[1247] in Commercialization & Privatization of the Internet
Re: technical details
daemon@ATHENA.MIT.EDU (Martin Schoffstall)
Wed Aug 28 13:08:48 1991
In-Reply-To: <CMM.0.90.2.683319946.ittai@shemesh.ans.net>
Date: Wed, 28 Aug 91 10:53:43 -0400
To: "Ittai Hershman" <ittai@shemesh.ans.net>
Cc: "William Schrader" <wls@psi.com>, com-priv@uu.psi.com, lear@turbo.bio.net
From: "Martin Schoffstall" <schoff@mail.psi.net>
Reply-To: schoff@psi.com
>
> 1. ANS offers a way in which mid-levels can leverage their
> existing NSFnet-funded attachments to also carry
> commercial traffic (and avoid the work of dealing with
> the routing issues).
This is your best spin. However, most people I talk to use leverage in a
different manner when referring to all of this - basically they believe
that you have leveraged off the NSFNet and government funding. Actually
the words they use are stronger than leverage. I continue to counsel
that they not take action on this. However statements like this are
like waving a red flag in front of a bull.
As for routing - that is a red herring. You can always simplify routing by
having one single connection/path to "default", people have done this
for almost a decade. This is the way that most retail connections are
handled; however, many of the midlevels have always had rich
interconnection environments, and have dealt with "routing issues" from
day one - before NSFNet v2.0 was even in place. Quite frankly I
remember the mid-levels explaing to your crew how to do this in the beginning.
Effectively you are making a competitive advantage in your many messages
and ANS's various dealings of being the top of the default hierachy.
This ensures that many people will work very hard to make sure that this
is divorced from whatever you are in charge of asap.
>
>
> 3. As Marty recently pointed out, ANS is allowing the NSF
> to build a better, stronger, faster, network than they
> could afford. That better, stronger, faster, network
> is in-part funded by commercial connections (either
> directly via ANS CO+RE, or through the mid-levels).
Obviously I would never say this. I don't repeat TV ADS as part of
philosophy to live by. However, this is your 2nd best spin. A lot of
people would beg to differ with the adjectives; however, in your
"leveraging" off the NSFNet, you have introduced significant instability
as one of the conditions to do this appears to be your trapezee-like
settlement methodology. This introduces variable rate charging into the
INTERNET. You pretend that it doesn't by saying that it doesn't apply
to the retail connections. But it DOES apply to the midlevels, got any
good suggestions how they pass this on to their customers, in a
non-variable manner? Effectively your model is a trojan horse.
Of course one of the principal points of your rhetoric is that "you" put
in infrastructure, and either implicitly and explicitly stating that no
one else is. This of course is pretty insulting to both the midlevels
and the "campus" networks, whose investment dwarfs yours. This is
despite the fact that we're using emminently more financially efficient
equipment, bandwidth, operations, etc than anything used to date within
"your" network.
Now, given that "you" are in charge of the top of the default hierarchy
and a substantial port of the international links, we need to see that
they get moved because there is no sense in paying you (twice in many
cases) for this "service", which should be done by a qualified neutral
group that the community selects in some manner.
As for the NEX, described to date, it is emminently compatible with the
CIX and in fact architecturally looks identical to the CIX except that
it master the top of the routing hierarchy, as I understand it,
I wholeheartedly support the NEX concept. I perceive the NEX has a
threat to your model and marketing in a substantial manner.
Lastly - "you" and "your network". I view ANS as a contractor in
providing the NSFNet (how this happened, who knows), however, some of
your actions to date are not that of a contractor but that of a public
policy maker.
Marty
