[ntp:hackers] Standardizing NTP...

Terje Mathisen terje.mathisen at hda.hydro.com
Tue Jun 14 04:07:05 PDT 2005


hackers-bounces at support.ntp.org wrote:
> Rather than arguing with you Brad I would propose that Standardizing NTP is
> the goal of this list of people and since its the group spawned though
> NTP.ORG from Dave Mills it seems to me that this is the right place to own
> NTP.
> 
> I suggest that there are two core areas of focus. That of the continued
> expansion of the codebase and the features that make up NTP, and another
> that would be specific to standardizing and owning the operational practice
> statements as to how NTP is to be used. This is where Terje that your
> operations model becomes of some value to the group in your documenting it
> as an operational practice. It's obvious where the bulk of the group's
> efforts today lie and that's cool.

Todd, I have read all of your emails, including the statements about
patents etc, and it is obvious that you have spent some time considering
this.

I still, most repectfully, believe that you are wrong when you claim
that nothing but your own patented/certified time model is correct.

I.e. you claim that the computer logs generated by the current ntpd are
legally worthless, while we all know that exactly these kinds of logs
from other systems/programs have been admitted as evidence.

I assume you never bothered with a bound notebook to write down your
work when trying to come up with a patentable concept?

Written logs being legally worthless?
> 
> My point is not to fight with anyone here, but rather to build something
> that allows me to continue my admittedly commercial interests, in packaging
> and designing interoperability models in global (national timing services)
> solutions.

So why don't you do just this?
> 
> What I propose is that I am willing to work to create the audit profiles and
> controls as well, and that by merging all of these efforts, there is a way
> to make NTP self supporting...

NTP is self supporting, it just isn't Todd supporting. :-)

OTOH, I really cannot see what's stopping you from taking the current
ntp-dev source code, fork it and generate a version that works the way
you think it should do.

You can't sell the original NTP source code, but you can probably sell a
service which consists of set of patches applied to a given version of
ntpd, plus a nicely framed certificate to put on the buyer's wall.

Terje

PS. Even though I'm norwegian, I have been asked to offer expert
testimony in a US patent dispute.

Of the 10 patents that were assumed to be infringed upon, 6 of them were
wrongly granted imho, because they were _far_ too obvious, not just to
'someone skilled in the field'.

2 more were patents allowed on standard textbook (Knuth etc) algorithms.

The final pair actually deserved to be patented.
-- 
- <Terje.Mathisen at hda.hydro.com>
"almost all programming can be viewed as an exercise in caching"



More information about the hackers mailing list