Alexandru,
I am for adopting this draft as WG document.
Great!
It says that RH types should be allocated through "IETF Review" or "IESG Approval" - these methods seem to me the most appropriated because offer the widest possible review through IETF - IESG and WG. This is needed knowing that RH has large importance in the sequence of headers.
Right, particularly if we get it wrong and a security hole opens up ;-)
The other methods are "Private Use", "Experimental Use", "Hierararchical Allocation", "FCFS", "Expert Review", "Spec Req'd", "RFC Req'd", "Standards Action" (rfc5226). The first four seem not adapted for such an important header while the latter three seem too much work for such simple action as properly reserving a number.
Yes, though RFC 4727 does already allocate experimental use values for routing header types. I view these as a useful thing, allowing people to try out weird things with numbers set aside for such a purpose. RFC 3692 explains how such numbers need to be used carefully to avoid breaking other things on the Internet.
And yes, Standards Action is too strong requirement for this type of a method. We might want to publish an experimental specification that uses a new routing header type, for instance.
Jari -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
