Re: [DNSOP] I-D Action: draft-ietf-dnsop-attrleaf-12.txt

[Dave Crocker <[e-mail]>]

For completeness:

      Absent further discussion and agreement in the wg, I taking this 
exchange as producing /no/ changes to the spec.

d/


On 7/24/2018 7:58 AM, John Levine wrote:
> In article <[e-mail]> you write:
>> -=-=-=-=-=-
>> In table 2 on page 9, the draft refers to RFC 2782 for _dccp and _sctp (SRV), but those “_node names”
>> are not even mentioned in the RFC. Are they defined elsewhere?
> 
> RFC 2782 says that SRV's are named with _proto where proto is is a
> protocol name.  RFCs 3588 and 6733 say to do _sctp SRV lookups, but
> don't further define them, and only have 2782 as an informative
> reference.  No RFC mentions _dccp.
> 
> It seems to me that 2782 is the right reference for _sctp.  For _dccp
> we've had arguments about whether 2782 says that a SRV can be named by
> any protocol so maybe we should put in every protocol in the IANA
> registry.  That's a lot of dead protocols.  A reasonable compromise
> would be to start the registry with the names that have some evidence
> of being used somewhere, so we could drop _dccp
> 
>> In the same table, the draft refers to RFC 7553 for a number of URI _node names, but the references are quite
>> indirect. Could references to relevant IANA registries be added?
> 
> Since RFC 7553 is the place that says that the enumservice names turn
> into _node names, I think that's the right reference.



-- 
Dave Crocker
Brandenburg InternetWorking
bbiw.net