RE: Please review draft-carroll-dynmobileip-cdma-04.txt

["Nelson, David" <dnelson@enterasys.com>]

Thomas Narten writes...

> It describes what has been deployed. The authors were never willing to
> make changes that impacted the on-the-wire behavior.

I guess the horse was stolen quite some time ago, and we're still here
quibbling about locking the barn.  :-)

As I understand it, the IESG is being asked whether the subject draft is
"so bad" as to request the RFC Editor to block publication. The AAA
Doctors are being asked to comment on "how bad is bad?".

I stand by my earlier comment that the RFC series should never be used
to undermine the global interoperability of Standards Track RFCs.
Protocol extensions are fine, but normative requirements violations are
not, IMHO.

Having said that, since this protocol has been designed, implemented and
deployed widely without anyone in the IETF (apparently) being asked to
comment on the issues of RADIUS RFC compliance, we are left with the
decision of whether to document "what got implemented" or to leave it
undocumented (or documented someplace other than in the RFC series).
Failing to describe the implementation, if it is widely deployed, is not
particularly productive.  Just as we don't rely on "security by
obscurity" we ought not to expect "protocol quality by obscurity".  :-)

OTOH, publishing this protocol "extension" in an RFC will give it
something of the imprimatur of the IETF/IESG/ISOC.  Adding a protocol
quality disclaimer to the front matter will help somewhat, but not
everyone really understands the implications of those notes.

If there is no disincentive to publishing poor quality "extensions" to
Standards Track protocols, other than an extended IESG review cycle and
a protocol quality disclaimer, are we not sending a mixed message?

-- Dave