET'98 Web-site was designed by
Designright, Inc. http://www.designright.com