On 10/9/2018 2:45 PM, Adam Roach wrote:
This is based on an assumption that document authors who add
enumservices are more likely to notice the need [1] to add their service
name to two tables than the IANA are. Given the respective levels of
rigor, that seems like a losing bet.
There is certainly a substantive discussion to have about this, since
the working group did.
But I'll suggest something simple, in the hope that it actually
simplifies things in process terms:
This issue was discussed at some length within the working group,
including disagreements of the sort you raise now. Eventually the
working group finally settled on the choices made in the draft.
That is, this is a matter of some tradeoffs that the working group
considered.
d/
ps. FWIW Personal comment: having to coordinate tables based the SRV
spec is unfortunate. Thinking about having to factor in the URI details
pretty much blew me by the fact of it's using /two/ tables and how it
used them.
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
