Re: [homenet] I-D Action: draft-ietf-homenet-hncp-bis-00.txt

[Ralph Droms <[e-mail]>]

> On Jul 8, 2016, at 12:30 PM 7/8/16, Ray Bellis <[e-mail]> wrote:
> 
> 
> 
> On 08/07/2016 17:25, Ralph Droms wrote:
>> I took a quick look at draft-ietf-homenet-hncp-bis-00, including a
>> diff; seems the only change is to solve the .home problem.  I don't
>> think I quite understand the new text and here's a suggested
>> clarification:
>> 
>> OLD:
>> 
>> A default value for this TLV MUST be set, although the default value
>> of the Domain-Name TLV (Section 10.6) is out of scope of this
>> document, and an administrator MAY configure the announcement of a
>> Domain-Name TLV for the network.
>> 
>> NEW:
>> 
>> The administrator MUST configure the announcement of a network-wide
>> zone suffix through the Domain-Name TLV.
>> 
>> As far as I can tell, draft-ietf-homenet-hncp-bis-00 does not address
>> errata 4718.  While this errata has not yet been verified, in my
>> opinion *something* has to be done to correct the text around
>> "Multicast DNS Proxy".  If "Multicast DNS Proxy" is intended to refer
>> to "Hybrid Proxy" in draft-ietf-dnssd-hybrid-03, the appropriate
>> normative reference will constitute a downref in
>> draft-ietf-homenet-hncp-bis-00.
> 
> The instruction to the authors were to incorporate the original .home
> errata verbatim, and also to fix the error with options 37/38 being the
> wrong way around in two diagrams.

I honestly can't understand the new text.  Why a friendly suggestion for a simplification need AD intervention?

> 
> If there is to be further wordsmithing on that we'd need to take that up
> with our AD.
> 
> As for errata 4718, I fully expect that it will be incorporated in a
> further revision just as soon as it has been verified (and subject to a
> resolution of any resulting downref issue).

OK. As an aside, *something* will have to be done with the text regarding "Multicast DNS Proxy" regardless of whether or not the errata is verified.  I see a couple of ways to read the existing text and the citations of RFC 6762.

One way to read RFC 7788 is to assume "Multicast DNS Proxy" refers to the "Multicast DNS Proxy Servers" defined in RFC 6762, in which case RFC 7788 is specifying that an HNCP device should participate in whatever election protocol "Multicast DNS Proxy Servers" use to elect the proxy server for a link.  That election protocol needs a reference

Another way to read RFC 7788 is that there is no citation for a definition of "Multicast DNS Proxy" at all (assuming the citations of RFC 6762 apply just to mDNS and not "proxy"), in which case RFC 7788 needs to be amended to include that definition of "Multicast DNS Proxy".

Seems like the WG might want to go ahead and figure out which way it wants to fix this issue as doing nothing isn't an option.

- Ralph

> 
> kind regards,
> 
> Ray
>