Re: [art] draft-ietf-dnsop-attrleaf

[Adam Roach <[e-mail]>]

On 8/9/17 8:15 PM, Dave Crocker wrote:
> On 8/9/2017 4:25 PM, Adam Roach wrote:
>> I would recommend documenting and registering the underscore leaf 
>> nodes used by TXT records as a completely separate concern and in a 
>> different IANA table
>
>
> Adam,
>
> (Leaving for later, all other comments about the note, other than to 
> say thanks for reading the draft, thinking about it, and commenting...)
>
> Your above seems to be calling for two registries that assign out of 
> the same namespace.  (Being specified in the same document doesn't 
> matter if there are two tables.)
>
> I'm hoping I've seriously misunderstood, since the entire goal of my 
> simplification proposal is to avoid exactly that.
>
> Please to set me right. 

So, I'll reiterate explicitly what I've only implied to this point: I'm 
not a DNS expert; I've simply dealt with protocols that use DNS, and 
I've managed DNS zones.

What I do know is that I can ask for an RR of type FOO for 
something.example.com and get a response with an answer, and I can ask 
for an RR of type BAR for something.example.com, and get a response with 
no answers. I've internalized this as FOO and BAR being different things 
-- different scopes, effectively -- but (based on your and John Levine's 
and Mark Andrew's responses) clearly have an incorrect mental model of 
what's considered "namespaces," at least as compared to people who deal 
with DNS in more detail.

So ignore the aspect of my proposal that suggested two tables. I think 
the remainder is sound, which probably just resolves to cleaning up the 
table so that it contains only seven entries and removing discussion of 
NAPTR RRs.

/a