Re: IPv7 Selection Criteria

[Frank Kastenholz <[e-mail]>]

 > Hence, let me suggest explicitly distinguishing those criteria that
 > have clinical benefit, but lack objective assessment, from those 
 > whose assessment can be made mechanical.  The clinical criteria could
 > be in their own section, some sort of qualification explaining why they
 > are included and how they should be used.

The criteria document is divided into two major chapters, Chapter 4
entitled General Principles which has the harder-to-quantify stuff in
it (Live Long, One Protocol to Bind Them All, Architectural
Simplicity, etc) and Chapter 5, entitled Criteria, which has things
that Craig and I believe are easier to quantify.

I think that the evaluation of each proposal against the contents of
Chapter 5 can be reduced to a simple mechanical process is not
likely.  Some of the process can be mechanistic. Other parts will not
be. The main reason is that if we make Chapter 5 completely
mechanistic then it is conceivable that the people backing proposal
foo will fight against including any criteria that would elminate
their proposal.  (e.g. Suppose that someone offered, as a criterion,
that IPv7 must be derived from ISO work. For the sake of argument,
pretend that there is a "good" reason for this. I imagine that the
SIP/IPAE people would not like this criterion and would devote
considerable energy in making sure that it did not get into the
document).

The overall goal of the Criteria is to provide feedback to the
protocol designers as to what the community wants in IPv7 AND to
provide a framework for the "my protocol is better than your
protocol" discussion that is sure to happen.



--
Frank Kastenholz