Re: [tsvwg] Comments on draft-ietf-tsvwg-transport-encrypt-14

[Gorry Fairhurst <[e-mail]>]

On 06/04/2020 16:00, Joseph Touch wrote:
> Notes below...
>
>> On Apr 6, 2020, at 3:23 AM, Gorry Fairhurst <[e-mail] 
>> <mailto:[e-mail]>> wrote:
>> On 05/04/2020 23:50, Joseph Touch wrote:
>>>
>>>
>>>> On Apr 5, 2020, at 3:32 PM, Black, David <[e-mail] 
>>>> <mailto:[e-mail]>> wrote:
>>>>
>>>> ...
>>>> ...
>>>> OLD
>>>>     UDP-based protocols often do not use well-known port numbers.
>>>> NEW
>>>>     UDP-based protocols often do not use well-known port numbers,
>>>>     and use of a well-known port number is not limited to the
>>>>     protocol for which the port is well known [RFC7605].
>>>
>>> RFC7605 doesn’t say this, or at least most of this.
>>>
>>> Yes, there’s no rule that well-known port numbers are used. But 
>>> that’s not either unique to nor biased towards UDP vs TCP.
>>>
>>> Joe
>>
>> I understand, I think we can mention RFC7605 in 3.1.1 by adding this 
>> to the para about ports to read something like:
>>
>> "In some uses, an assigned transport port (e.g., low-numbered port)
>>
> (e.g., 0..49151)
>
> [I would avoid the term ‘low-numbered’, which could be interpreted 
> as meaning 0–1023]
>
>>  identify the protocol [RFC7605]. However, port information alone is 
>> not sufficient to guarantee identification. Applications can use 
>> arbitrary ports and do not need to use well-known port numbers. The 
>> use of a well-known port number is also not limited to the protocol 
>> for which the port is well known.
>>
> Well-known often implies System.
>
> As per RF7605, I’d encourage the use of “assigned” for the entire 
> range from 0..49151.
>
> Joe
>
>
OK. We now have:

In some uses, an assigned transport port (e.g., 0..49151) can
identify the protocol [RFC7605]. However, port information

alone is not sufficient to guarantee identification....


Gorry