I would like to reiterate my opposition to four requirements of this
draft (as described in my mail of 18 July):

 - the requirement for a new ND option;
 - the requirement for a RESTful management API;
 - the requirement for a local DNS resolver on each link;
 - the requirement for a ULA.

Speaking as the author of one of only two implementations of HNCP, I would
like to be very clear that I shall not implement a RESTful management API,
and I shall not implement a local DNS resolver.  I will also not make it
compulsory to deploy a ULA, although I'm willing to make it the default
behaviour.  (I have no objection to implementing the ND option, assuming
that somebody else gets it deployed in host implementations.)

I believe that it would be premature to adopt this draft now, at least
until people have had a chance to consider the consequences for
implementations.  The best way IMO would be to have a prototype implementation,
which I will be glad to review.

I shall not make too much of a fuss should the WG decide to adopt this
draft notwithstanding my objections.  However, I reserve the right to say
"I told you so" when we eventually run into trouble.

-- Juliusz

P.S. Our mail server is down until Thursday, please reply to the list.

_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to