Patrik,
On May 29, 2014, at 2:41 AM, Patrik Fältström <paf(_at_)frobbit(_dot_)se> wrote:
Just for clarity, the root server operators are under no obligation to do
anything.
But that is not the problem of the IETF.
True, which is sort of my point.
If the IETF come to conclusion that root server services "must" support IPv4
and IPv6, so be it. It should be in the RFC.
BCP177 would appear to cover the IETF's interest in ensuring both IPv4 and IPv6
is supported.
Policing will not happen without a spec that services can be compared against.
My impression is that the IETF hasn't been particularly effective in coming up
with operational service specifications.
And lack of policing (which seems to be what you talk about) is I think a
separate issue.
That's not really what I'm talking about. What I am saying is that it I don't
really see the point in the IETF attempted to demand stuff outside of its
control. In most standardization/protocol definition contexts, the IETF
specifying "MUST" or "MUST NOT" usually makes sense since folks generally have
a choice in obtaining equipment/software/service that is implementing those
standards.
This is NOT the case with the root servers.
By and large, the Internet community gets what the root server operators deem
at their sole discretion (perhaps informed by outside-the-IETF contractual or
other obligations) within their interests to provide, nothing more and nothing
less.
In April 2012, the IETF, via BCP177, already stated "IPv6 is required" yet B,
E, and G still do not support IPv6 (yes, I'm sure there are reasons, that's not
the point). RFC 2010 was published in 1996 and was mostly ignored by the root
server operators. RFC 2870 was published in 2000 and was mostly ignored by the
root server operators. What has changed that makes you believe a new RFC is
going to have a different impact?
Regards,
-drc
signature.asc
Description: Message signed with OpenPGP using GPGMail