[ntpwg] [ntp:hackers] Announcing the NTP Forum

[Paul Vixie]

dave, i know that you have no reason to take my views on ntp seriously, since
for every million lines of code you have contributed to time protocols, i've
contributed about 0.5.  but if you will allow me to draw on my dns experience,
i have some observations i'd like to share.

> ... That's the conventional approach and what most of
> us academic types have been doing for our adult lives.

i know that "time protocols" was once a purely academic matter, and that there
is still some pure academic work to be done at the level of improving the
general understanding of the problem spaces, investigating new approaches,
measuring and improving current approaches, and so on.

but "time protocols" also has operational and business facets.  recurring
revenue and expense, with dedicated management and technology resources, and
designated responsible individuals, and "big-tent politics", must also appear
if "time protocols" is to be a viable ecosystem, and to be a vibrant part of
the larger "internet ecosystem".

(all of which is true of "domain name protocols" which is why i claim to have
some wisdom to share on this general topic.)

i've seen that some of your investigations over the last few decades have used
your "reference implementation" as an experimental test bed.  you're a real
engineer as well as a real scientist, and you aren't satisfied with proofs of
correctness, simulations, or whiteboard exercises.  many of us are happy
running code with features that may or may not appear in an RFC some day.

however, the internet is far larger than academia or even geekdom, and to
support "time protocols" in this larger non-academic context, it's necessary
to reach beyond the protocol development, advanced development, open source,
and academic communities, and to reach out to the business, security,
operations, government, and legal communities.

because of harlan's dual relationship to both udel and the ntp forum, this is
not a code-fork event.  it's our hope that there will always be a reference
implementation whose technical correctness has mills at udel.edu as its final and
absolute judge.  but since there are opportunities and obligations in "time
protocols" that go beyond udel's interests, we're creating an ntp forum as a
vehicle to address those opportunities and obligations.

mills at udel.edu will always be "father time" to us.  i say this not only as the
one time operator of clepsydra.dec.com, but also as president of isc whereat
the ntp forum secretariat is presently housed.

paul