At 11:58 PM +0200 11/25/07, Jari Arkko wrote:
Paul,
They still should (strongly) consider checking the validity of the XML
by comparing it to what the IESG approved.
Yes, and they do compare to what IESG approved. Substantial changes are
brought to the AD's approval. This is what caused us to find the problem
in this case.
I'm confused. Why should the RFC Editor accept XML with any
substantial changes? That's inherently prone to error. They should
start with what was approved.
(But note drafts are often approved with some remaining "Comments" from
the IESG review. It is up to the AD in charge and the authors to
determine whether those should be addressed. And that could happen by
submitting a new draft version before the approval announcement is sent,
with RFC Editor notes, in AUTH48, or perhaps even by tweaks in the XML
file. In all cases the AD should be checking that no inappropriate
changes are being done.)
The "tweaks in the XML file" seems like a Really Bad Idea. I have
never seen IESG comments that would take the RFC Editor more than
half an hour to incorporate by hand from the IESG announcement.
--Paul Hoffman, Director
--VPN Consortium
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf