Re: [tsvwg] [Fwd: Re: AD review: draft-ietf-tsvwg-port-randomization-05]

[Joe Touch <[e-mail]>]

Tom,

The problem is that the two ideas (the list below, vs. Sec 7) have
completely different purposes.

Section 7 describe IANA policies and changes therein.

The above describes expert review, which is a separate matter.

I hesitate to encode expert review rules - even the "most common" ones -
because then can too easily change, and putting them in a doc promotes
them as static. Further, this doc is about IANA issues, not the expert
review process.

Joe


tom.petch wrote:
> ----- Original Message -----
> From: "Michelle Cotton" <[e-mail]>
> To: "Fernando Gont" <[e-mail]>; "Lars Eggert"
> <[e-mail]>
> Cc: "tsvwg" <[e-mail]>
> Sent: Monday, February 01, 2010 11:39 PM
> 
> 
> Fernando/Lars,
> 
> Following-up on the thread below.
> 
> Port Number requests are rejected for many reasons.  Here are three of the most
> common reasons:
> 
> 1 - Duplicates a function or protocol already in existence
> 2 - Is the secure version of a port already applied for or already in existence
> 3 - Port applied for is of local use only and traffic does not flow over the
> public Internet
> 
> In general, a port number request is granted when an applicant can show a
> well-defined, public Internet protocol.  The port must identify a named service
> that allows sessions to be created over the public Internet.  The protocol, in a
> successful application, will be sufficiently documented to ensure that it is not
> local or a version of another service already in use.
> 
> I hope this information helps.  Please let me know if there is anything I can
> clarify.
> 
> <tp>
> 
> Michelle
> 
> How straightforward and clear the above paragraphs are.  Honest.  Brilliant.
> 
> I need to say this because recently I was critical of
> draft-ietf-tsvwg-iana-ports-04
> for the lack of just this information, for describing the bureaucracy of port
> assignment/allocation/registration/reservation/request/grant........
> without
> a) explaining what if anything is the difference between these terms
> b) describing what the point of the bureaucracy is.
> 
> Your paragraphs above do more good than the whole of section 7, so please ditch
> section 7, include something along the lines above, and place it right at the
> front of the I-D, section 2.  I would place reasons for granting before reasons
> for rejecting but otherwise I would change little.
> 
> Tom Petch
> 
> </tp>
> 
> 
> Michelle
> IANA
> 
> On 1/27/10 11:20 AM, "Fernando Gont" <[e-mail]> wrote:
> 
>>> Have there any cases in which use of a port has been rejected?
>> Yes.
>>
>>> If so, what has been that reason?
>> Depends :-) Maybe IANA can give some examples.
>>
>>> And what has been the criteria for actually "granting" the use of
>>> ports (as the above)?
>> Satisfying the Expert Reviewer.