> > People should generally reject religious fundamentalism, including
> > what you said there. If you have an a priori case for this or that,
> > fine, but you seem to be making a mystical argument for a wildly
> > generalized claim.
>
> for some, "let the market decide" is a religious statement. it's
> generally based on an unexamined faith in market conditions as an
> effective way of making a good choice among competing technologies.
>
> For some it may be so (I agree) but apparently not you are I.
>
> All I mean is that, for higher level protocols, letting people do what
> they will ("the market decides") seems to me to be the best option.
there are cases when this is true. I don't think the cases are as simply
described as "higher level protocols" or "applications". it depends, for
example, on whether competing apps interfere with each other or with other
valuable network facilities, whether particular solutions will do harm to
networks and/or end systems when deployed (say, due to poor security), and
on the size of the market for a single solution vs. the size of markets
for fragmented solutions.
it's a bit like an argument that there shouldn't be any standards for
automobiles. which leads to more pollution, higher accident rates, higher
casualties, higher insurance rates and/or more expensive vehicles for
everyone. there are corrective mechanisms for some of these (as in,
people will tend to favor cars that have lower insurance rates) but some
of an individual's risks due to automobiles aren't influenced by his own
decisions so much as others' decisions.
> Yes, using your example, IM protocols fragment, interop suffers,
> there's lots of crap --- so what? It looks like it will probably sort
> out. Fretting over how best to impose governance over the situation
> doesn't obviously accomplish anything.
well, nobody is talking about governance, since we don't have any
enforcement mechanisms.
> There are faster, better,
> cheaper ways I think to get universally adapted standards in motion
> than wrangling with committees.
it's pretty difficult to build something that works well in the diverse
environments in the Internet without extensive discussions among
experienced people with a wide variety of interests. to the extent that
it happens, it's either luck (rare) or genius (even rarer).
> The main practical utility of standards very high in the stack is that
> they are written clearly, widely reviewed, and generally agreeable to
> people. One doesn't need a complicated government to support the
> development of standards with that utility: a few mailing lists and
> archives do most of the trick.
again, nobody is talking about a government. we're talking about a
mechanism to support engineering work among diverse and often-competing
participants, that will produce results in which users/customers can have
a high degree of confidence.
> Processes in which there is nothing
> available to fight over seem like the most efficient to me.
that's because you aren't considering the cost _after_ that process.
Keith
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf