--On Thursday, June 19, 2025 14:02 +1200 Brian E Carpenter
<[email protected]> wrote:

> Martin,
> 
> On 19-Jun-25 13:45, Martin J. Dürst wrote:
>> Hello Brian, everybody,
>> 
>> On 2025-06-19 05:44, Brian E Carpenter wrote:
>>> On 19-Jun-25 05:37, Donald Eastlake wrote:
>>>> I want to support John Klensin here. I have always thought that
>>>> the ~immutability~ of RFCs was one of their greatest strengths.
>>> 
>>> True, but that ceased to be a simple property once we allowed
>>> multiple presentation formats. From then on, the property split
>>> into two: immutability of presentation (gone) and immutability of
>>> intent (hopefully still applicable). What we've been arguing
>>> about is how to precisely define immutability of intent.
>>> 
>>> A friendly amendment to:
>>> 
>>>>>           "Once published, RFCs may be reissued, but
>>>>>  only syntactic          changes that do not affect the
>>>>>  syntax for protocols          themselves may be
>>>>>  changed."
>>> 
>>> is:
>>> 
>>>            "Once published, RFCs may be reissued, but
>>>   only syntactic or          superficial changes that do
>>>   not affect the syntax for protocols          themselves
>>>   may be made."
>> 
>> I don't oppose the direction this is taking, but having
>> "syntax/syntactic" in the same sentence for two different things
>> may be confusing. Even more, RFCs don't only define the syntax for
>> protocols, they usually also include some semantic part. On top of
>> that, protocols are not the only thing we are defining.
> 
> That's all true, so a little more wordsmithing may be needed, but
> what I think John K and I are getting at is that the end goal is
> interoperability on the wire (or the wireless), so it is all
> matters related to such interop that absolutely must be immutable.

Yes.
Would "do not affect the operational specification of protocols
themselves..." be a bit closer?

   john



_______________________________________________
rfc-interest mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to