So yes Dean, I think you elude to the central issue - what is the common
interest, and as the community was propelled almost forceably, and
inexorably by market forces from a world where as Randy put it
"operators cooperated" together, in a non-commercial endeavor based on
very non-commercial values, into a very commercial world, did the
community ever have a chance to take a big breath, inspect what has
happened, and where it really wants to go from here.
Ok, here is MY opinion on the central issue:
Deployed communications systems are either military or commercial. Pure
research doesn't deploy systems. OC-192 equipment and millions of miles
of fiber aren't cheap. Those complaining about how great things were when
they were non-commercial were preceeded by people complaining about
working for the military.
The people who don't want to service the commercial sector, and don't want
to service the military sector, really don't realistically have much of a
place in our society outside of pure research, and shouldn't be given
control over the future commercial or military applications of the
internet.
But it comes down to a question of whether science serves public policy or
whether public policy serves science. Which is the tail, and which is the
dog? There is room for some pure science for nothing other than the sake
of pure science, but most of the research has to be done for productive
commercial or military purposes. I think this is a pretty general
statement that is as true of biotech research as it is of communications
research. And since the internet is now vastly commercial, and
international, the leaders of the IETF ought to focus research on things
that will be commercially useful.
The present situation, in my opinion, the tail is wagging the dog. This
should be changed. The IETF mission should make clear what the
constituencies are, what the goals are, and what the priorities are, so
that the tail does the wagging.
It used to be that engineering and operations within a company "cooperated
together", sharing work and, of course, passwords. In small companies,
this is still be true. But in many large telecom companies, engineering
is denied access to the production systems, and only the operations staff
can make changes. This is necessary to ensure commercial stability, since
custom engineering changes don't scale well, and don't lead to widely
stable systems, since the custom changes may not be made on all systems.
Large groups work at arms length, even within a company, with well defined
protocols and policies regarding who can do what. Of course, there are
inevitable disagreements between engineering, operations, sales, etc.
These are resolved civilly, within the organization, and may include
appeals to senior management. Those that know me, know that I'm usually
working in the engineering group, which means I am sometimes arguing with
the operations group for access to a newly deployed production system. The
first few times, I usually win. But I either have to stop asking or lose
these disgreements, in favor of documented troubleshooting procedures for
the benefit of improved stability. There are good reasons that practices
that might have been nice for small companies are abandoned by big
companies. The IETF has this same issue.
Clearly, there are going to be even greater problems when many companies
come together to work on the same problems. It is not reasonable to
expect that a few people can just "cooperate together" and work something
out for a large community. What worked for the Internet many years ago
isn't going to work again--ever again--anymore than it would work for a
large telecom company to share passwords between engineering and
operations.
--Dean