Re: [DNSOP] Last Call: <draft-ietf-dnsop-onion-tld-00.txt> (The .onion S

On 14 Jul 2015, at 22:16, Ted Hardie wrote:

> Further, I believe this stretches the "special handling" requirement of RFC 
> 6761 to the breaking point.  This does not describe special handling _within 
> the DNS_, but instead removes a portion of the global namespace from the DNS 
> at all.  To me, at least, this does not seem to me to meet the analogy RFC 
> 6761 provides to IP multicast ranges or local addresses.   Whether it is 
> permitted or not by RFC 6761, it is a bad idea.
Speaking personally I think the intention of RFC6761 was to allow any kind of 
splitting of the namespace used by for example the global DNS so that there 
would not be any confusion and/or questions on how to resolve a specific 
string. See other similar specifications regarding .EXAMPLE, .LOCAL and such. 
And I think this is exactly the kind of need that exists for .ONION. The 
reference (normative or informative) is something I agree can be discussed 
whether it should not have been normative.

If the wording in RFC6761 is such that we do not agree that this is the 
process, but instead that RFC6761 is only for specifications that use the DNS 
for resolution, then I think RFC6761 should be made more clear, and not this 
document.

RFC6761 should be about namespace management, which I think is what we discuss 
here.

Because of this, I think this RFC should be published modulo resolution of the 
question about the reference.

> ​My opinion only,
Likewise ;-)

   Patrik

signature.asc
Description: OpenPGP digital signature