Le 3 août 09 à 22:14, Christian Huitema a écrit :
In view of the various arguments made, here is IMHO a good
combination:
- IPv6 hosts MUST create UDP datagrams with non-zero checksums.
(Nothing new here.)
- IPv6 hosts MAY accept UDP datagrams with zero checksum.
((1) Application of the classic principle "be strict in what you
send, tolerant in what you receive"; (2) useful for some UDP
datagrams that cross some v4-v6 translators.)
Except you do not get to specify that as part of the translator
specification.
The translator spec applies to translators, not to hosts.
Agreed.
Translators must assume that IPv6 UDP packets with null checksum
will be dropped by the receiving host.
Different understanding here.
A better sentence IMHO would be "Translators MAY assume that IPv6 UDP
packets with null checksum are accepted by some receiving hosts (or
will be in the future)."
For a translator, forwarding such a packet with a null checksum is
pretty much the same as dropping it.
If all destination hosts drop them, the effect is indeed the same.
But if some hosts start being open to receive them (a backward
compatible extension), then it makes a useful difference.
The proposal is a combination of backward compatible extensions in
hosts and translators:
- They make no harm in general.
- They are useful when used in combination.
Regards,
RD
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
