[ntpwg] [dhcwg] NTP option: IP address and/or FQDN

[TS Glassey]

Harlan  - thanks , let me answer your key comments?

----- Original Message ----- 
From: "Harlan Stenn" <stenn at ntp.org>
To: "TS Glassey" <tglassey at earthlink.net>
Cc: "Alain Durand" <alain_durand at cable.comcast.com>; "Richard Gayraud 
(rgayraud)" <rgayraud at cisco.com>; "Danny Mayer" <mayer at ntp.org>; "MORAND 
Lionel RD-CORE-ISS" <lionel.morand at orange-ftgroup.com>; "DHC WG" 
<dhcwg at ietf.org>; "Ralph Droms (rdroms)" <rdroms at cisco.com>; "NTP Working 
Group" <ntpwg at lists.ntp.isc.org>
Sent: Monday, December 10, 2007 1:39 PM
Subject: Re: [ntpwg] [dhcwg] NTP option: IP address and/or FQDN


> Todd,
>
>> Boy is this commentary going to make me unpopular...  but the key to this
>> problem isn't technology - its the License. The real failing in this 
>> matter
>> is that the NTP License allows people to mis-use NTP  and to do whatever
>> they want in hard coding the addresses into the product and based on the
>> existing license and the modified license created by the IETF there never
>> will be anyway to control the mis-use of NTP.
>>
>> By the way - as I noted, the IETF's "Any and All" uses model further 
>> creates
>> this same problem by making those that misuse of the protocol still 
>> within
>> their licensing models so while it's nice that the WG is worried about
>> bad-NTP users or vendor's who do things that damage those of us operating
>> public access NIST servers.
>
> I think in some ways you have a great point here.

Thanks!

>
> And as proof of concept, I invite you to show how well this will work by
> using these very same techiniques to solve something like the spam,
> phishing, or malware problems.

No can do - This type of thing wont work for those particular problems. 
Unfortunately for all of us SPAM traffic is probably key to keeping the 
merged data network "inflated" properly.

> I won't ask for a "finish-line" solution

Why not - A licensing requirement requiring a proper operating environment 
and a BCP for its operations is what it needed today.

> (ie, "At 200902281234, the last piece of spam email was sent", but if
> you can get that race actually started then I'll have much more
> confidence that this will work.

Uh OK let me respond - "in the US no Public Company can knowingly purchase 
and deploy any softwares in violation of those softwares use licenses. So as 
it happens a more restrictive license really does work IMHO.

>
> And just to ask, what are some of the 2nd and 3rd order effects of your
> proposal?

As far as what? As far as drafting a license for the use of NTP which 
precludes bad behavior? - OK - from my perspective there are two key effects 
from this as 2nd and 3rd order effects: the Second Order effect is that 
Wisconsin and the SNAFU last year would never have happened and the third 
order effect is that I believe that the Networking and OS Firms would have 
all ponied up some denario ($$$) to support their formal operations here  in 
NTP land and you would have had all this advanced about a light-year farther 
than it is now.

So my take is that the more restrictive license is key to forcing the proper 
operations of the tool suite.

> I'm talking holistically here.

yeah - me too...  :-)

> Like "the wall" in a General
> Systems Theory analysis.


You think? As I recall a "Complex System" consists of some number of 
independent or freestanding components, and yes these components interact 
locally. The system's overall behavior is generally independent of the 
internal structure of the components or the individual behaviors of the 
component elements, and the overall behavior of the system is well-defined.

Myself - I thought of it as more domino theory in a cause and effect model, 
but hey a simple "Wall" is OK 4 me too.


T/

>
> H