--On Saturday, January 10, 2009 12:52 -0800 Dave CROCKER
<dhc2@xxxxxxxxxxxx> wrote:
>...
> I can't begin to guess at the logic that uses Larry's somewhat
> bizarre assertion as a basis for trying to press approval of
> this clearly and substantially problematic proposal.
>
> To create a paraphrase, what part of "neither short term nor a
> fix" did you not understand?
>...
With the understanding that this is _not_ a fix to any
short-term problem, I think I've learned something from the IPR
WG process and its "too few people really understood the
documents and their implications" consensus and the resulting
IETF consensus that produced 5378. That lesson goes beyond "we
don't evaluate legal language at all well" and into some defects
in how we approve procedural documents.
Independent of the much more controversial question of which
body should be approving these things and under what process, I
note that we require "IANA Considerations" sections for
technical documents so that IANA can check to be sure that the
instructions make sense and are consistent with other registries
and practices. It is now clear to me that any document that has
procedural, administrative, or legal provisions or consequences
should be required to contain:
(i) Explicit "IASA Considerations" and/or "IETF Trust
Considerations" sections, as appropriate.
(ii) A requirement for explicit signoff, or at least
formal comments, from those bodies before the IESG makes
a decision about IETF consensus.
While it isn't clear that the Trustees would have identified
these implementation / transition problems during this review,
giving them the opportunity and obligation to spot
implementation problems --and to think through how they would
implement a particular set of requirements before those
requirements are cast in stone-- would have at least eliminated
the "not our problem; we are just doing what we were told"
attitude and comments that resulted in our looking at proposals
in January rather than before the holidays.
The general idea would be to explicitly put the Trustees (or the
IAOC for other situations) into the position of being able to
say "nice idea, but it won't work legally and/or procedurally",
just as IANA can say "we have read that specification and have
no idea how to implement it in practice"... and to give them the
obligation to review a document, pre-approval, and decide
whether to provide that advice.
I believe that the IESG can impose these requirements and the
additional checking/validation cycle on its own initiative,
based on their having imposed the "IANA Considerations" and
others on their own initiative. This is a call for that
initiative or, if it is not forthcoming, for others to think
enough about the issue to determination of whether collaboration
on an I-D would be appropriate.
It doesn't need to be done this week, but it appears to me to be
something we had better have in place before anyone starts
considering 5378bis (or a forked 3978ter).
john
_______________________________________________
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf