[ntpwg] KoD 'backoff' BCP/direction

[TS Glassey]

Rob - PLEASE DONT TAKE my words personally - the are aimed at creating a 
strong and complete Standards Development Environment for NTP and to 
document and model its use in today's key use markets. :-)

I will also make this my last response to this on the open list. But I 
suggest you look at the retort's herein.

Todd


----- Original Message ----- 
From: "Rob Seaman" <seaman at noao.edu>
To: "TS Glassey" <tglassey at certichron.com>
Cc: <ntpwg at ntp.org>
Sent: Monday, June 02, 2008 9:19 AM
Subject: Re: [ntpwg] KoD 'backoff' BCP/direction


> Hi Todd,
>
> This has clearly passed the point of diminishing returns.

Why - because I suggested we get input from people you work for as to what 
they want here? :-)

Please dont take this personally - I am trying to figure out whether this is 
in fact an International Standards process or just a bunch of guys and gals 
that have an amatuer interest in timekeeping. It clearly makes a difference.

>
>> I dont want or need to evaluate the NTP Successor
>
> Ok.  Then my advice is to let this NTP update cycle finish and try to
> influence the next.  You don't appear to have any hope of gaining
> traction on these issues so late in the game.

The problem is that others have financial investment's in the time-spent 
here and that creates an issue.

>
>> the best way to move forward is to use a uniform set of development
>> assignments (IP Rights) and their owners across the whole process.
>
> You are discussing questions of design.  I am discussing questions of
> process.

I disagree. But its more an issue of the style - for instance - the code 
changes which are captured in the NTPWG's Document Effort are what we are 
talking about. So how are the IP rights to that codified in a manner which 
can then be assigned to the IETF through the NTPWG's operations ??? See my 
point?

> Any development cycle starts with design issues and ends
> with implementation issues.  It is simply too late to influence the
> current cycle (other than to slow it down).
>
>> The NTP.ISC.ORG operations are not doing that now but can easily be
>> setup to do so.
>
> Write a white paper now, to influence decisions later.

My comments above clearly address this as well.

>
>> And BTW if you want to see if your organization supports that or
>> your assertions or mine here, lets get your general counsel
>> involved...
>
> I don't recall making any assertions.  It would be pointless to
> construct an opinion on the skimpy basis available at the moment.

So then lets ask ALL of our master's what they formally want from NTP and 
its implementation... A simple Survety would do this and add a huge amount 
of credibility to the process - (Look you may not like my attitude on this 
stuff but these ideas are priceless).

> From your association with Certichron it is clear that you have a
> horse in this race.  I'm personally only interesting in seeing the
> race actually occur.

Cool - I think that is a good goal. One I support.

>
> (BTW, I might find some other analogy than "trust-anchor".  It really
> conjures visions of walking the plank and having ones cargo of trust
> carried downward into the inky blackness.

Did you watch the Black Pearl last night?

> An anchor is also only a
> temporary mooring, sometimes slips on the sea floor, and occasionally
> has to be cut loose and left behind.

Trust Assertions are temporary in form. They exist as long as the anchor is 
real and properly rooted in its trust hierarchy.

> A single anchor leaves the
> vessel free to swing willy-nilly in a large arc with the wind and
> water.)

Agreed which is why AutoKEY by itself isnt enough.

>
>> As I recall Carrie-Nunes was the Assistant/Acting GC for you folks,
>
> Not sure who this is or who "you folks" refers to.

NOAO.

BTW Did you know my academic minor is Astronomy? I also used to work in 
sub-picosecond physics at SLAC and another 'place' ... Bubble Chamber type 
stuff, so timing is something near and dear to my heart.

>
> In any event, legal and economic use cases are indeed significant.

Thanks for acknowleding that - can you perhaps also answer why they are 
ignored by this WG in its mad dash to publish?

> I've spent eight years arguing this regarding leap seconds.

Ahh - yes the Leap Second Bingo Game - I wrote a white paper for the 
Securities Industry on this a couple of years ago. A number of larger 
Securities Traders have taken up a two layer timing model to address this, 
its a solution I am very proud of as is my replacement for ACTS solution.

> However,
> you haven't connected the dots to show that NTP is unresponsive to
> these.

Leap Seconds are only relevent to logging operations which need to properly 
codify a process to reintegrate those offsets into them to make the logs 
'propah'. In fact since UTC is calculated and proclaied from BIPM some 30 
days after that point in time this commentary of mine is dead-on here.

> In fact, your participation here suggests that no other
> technology is more mature for your purposes.

true.

> That being the case, it
> is to your own advantage to move the NTP standards process forward to
> provide yourself the opportunity of hopping aboard the next time the
> carousel goes by.

Not if that means 'breaking it' in the interim. Its better not to have to 
say "I am sorry" and fix something that was done just to get someone elses 
efforts into print. Its better to stop those from being printed and fix the 
problems. The issue here is that some of this group has a vested interest in 
publishing those document's whether they work or not. That is a conflict of 
interest since who is being benefited is the operator of the carousel and a 
very few riders who live under the publish or perish idiom IMHO.

Todd
>
> Rob
>
> _______________________________________________
> ntpwg mailing list
> ntpwg at lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/ntpwg