Re: Last Call: <draft-ietf-tsvwg-iana-ports-09.txt> (Internet Assigned Numbers Authority (IANA) Procedures for the Management of the Service Name and Transport Protocol Port Number Registry) to BCP

[Alexey Melnikov <[e-mail]>]

Hi Paul (and Cullen),

Paul Hoffman wrote:

> On 1/29/11 9:34 PM, Joe Touch wrote:
>
>> On 1/29/2011 8:54 PM, Cullen Jennings wrote:
>>
>>> On Jan 27, 2011, at 17:10 , Joe Touch wrote:
>>
>> ...
>>
>>>> AFAICT, the experts team reports to IANA. We make recommendations to
>>>> them. They are the ones who have the conversation with the applicant.
>>>> They can take our advice or not - that's their decision.
>>>
>>>
>>> I think you are pretty misrepresenting the situation. My impression
>>> of the reality of the situation is that if the IANA did not like the
>>> advice of the expert reviewer, they might ask the AD to override but
>>> short of that they pretty much do whatever the expert says.
>>
>>
>> As per my other note:
>> RFC2780 specifies Expert Review as *one* of the viable means by which
>> IANA can decide on transport protocol port assignments. The term "Expert
>> Review" is defined in RFC 2434.
>>
>> Neither document binds IANA to use the advice of a reviewer.
>>
>> Further, there is no single reviewer - we have a team, consulting each
>> other on occasion, and all decisions are seen by multiple reviewers.
>> However, none of that is worth codifying. If IANA or the IESG doesn't
>> like how we serve them, they can replace us - at any time, for any
>> reason, and there is an appeals process for decisions of the expert 
>> team:
>>
>> Any decisions made by the designated expert can be appealed using the
>> normal IETF appeals process as outlined in Section 6.5 of [IETF-
>> PROCESS].
>
> Now that this has been made clear to me, I am *much* more worried 
> about the wording in the current draft. The above emphatic statements 
> means that IANA can reject a request for an IETF-approved protocol 
> that needs two ports without recourse. As Cullen and others have 
> pointed out, there are sometimes very good technical reasons for a 
> particular protocol to need to have two ports.
>
> The document should be amended to say that protocols with IETF 
> consensus should get as many ports as it needs, regardless of what 
> IANA or the expert reviewer thinks. This makes it the responsibility 
> of the IETF consensus process to follow the guidelines in this document.

I think the document (-10) was amended as requested, but can you please 
double check?