> >> Given the number of remaining entries, the task of garbage collection > seems of little value. > > Until the day when it seems urgent... > 109 available out of 256, or 42%.
The catch is that when this breaks, it will break lots of big stuff. This 1-byte value is shared between IPv4 and IPv6(!), so exhausting it in one will also exhaust the other. And it occurs in every IP packet, describing the protocol of what it contains (e.g. UDP or TCP, or an IPv6 packet, or an IPv6-Fragment header). Or, perhaps the IP packet contains ISO-TP4, MERIT Internodal Protocol, the "Cross Net Debugger", or "TRUNK-2" from the ARPANET days. That one-of-a-kind stuff really should get cleaned out. I have asked several people who allocated ancient entries to determine if those early protocols are dead, and if so, to formally relinquish the allocation via email to IANA. You'll see several marked as "deprecated" which have resulted from those asks. Some of those people, however, are dead now, and there's no obvious way to clean up the detritis they left lying around in the IP protocol. For example, I have an ask pending to Barry Boehm, who seems to only have been the formal owner (as head of the DARPA IPTO office) and not the actual user (those entries were allocated by Barry Leiner, then moved to Saul Amarel, then to Barry). And I haven't heard back. He has four, almost certainly obsolete, protocol numbers, and IANA doesn't even seem to have his current email address ([email protected]? But he appears to be at USC now.). John _______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
