In your letter dated Tue, 19 Jul 2011 09:36:05 -0700 you wrote: >On 7/19/11 8:51 AM, Philip Homburg wrote: >> How can it be persistently incorrect? > >Depends whether it is truly random, or pseudo-random, or random initial >value plus an monotonic increment.
Truly random sounds a bit expensive for a translator. Yes, pseudo-random is what you want. And any kind of counter, no matter how it is initialized, doesn't count as random. >RFC 2460 doesn't use MUST/SHOULD/MAY RFC 2119 language, hence it is a >bit hard to gauge the strength of its text. >RFC 2460 does contain this text: > * "recently" means within the maximum likely lifetime of a packet, > including transit time from source to destination and time spent > awaiting reassembly with other fragments of the same packet. > However, it is not required that a source node know the maximum > packet lifetime. Rather, it is assumed that the requirement can > be met by maintaining the Identification value as a simple, 32- > bit, "wrap-around" counter, incremented each time a packet must > be fragmented I guess the most fun can be had if a host implements a counter, but then accidentally writes out the counter as a little endian value. But I guess that is easy to solve. Harder to solve would be the case where the IPv6 host is actually a busy server, say a DNS server, that doesn't want to keep any state and just has the link MTU set to 1280. -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
