On 2013-05-28, at 3:38, SM <sm(_at_)resistor(_dot_)net> wrote:
In theory the IETF does not publish RFCs to suit the regulations of one
country (see use-case in draft-jabley-dnsext-eui48-eui64-rrtypes-04). In
practice, the IETF has published a RFC to suit the requirements (it was a
voluntary measure instead of a formal requirement) of one country.
Note that there's no suggestion that these RRTypes are required by the
CRTC. The example given was for a situation where Interop would have
been beneficial (so that cable resellers have an obvious, stable and
supported way of encoding this kind of information.
draft-jabley-dnsext-eui48-eui64-rrtypes-04 is an odd case. My guess is that
the requirements were set because of a problem of monopoly.
The opposite actually: cable operators are required to provide access
to subscribers on behalf of third parties in order to promote
competition. There are multiple such cable providers and multiple such
resellers.
(TekSavvy is one such reseller of multiple cable companies' access networks.)
I have not looked into whether the transfer of data violates the expectations
of the user.
The CRTC's decisions are public, and one might hope they would show
their working. The Canadian courts have been pretty protective of user
privacy in general, I would say.
I understand that the draft is about standardizing [1] a data format and not
the transfer of data. Section 8 of the draft says everything correctly
except that it doesn't provide adequate security guidance.
Feel free to point out the gaps, and/or to suggest text.
I believe that Joe tried to do the "right thing". I am not comfortable
objecting to publication as I don't know the "path forward". I personally
would not support publication. That can easily be overcome and I won't do
anything about it.
Thanks for the thoughtful review.
Joe