Re: The Trust and licensing [Re: New version of my proposal posted]

[John C Klensin <john-ietf at jck.com>]

--On Friday, 04 November, 2005 20:57 +0100 Brian E Carpenter
<brc at zurich.ibm.com> wrote:

>...
> b) it's my personal intention to insist that the Trust's
> procedures
> include a commitment to follow the IETF's wishes in matters
> of licensing.
>...

This intention would ease several concerns.  However, section
9.5 of the trust agreement contains provisions that appear to me
to be equivalent to those of the GPL copyleft only retaining the
rights for the Trust rather than the public.  Specifically, the
Trust cannot license or assign any rights in Trust-owned IPR,
including the right to to create derivative works, unless the
licensee agrees that those derivative works will belong entirely
to the Trust.  In addition, not being a lawyer, I'm not certain
what 9.5(b) means in practice, but it appears to mean that, if a
licensee were to marginally profit as a consequence of using the
IETF's IPR, the Trust would have a claim on a share of those
profits.

It seems to me that this raises two issues that have been hotly
debated in the past in another context.  The restriction would
appear to require individual negotiation of licenses, not a
blanket grant, which some members of the community have argued
is too onerous for other types of IPR.  It also creates a "what
is mine is mine and what is yours is mine too" condition, to
which many organizations have taken strong exception in the past.

I note that this working group was not consulted on that
language or on the appropriateness of these licensing
constraints.

I hope we can avoid re-debating either of those substantive
issues here.  But I  think it is important to point out that,
should the IETF change its mind in this matter, or reach a
conclusion about terms different from those specified in the
trust agreement, and do so by some mechanism that clearly
reflected community consensus:

(1) The IAOC, in addition to or instead of the IESG, would need
to ratify that consensus.  While I agree with Brian's
observation that 

> 	a) the IETF can in fact fire most of the Trustees if
> 	they misbehave, and I think that ignoring an IETF or
> 	IESG consensus on licensing would be a firing offense;

the "running code" is that the community has found "firing" to
be a much too difficult and unpleasant procedure to deal with
any but the most obnoxious of behavior.  It is not yet
demonstrated that behavior that obnoxious is even possible
because the mechanisms have never been exercised at all.   If an
IAOC concluded that it simply knew better than the community
consensus, perhaps on the basis of information available to them
and not the community, experience so far would lead them to
reasonably predict that recall/ firing is not a sufficient
threat to prevent following their judgment rather than the
community's.  

More generally, if we have designed a procedure for which we
need to use the threat of recall as the only mechanism to ensure
that an IETF consensus on licensing is followed, we are probably
in bad trouble.  I suggest that, if the IETF believes that IETF
consensus, especially on the policies to be followed by the
Trustees, should be binding on the IAOC and those Trustees, the
IETF needs to insist on that as a requirement.

Unfortunately, under the Trust Agreement as I read it, the IETF
doesn't get to impose that requirement, as least not as an IETF
action.  

(2) Any changes to the basic licensing requirements or rules of
the Trust, such as changes to the licensing provisions of 9.5
described, should the IETF decide to make them, require the
consent of both CNRI and ISOC unless we are prepared to wait
until July of 2010.  These are requirements of the Trust
agreement and CRNI and ISOC must agree to any amendments to that
agreement.   If they don't agree, they don't need to give
reasons.  They are not even obligated to indicate whether they
agree or not within any specified period of time.  Should one of
them not agree, then, as the Trust Agreement is written, IETF
consensus is worth exactly nothing.   And there is no mechanism
for "firing" either of the Settlors, at least without the
consent of both of them.

As I have communicated privately (and in much more detail) to
the IAOC, I believe that the Trust Agreement sets policies for
the IETF and gives away decision-making authority over which the
IETF has believed that it _as a community_ had consensus
control.  Because it involves setting policies, rather than
"managing" the IPR portfolio, I don't believe the IAOC has the
right/authority to do any such thing without explicit community
consensus.   Whether their doing so is a matter that would be
considered equivalent to one of the activities for which Brian
believes people should be "fired" is something that he should,
IMO, comment on.  And it is one on which the community should
reach some conclusion... a conclusion that I believe is
ultimately about whether IETF decision-making, at the most basic
of policy levels, is bottom up or top-down, controlled by the
community or controlled by a few of its appointed leaders acting
on their own judgment rather than on community consensus.

    john


_______________________________________________
Ipr-wg mailing list
Ipr-wg at ietf.org
https://www1.ietf.org/mailman/listinfo/ipr-wg