POISED '95 BOF (POISED95)

Reported by Paul Mockapetris/Information Sciences Institute

Three sessions of the POISED '95 BOF were held at the IETF in Danvers.

The Monday session, chaired by Paul Mockapetris, sought to create a list
of issues.  The Tuesday session, chaired by Carl Malamud, was limited to
non-IPR issues, but in fact concentrated on the NOMCOM and its
procedures.  The Wednesday session, chaired by John Curran, concentrated
on the reform of RFC 1602 regarding Intellectual Property Rights (IPR)
such as patents, copyrights, etc.  The object was to create new
procedures which would replace the enduring fun occasioned by the Sun
RPC/XDR episode.



ISOC/IETF Relationship

Most felt that the ISOC is the only choice at the moment to execute
legal agreements, but most agreed that progress on these issues was too
slow and RFC 1602 is broken.  The Wednesday session on RFC 1602
intellectual property rights will address this.

The insurance coverage is still uncertain, the aggregate coverage is
about one million dollars, and as of the meeting, covers the ISOC Board
of Trustees, IESG, but not the working group chairs, widely seen as
involved in progressing standards.  The ISOC is progressing the issue.

At least one person is concerned that the ISOC is competing for research
grants and may come into conflict with certain IETF members seeking the
same grants.  At present no conflict exists, and most of the meeting
participants felt that this was not an issue.

The ISOC has proposed support of IETF, and the number 250 thousand
dollars has been mentioned, but no funds are available to the IESG,
Secretariat, or IETF. The ISOC Board of Trustees may act.  A partial
dream list for spending such funds is:  expenses related to startup of
the ``hostless IETF'' meeting scheme, legal advice for the IESG/IETF
(ISOC did fund legal talent at the BOF), funds to create standards
tracking systems, professional editing help, and development of
technology to support multimedia (i.e., non-ASCII text) standards.  A
lot of suggestions started to emerge about the use of the Web, but this
was ruled out of order for POISED.

The ISOC, through the IAB, sponsors and will create new groups that are
not open, but are self-selecting.  Some feel this is appropriate, some
not.  Most did feel that if the IETF is held to a standard, so should
other groups.

Ownership of the IETF, the address space, etc., seems to be a continuing
issue.


Funding Models

There is a general request for information about the IETF and
Secretariat financing.

In the future, some feel that integrated funding for the IANA, IESG,
InterNIC, etc., is desirable.  There are various opinions about how this
should happen.

One prominent question was non-US support.  Some feel that financial
support will not be forthcoming until an acceptable framework, including
reporting and control to some degree, is set up.  Others say that until
there is money on the table, why set up the framework?  A bit of a
chicken and egg problem?



Suggestions for POISED '95 IPR

The discussion regarding the repairs necessary to RFC 1602 to deal with
intellectual property rights highlighted several issues:

A single cookie-cutter agreement, particularly one that was as
unfavorable to the technology donors as the current RFC 1602, is
unworkable.  In particular, no business is seen as willing to warrant
that all technology is free of patent problems, especially since such
problems may be unknown at the time the technology is donated.  It is
clear that we need some combination of:


   o Model agreements for a company that wants to hand over ownership of
     a technology to the IETF. The Sun RPC/XDR agreement may be a good
     start here.  In general, this can be both good for the IETF,
     because it acquires a useful technology, and for the company, since
     it will presumably sees its investment and leadership in the
     donated technology rewarded in the marketplace.

   o A simpler agreement whereby the holder of a patent does not give it
     over to the IETF (i.e., change control), but agrees that the IETF
     can use the technology to implement some larger goal.

   o Somebody empowered to negotiate variants.  The general idea seemed
     to be that as we built up a history of agreements, any new
     agreement could follow precedent, but where precedent was lacking,
     someone had to be empowered to negotiate.  Clearly, the
     desirability of the technology, availability of alternatives, etc.,
     would have to be taken into account.  There was a vigorous debate
     about whether licensing fees were appropriate, and what form they
     should take.  However, a strong preference for unencumbered
     technology should become a stated objective of the IETF.

   o The question of who is party to the agreement was discussed.  Some
     feel that unilateral grants would be an effective mechanism, but
     others noted that somebody would still have to take the
     responsibility for deciding whether the terms of a unilateral grant
     were acceptable when it did not follow a clear precedent.


(There is some subsequent work from the IAB/IESG to develop an ``escape
clause'' procedure, or a ``procedure for variation negotiation.'')


Possible Improvements To The Standards Process

Several people believe that once an Internet-Draft or other matter
enters into the hands of the RFC Editor, InterNIC, regional registry,
IANA, IESG, or IETF Secretariat, it can be difficult for an ``outsider''
to determine where it sits, and who has the responsibility for
``progressing'' the matter.  The chair mentioned that the Secretariat
was working on developing a Web-accessible tracking system for IETF
documents, that would provide such tracking for IETF documents.  The
other processes are probably beyond the control of POISED.

Openness in the actions of the IETF, IESG, InterNIC, are seen as
essential.

There seems to be a need to create a new type of standard, which some
call a ``policy'' standard.  This would be used for things like the
RFC 1602 follow-on, address allocation procedures, etc.  Opinions
differed on the best way to proceed.


IETF Meeting Site Selection and Criteria

What are the criteria for an IETF site, and how should sites be
selected?

As the IETF has grown, the difficulty of getting meeting sponsors has
increased.  Various estimates suggest that Sun and FTP Software spent
50-75 thousand dollars or more on the most recent IETF meeting for the
terminal rooms, MBONE support, etc.  The effort and hassle are also
daunting.  As a result, future IETF meetings may be ``hostless,'' with
meeting fees going up to cover the expense of hiring the terminal room,
etc., out.  This is seen as acceptable, and will probably happen in
1996.

The criteria for a meeting site were seen as:


   o A state of the art terminal room (X-terms, appletalk, Ethernet)
   o MBONE coverage for multiple parallel sessions
   o Capability to house about 1000 participants nearby (not bussed)
   o A reasonable airport nearby


Minority opinions included requirements that meetings be scheduled two
years in advance, that the spread of laptops meant a declining need for
X-terminals and the like, and that geographical diversity should be
given more weight.  The current Secretariat was seen as doing a good
job, but there is always a possibility for improvement.


NOMCOM

There is a general need to finish formulating the NOMCOM procedures.
Several NOMCOM members and even past chairs felt that better guidelines
would be welcome.

One issue is ``tailoring'' the NOMCOM membership.  It was observed that
the membership of the three NOMCOMs seemed to be fugitives from the law
of averages.  This is due to the small ``gene pool'' of volunteers.  It
was recommended that the IETF Chair and the IETF in general attempt to
broaden the gene pool by aggressive attempts to get people to volunteer.
Another method of tailoring is to prohibit participation by candidates
for office, sitting members of the IAB/IESG, members of the prior
NOMCOM, or various other criteria.

There was also debate about whether nominations, candidacy, etc., should
be kept secret or not during the NOMCOM process, and who should announce
results.  Consistency is not a hallmark of past NOMCOMs.


Other

One participant mused for the good old days when the IESG was a
tight-knit team, and believed that the present IESG was not this
Camelot.  More cynical types discussed the tradeoffs between IESG as
Borg vs.  IESG as a bunch of Dirty Harrietts/Harrys.  The truth is
probably somewhere in between.