Re: Clarification on recent comments as to ISOC being in charge of Internet

Einar Stefferud <Stef@nma.com> Thu, 28 October 1993 22:23 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa17950; 28 Oct 93 18:23 EDT
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa17946; 28 Oct 93 18:23 EDT
Received: from mhs-relay.cs.wisc.edu by CNRI.Reston.VA.US id aa24333; 28 Oct 93 18:23 EDT
Received: from cs.wisc.edu by mhs-relay.cs.wisc.edu with SMTP (PP) id <23172-0@mhs-relay.cs.wisc.edu>; Thu, 28 Oct 1993 16:37:59 +0000
Received: from ics.uci.edu by cs.wisc.edu; Thu, 28 Oct 93 16:37:50 -0500
Received: from nma.com by q2.ics.uci.edu id aa00426; 28 Oct 93 13:03 PDT
Received: from localhost by odin.nma.com id aa10665; 28 Oct 93 11:19 PDT
To: ietf-osi-x400ops@cs.wisc.edu, Richard Colella <colella@osi.ncsl.nist.gov>
Cc: AIKEN@ccc.nersc.gov, pays@faugeres.inria.fr
Subject: Re: Clarification on recent comments as to ISOC being in charge of Internet
In-Reply-To: ietf-osi-x400ops message of "Wed, 27 Oct 1993 07:59:09 PDT." <931027075909.218197e9@CCC.NERSC.GOV>
Reply-To: Stef@nma.com
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Einar Stefferud <Stef@nma.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 28 Oct 1993 11:19:21 -0700
Message-Id: <10663.751832361@odin.nma.com>
X-Orig-Sender: stef@nma.com

Hi All -- I do not recall making this statement.

>  . as said by Stef, is there any law preventing an operator to
>     avoid a network service falldown by advising against
>     offending products?

I will only agree with it in the context of a specific ADMD or PRMD
guarding its own QoS.  Certainly each ADMD operator has so far found
it important to certify interoperability, and each has set up
certification programs for this purpose.  No one that I know of has
objected to any ADMD (or PRMD) doing this as an individual action.

Also, in the US, the NIST has attempted to register claims of
interoperability, which would appear to provide some of what we are
seeking, but I am not sure this has become terribly productive as yet.
Richard Collella can speak to this.  He will be at Houston.

I would expect PRMD operators to do the same for their own protection.
So, I expect that it will be fine for any PRMD to announce in public
which systems with which they will be willing to exchange X.400 mail.

I would also map all this over to the SMTP Internet, though we do not
have ADMD and PRMD entities per se in SMTPland.  We do however have
corporate enclaves which can and should exercise some control over
their SMTP systems and the gateways there-to.

We also have Commercial EMail Service Providers (who are also ADMD
Operators) who should be just a fussy with their SMTP as with their
X.400 (as I see it), but who often are not.

As I have noted in other places, The Internet is not a Service
Provider!  The Internet is a Marketplace, containing Vendors and
Customers who share access to each other via "public" thoroughfares,
which in our Internet case are provided by the interconnected IP
service providers, who are funded by their customers.

A big difference that I keep pointing out between an Internet full of
mail exchangers, and the X.400 model of ADMD service providers
providing all the interconnections between PRMD service providers, is
that in the Internet, both ends are responsible for assuring that
their ends, including their IP service providers, are performing
properly to provide good IP service to the backbone (soon to be NAP
based).  The other end is responsible for the other half of any
connection.

This is very different from the PTT model where the 3rd party ADMD
claims a major role in the EMail interconnection, and talks a lot
about QoS and Service Level Agreement enforcement, and sort of implies
that they provide end-to-end delivery.  Of course, we know they cannot
deliver end-to-end service if they are dumping mail off to a PRMD!

So, in the Internet, we only have peer pressure.  What might be a good
approach would be for Postmasters to monitor performance of
interconnecting SMTP servers and notify poor performers of their lack
of measured quality.  They can also notify their own users about why
certain other EMail domains/sites are not providing good service.

We all know about certain domains where MSMail or cc:MAIL gateways are
badly misbehaving.  Postmasters might do well to post such information
on local BBoards so that users can tell their correspondents about it,
and stimulate local pressures on the other end to get things fixed.

Surely no one can sue me to reporting to them the measured results of
Email transfer transactions.  The reports might even identify the
software that is failing on the reported connections.

I suspect that any postmaster worthy of the name will respond by doing
something about a continuous stream of low quality reports.  Just as I
would expect an ADMD to do in the face of the same thing.

Of course, there are and always will be, postmasters who are not
worthy of the name;-).

Cheers...\Stef