DNSO Mailling lists archives


<<< Chronological Index >>>    <<< Thread Index >>>

RE: [ga] Letter from ICANN to New.net

> From: Kent Crispin [mailto:kent@songbird.com]
> Sent: Thursday, July 26, 2001 10:09 PM
> On Thu, Jul 26, 2001 at 06:48:42PM -0700, William S. Lovell wrote:
> > A telling presumption exhibited here: if the "protocol community"
> > doesn't like something, ICANN should dump it.
> Yes.  That is the same presumption you would use to say "if heart 
> surgeons think a procdure is dangerous, dump it". 

No it is not. The fundimental difference is that hearts aren't artifacts and
heart surgeons aren't [yet] artifacers. Computer Science isn't a science.
It is an art, wholly created by man. As such we understand it very much
better than we understand medicine. We can't create organic neurons, but we
can create cybernetic neurons and even entire neural networks. The entire
field of Computer Science is a fabrication of mankind, a pure artifact of
our civilization. It is arguably, the only field where this is the case.

> >  However, the
> > Internet does not exist for the benefit of the "protocol community"
> > or ICANN; those two entities exist for the benefit of Internet
> > users.
> Right.  Medicine doesn't exist for the benefit of doctors.  But if 
> doctors tell us something is a bad idea, we generally listen.  If we 
> got practically universal agreement among doctors that a 
> procedure was a bad idea, we should almost certainly dump it.

The analogy breaks down very badly here. As Computer Scientists, we both
recognise that we have vastly more control over our work product, which we
create, than Medical Doctors whom, by comparison, aren't much better than
technicians. As a result, the same limits do not apply.

> The problem is protocol engineering really is a species of "rocket
> science"* -- it takes a long time to really understand the 
> issues.  And, 
> despitewhat you hear, most of the participants on these lists really 
> aren't rocket scientists of the proper variety.
> (*) for those who may not be familiar with the idiom, "rocket 
> science" is a slang phrase meaning "truly esoteric and specialized
> speciality". 

I don't know about you, but I've been writing communications drivers since
long before even the IAB existed. This was ages before the IETF existed,
which was mostly populated by academics and not working engineers. The
engineers hung out in the IEEE, in those days, even if some of us did join
the ACM. When was the last time you baked a seven-layer ISO cake, if ever?
But I digress, there are probably only three or four of us here. That's
nowhere near a majority. How many of us are active in the IETF? The IETF
includes a very small portion of the Computer Science community.

The fundmental difference, between rocket science and computer science is
that rocket scientists are limited by physics, Computer scientists aren't.
Yes, it may take a while for computer manufacturing capability to catch up
to some of our artifacts (ie. rendering engines), but that doesn't stop us
from conceiving and building the artifacts. The only limits are E. Dykstra's
seven non-computable tasks.

The problem is that the IAB and the IETF have forgotten this and are
treating, long established, arbitrary convenience limits as if they were
laws of physics (they aren't). When that fact is pointed out to them, they
shout it down with religious zeal and the opposition is so strenously
harassed that they are encouraged to leave participation in the IETF. IOW,
it is a very hostile environment.

Given this environment and those conditions, it is small wonder that the PSO
is exclusively populated with those of similar mind-set. They've
brain-washed each other that way. To then hold up that body as a true
representation of all the experts in the world, in that speciality, is, to
say the least, absurd and disingenuous in the extreme.
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html

<<< Chronological Index >>>    <<< Thread Index >>>