A question about the role of the IAB

RJ Atkinson rja at extremenetworks.com
Wed Jan 8 11:42:22 CET 2003


On Wednesday, Jan 8, 2003, at 11:26 America/Montreal, Brian E Carpenter 
wrote:
> Assuming this means RFCs, neither committee has *editorial* control.
> That lies with the RFC Editor. But again the separation of powers is
> in fact clear in reality: the IAB oversees the RFC editor contract
> and statement of work (formally signed, and funded, by ISOC),

yes.

> and the  IESG approves (or disapproves) individual documents.

Precisely true for documents that (originate in an IETF WG or
are on the IETF standards-track) only.

For non-standards-track documents that are individual submissions
(and not from IAB, IRTF), then RFC Editor makes the final decision on
whether to publish.  IESG may advise against publication of such an RFC.
IESG *may* insert an IESG Comment into such a document if it has 
concerns
about a document and the RFC Editor decides to publish the document
anyway.

There is a more clear documentation of how IAB documents are published
(either in preparation or recently published).  The IRTF publication
process ought to be more clearly documented than it seems to be, but
new documents go via IRTF Chair and RFC-Editor rather than direct to
the IESG.

AFAIK in all cases, IAB/IRTF documents are circulated for collegial 
review,
prior to publication.  So IESG ought not ever be surprised by some RFC 
that
appears.  In my experience, IAB documents have always benefited from 
such
review.  I'm not aware of any IAB documents that have been published and
caused severe anxiety within the IESG as a whole in their final 
as-published
form/content.

> RFC 2850 is certainly not perfect, but I have trouble seeing how
> it could be significantly more precise, given the nature of the IAB's
> job.

I agree with that.  The current IAB is certainly interested in hearing
any specific suggestions that folks have with the current IAB Charter.
Those comments could reasonably be sent to the IAB directly if one
wished -- or posted here if the WG Chairs think it is in scope here.

Ran



More information about the Problem-statement mailing list