[8445] in Commercialization & Privatization of the Internet

home help back first fref pref prev next nref lref last post

A Trojan Horse? Was: Russian chemists harmed by ANS-CIX dispute?

daemon@ATHENA.MIT.EDU (Martin L. Schoffstall)
Fri Nov 19 12:23:06 1993

Date: Fri, 19 Nov 1993 12:29:49 +0000
From: "Martin L. Schoffstall" <schoff@psi.com>
To: "Peter S. Ford" <peter@goshawk.lanl.gov>,
Cc: com-priv@psi.com, schoff@psi.com


> Lars, 
> 
> As your note points out, we both seem to be speculating as to what the 
> CIX AUP is (for the sake of our discussions we should probably use 
> NSFAUP and CIXAUP).  Perhaps we can get a copy of the CIX agreement and 
> a full explanantion of the CIX filtering service before we get much 
> further in the discussion.   

There is no CIXAUP, you know that.

The "filtering service" is due to one oddball connection to the CIX under a 
different agreement then anyone else ever signed or ever will, with ANS.  
This agreement with ANS was signed due to a member of the board of the CIX 
believing that it was "good for the community" to have them participate at 
any level "way back" in 1990.  That agreement was a bad idea, and will soon 
run out.  I admit to my own moment of weakness during that whole process in 
letting it happen.

No one other than ANS has ever been filtered since the beginning of time.

So we will (soon) be back to the model of everyone's CIX agreement being the 
same, ie a real even playing field of peer organizations.
 
> 
> You did not quite make it clear in your note why being CIX connected is 
> now the deal.  It seems odd to me that everyone has to figure out how 
> to get connectivity to the Wiltel POP in San Jose and pay some 
> organization to consider themselves connected to the Internet.  
> Many Internet providers (including many CIX members) are also working 
> on this thing called the GIX (global internet exchange) which seems to 
> have a sole focus on Internet interconnectivity with a minimal amount 
> of politics. 
> 

I find it odd that you would take such a position.

You know the difference between the CIX and the GIX (aka meta-NSF NAP), but 
let's review it here

(1) the CIX agreements says that you MUST swap traffic with the other members 
on a non-settlement basis and that any and all traffic is just dandy.  I 
believe you have cast this as "the coercive model", and yet most all of the 
commercial providers participate (except ANS CO+RE of course)

(2) the GIX/NSF-NAP model is that you attach to the media, but you don't have 
to talk with anyone.  In fact there are a number of examples of this to date, 
and I believe more in the future.

Now (2) is the offical goal of the NSF solicitation and one would assume the 
goal of the red/white/blue Big Brother "system". Why?

Having lived in Troy (NY) in the past, I am somewhat sensitive when people 
roll up a big wooden horse to my door and say "this is for you".  I'm sure 
its just a personal problem.


I think (2) is the open invitation (through time) to balkanize what is the 
power and the glory of the original Jeffersonian Internet.  What it 
provides/allows/encourages is pair-wise relationships where reachability is 
in no wise guaranteed if you have "a Internet connection".  This is X.25 
service throughout all of its history.

What the CIX did was establish a Virtual Private Internet (VPI) where the 
politics (esp AUP politics) did not play and the commercial traffic could 
happen.  The assumption is that the NSFNet was there to provide the "net" for 
the rest of the (assumedly) non-commercial world.

This equivalent to having private medical insurance in the US (the CIX) 
believing when you hit 65 Medicaid (health care net for the Seniors) will 
kick in (the NSFNet).  In fact this is how PSI runs its network, every bit of 
data we can throw into the CIX we do.

You have rearchitectured the NSFNet so it is no longer a "net" for everyone 
to fall back on.

I have already heard a number of providers saying we will not be talking to 
XYZ because of ABC, when we get to the NAP model.

The "CIX Filtering service" is a single snow flake compared to the snow storm 
you have set us up for in my estimation.  The NSFNet solication with NAPs et 
al establishes an architecture for the Internet (you should have stayed out 
of architecture), that architecture has a major change and influence for 
balkanization.

So why?  Are you being hoisted on your own petard or are you hoping that I 
will do that to myself?

Marty



home help back first fref pref prev next nref lref last post