From: Keith Moore <moore(_at_)cs(_dot_)utk(_dot_)edu>
... The IETF, through the processes
defined in 2026 and other documents, has change control. ...
Ok, then let's revoke the changes made in the Pacific Northwest to the
purely IETF protocol, PPP, starting with MS-CHAP and DNS-in-LCP. Others
can offer other examples of changes to purely IETF protocols from the same
and other industry leading vendors that bypassed the IETF's change control
and that many of us agree should be reconsidered.
I support the jihad against NAT boxes (although there are much worse such
as redirecting proxies). However, advocating or hoping to deal with NAT
or anything else by appealing to the IETF's change control authority is
worse than a distraction, because people outside The Standards Process
see it as obviously silly and useless. People remember that the IETF
started as a finger in the eye of the folks who had de jure and de facto
change control for network protocols and Architecture.
The only real change control of the IETF or any standards outfit is the
ability to offer something somehow, not necessarily technically better.
Pointing out an evil must be only a preamble to describing a better
alternative, or those who say they are "implementing NAT" will quite
rightly continue pointing and clicking their way through GUI's.
Discouraging the use of a bad thing without offering an alternative is
not just a waste of time, but a "crying of wolf" that reduces the IETF's
ability to control change.
I guess the fix for NAT is IPv6, so let's all drink to 128 bits
but let's *please* not worry about the change control authority
that exists only among standards committees.
Vernon Schryver vjs(_at_)rhyolite(_dot_)com