[dns-operations] RRL specified in a stable place?

[Doug Barton]

On 02/04/2013 03:16 PM, Paul Hoffman wrote:
> On Feb 4, 2013, at 2:32 PM, Edward Lewis <Ed.Lewis at neustar.biz>
> wrote:
>
>> Why an IETF document?
>
> Because that's where implementers look for such documents.
>
> Because there are implementers who are active in the IETF who might
> have valuable opinions on what the doc might say that would make it
> more valuable.

I agree with these reasons. Even if it was published as something like a 
BCP, it would fall into the IETF category, IMO. I'm not tied to the IETF 
path, but I think it's perfectly reasonable for people to look to us for 
DNS expertise.

As was pointed out elsewhere in the thread, there seems to be confusion, 
even in our community, about what RRL is, and what it isn't, so 
clarifying that would be a good additional purpose for the document 
(above and beyond simply describing what/how/why).

>> In what way does Response Rate Limiting impact interoperability of
>> implementations?
>
> It does not. So what? Did you somehow miss all the operational
> documents that the IETF helps produce?

I am not sure it would be defined strictly as an interoperability issue, 
but I could see utility in a document that answers the questions that 
may arise downstream when someone sees the effects of RRL and doesn't 
understand what they are seeing.

>> If this is not an independent submission, how does it fit into a
>> working group?  The implementations are pretty much out there,
>> what's to work on?
>
> There are two implementations. Are you saying there should be no
> more? Or, if there are more, that the implementers should have no
> clue about what earlier implementers thought about?

+1

If this document takes shape I would gladly volunteer to review it.

hth,

Doug