[ntpwg] Pending NTP WG Last Call on Autokey

[David L. Mills]

Danny,

In absolutely no way whatsoever does my proposal cripple the protocol 
than does yours gloriously enhance future happiness. My proposal 
preserves current programming style as I designed it, evil as you might 
find it. Yours is as I consider uneccesary, wasteful of header space, at 
odds with current programming style and makes it most difficult to reuse 
code. I'm not going to convince you and you are not going to convince 
me. There you have it. You are invited to purify future versions of this 
or another protocol, but leave this one  alone. Get this puppy out the 
door and save your energy for another time.

Dave

Danny Mayer wrote:

> Rob Seaman wrote:
>
>> Danny Mayer wrote:
>>
>>> I have no clue what is meant by broad-distribution models and I have
>>> yet to see anywhere in the protocol where it's even mentioned, never
>>> mind required.
>>
>> Todd didn't answer any of my (pretty rhetorical) questions, but made
>> it clear that his third of this discussion centers on specifying legal
>> requirements. Dave (quite understandably) appears to want to simply
>> put this puppy to bed. I doubt I'm alone in finding this an
>> attractive option. Your "third way" position is:
>>
>
> Right. This draft is not a legal requirements document, it's a technical
> protocol document.
>
>>> That is my reason for suggesting that we take a bit out of the
>>> header so that the current autokey implementation continues to work
>>> while enable the correct layout to be implemented in parallel.
>>
>> My reaction to this is from http://www.dspcsp.com/tictoc:
>>
>
> TICTOC has a different agenda.
>
>>>> The NTP Working Group is currently standardizing the fourth version
>>>> of NTP for time distribution over IP networks. The NTP WG has
>>>> focused its deliverables largely on standardizing the currently
>>>> deployed NTPv4, while collecting requirements for future
>>>> extensions. These requirements will transition to the TICTOC WG for
>>>> further development.
>>>
>> Why now, rather than later? Wouldn't a staged deployment be prudent?
>> Or perhaps there is something here I'm missing? If so, I also doubt
>> I'm alone in this.
>>
>
> I want to get this out too. The trouble is that I want to do this right
> rather than do it over much later when it's much harder to recover. Dave
> is rarely wrong about NTP but we are talking about interoperability and
> extensibility and crippling the protocol at the beginning limits its
> usefulness.
>
> Danny
>
>
> _______________________________________________
> ntpwg mailing list
> ntpwg at lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/ntpwg