[ih] Protocol numbers (was IP version 7)

[Brian E Carpenter]

On 24-Dec-20 08:20, Jack Haverty via Internet-history wrote:
...
> From a larger view, IMHO the Internet "governance" (IANA et al) have
> released thousands of RFCs and assigned hundreds of numbers, but don't
> seem to have any mechanisms to track how such things, i.e., numbers or
> protocols, are actually used in the operational Internet.  So it's not
> clear to me how to "release" anything....

It's impossible in practice to assert that a given feature isn't used
any more. How can you determine that protocols 15, 32 or 125 are
*not* used in some remote corner of the netiverse? This is a Good Thing
if you believe in privacy and dislike arbitrary surveillance. So IANA
and the IETF have always been very conservative about releasing magic
numbers for potential re-use. The problem has been most acute for system
ports; iirc that's the only case where the IETF has found real cause
for concern.

Even though the protocol number space might appear to be relatively
full, I think we're a long way from a problem. But if you think there's
a problem, write a draft for the IETF to discuss. Certainly people on
this list could help in any effort to identify obsolete assignments,
but instructions to IANA have to come from the IESG on behalf of
the IETF.

(I thought about writing such a draft a few years ago, when somebody
wrote to me out of the blue enquiring about the status of protocol 32,
probably because I am an author of RFC7045. But there seemed to be
no interest at that time in doing the work.)

Regards
    Brian