On Tue, 20 Apr 2010, Tore Anderson wrote:
* Mohacsi JanosYes, that the effect of the change proposed in draft-arifumi-6man-rfc3484-revise-02, and also how Microsoft's getaddrinfo() currently behaves. I strongly support the proposed change. RFC 3484, however, currently says to prefer 6to4 over IPv4 NAT44 (or more precisely RFC 1918 IPv4).However all the existing implemention prefers IPv4 with RFC1918 addresses.That is not the case. Mac OS X and GNU libc+derivates are two prominent examples of getaddrinfo()/RFC 3484 implementations that follow the RFC to the letter. They prefer 6to4 over (RFC 1918) IPv4. That's what causing most of the client loss in my dual-stack measurements. Best regards,
Mac OS X does not implement RFC 3484.Linux implementation is not tested by me a while ago, but one of my college did and confirmed no problem. Maybe he made a mistake in the test. It seems to me
http://people.redhat.com/drepper/linux-rfc3484.html does not refer to this problem. I try to inform linux RFC3484 implementers about the problem. Best Regards, Janos Mohacsi