On Wed, 12 Nov 2008, Iljitsch van Beijnum wrote:
In my opinion, this is a bug. This is the default policy table from a FreeBSD
system, which is the RFC 3484 table IIRC:
You should probably bring this up on 6MAN WG list then.
Prefix Prec Label Use
: : 1/128 50 0 0
: :/0 40 1 646064
2002::/16 30 2 0
: :/96 20 3 0
: : ffff:0.0.0.0/96 10 4 0
The last line catches IPv4. It's two steps below the 6to4 prefix. However,
the fact that the label for the 6to4 prefix doesn't match the ::/0 label
means that IPv4 will be used.
This happens on FreeBSD and XP, and I assume also on Vista. But not on MacOS,
because it doesn't implement the policy table. I don't know about Linux. (If
you want to test, try to connect to 6to4test.runningipv6.net and see what
happens. Both addresses are unreachable, though.)
I'm not sure whether you're agreeing with me or something else; I
don't see where you're saying the bug is. But if we start talking
about issues in RFC3484, it should happen on 6MAN list.
Your test is inconclusive due to the fact that the A record is a
private address. Depending on whether the connecting host has a
global or private address, the results are different (see my mail to
Remi for more).
Pekka Savola "You each name yourselves king, yet the
Netcore Oy kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings
Ietf mailing list