Re: [core] DNS-SD service types for CoRE-RD (Re: AD review of draft-ietf-core-resource-directory-23)

[Christian Amsüss <[e-mail]>]

Hello Ted, Carsten and group,

thanks for your input; I take from it that DNS-SD discovered secure RD
access is realistic, and that while there are several ways to use it,
those are not specific to RD and thus we don't have specific security
guidance.

For the common usage of RD that I personally expect, UDP will be the
predominant usage pattern, and the other protocols could still be
entered into the registry when actually needed -- that would make the
whole thing simpler here as it's only a single new type. Text proposals
for both versions (coap/coaps/coap+tcp/coaps+tcp[1,2,3] and
coap-only[4,5,6]) are currently in the editors' copy branches, diffs and
issue tracker.

Given that what was previously discussed in the WG all points to
many-services, I'll merge and submit the four-protocol variety this
evening before cut-off, but if there are seconds for the argument for
simplicity in limited initial choice (with the option to extend without
RFC through IANA registration), I'd be happy to go with that.

Kind regards
Christian

[1]: https://core-wg.github.io/resource-directory/dnssd-moreprotos/draft-ietf-core-resource-directory.html
[2]: https://github.com/core-wg/resource-directory/pull/230/files
[3]: https://github.com/core-wg/resource-directory/pull/230
[4]: https://core-wg.github.io/resource-directory/dnssd-nomoreprotos/draft-ietf-core-resource-directory.html
[5]: https://github.com/core-wg/resource-directory/pull/231/files
[6]: https://github.com/core-wg/resource-directory/pull/231

-- 
To use raw power is to make yourself infinitely vulnerable to greater powers.
  -- Bene Gesserit axiom