Message-Id: <4.3.0.39.0.20000209080044.00c65f00@mail.bayarea.net>
Date: Wed, 09 Feb 2000 08:00:46 +0100
To: Keith Moore <moore@cs.utk.edu>
From: Dave Crocker <dcrocker@brandenburg.com>
Subject: Re: RFC822's Appendix D, and X- fields 
Cc: Keith Moore <moore@cs.utk.edu>, DRUMS <drums@cs.utk.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed

At 03:38 PM 2/8/2000 -0500, Keith Moore wrote:
>what you call chilling effects and what I might call strategically
>placed bars are sometimes useful.  some people think that the IETF

1.  Sometimes is not always.  X- is always

2.  For all of your concern about examples, what are the examples of 
BENEFITS from the artificial X- barrier?


>process imposes a chilling effect because people have to meet
>certain requirements to get their work declared as standards.

This isn't about meeting standards.  Those requirements are well enforced 
through other mechanisms.  This is about ALWAYS requiring starting over, 
and therefore losing all install-base momentum.  Always.

In effect you are insisting on a syntactic barrier for enforcing a policy 
barrier.  Overloading functions that way should be permitted only in the 
face of compelling requirement and a good ability to demonstrate that the 
requirement needs to be enforced that way.


d/

=-=-=-=-=
Dave Crocker  <dcrocker@brandenburg.com>
Brandenburg Consulting  <www.brandenburg.com>
Tel: +1.408.246.8253,  Fax: +1.408.273.6464
675 Spruce Drive,  Sunnyvale, CA 94086 USA

Gong Xi Fa Cai   /  Selamat Tahun Baru Cina