At 01:01 PM 2/11/2002 -0500, Keith Moore wrote:
> 1. You need to learn a little bit about tradeoffs between individual and
> group decision making.
I've had quite a lot of opportunity to observe it for myself, thank you.
In that case, you need to learn a bit about asserting behavioral
generalizations based on individual experience. The quick summary is that,
at best, it is nearly useless.
Start with the construct of sampling error.
And, by the way, you appear to be confusing the difference between design
by committee and review by committee.
> 2. Extending your logic means that we take away all individuals
> rights. Unless an action is approved by the majority -- and hence the term
> "tyranny of the majority" -- then it is to be prohibited.
You and I both know that the IETF doesn't work that way.
We are not talking about an IETF standards process. We are talking about a
label registration process. They are different, though you do not want
them to be.
> >Yes they do. But most protocols don't allow arbitrary extensions from
> >other parties.
>
> Yes, I've noted this rather peculiar line of concern from you. It seems to
> suggest that cooperation among consenting hosts is not to be permitted
> unless it has IETF approval. Hence my term "parental".
Protocols are carefully constructed ...
Keith, many thanks for the tutorial about the nature and requirements for
protocol construction.
It was, of course, irrelevant to the question of permitting consenting
hosts to use a public registration venue for mechanisms for that are not
(yet) standardized.
> If registration requires citation of a specification, then community review
> is inherent.
No it's not. Few people will bother reviewing a specification when it
is first introduced unless there is some chance that their review will
do some good.
Keith, you keep inventing new requirements or concerns. The more
requirements you add, such as early review, the more it is made
compellingly clear that you want a full standards process.
d/
----------
Dave Crocker <mailto:dcrocker(_at_)brandenburg(_dot_)com>
Brandenburg InternetWorking <http://www.brandenburg.com>
tel +1.408.246.8253; fax +1.408.273.6464