[1910] in Commercialization & Privatization of the Internet
Re: Understanding Combits
daemon@ATHENA.MIT.EDU (John Gong)
Tue Jan 7 17:18:25 1992
To: lars@spectrum.CMC.COM (Lars Poulsen)
Cc: com-priv@psi.com
In-Reply-To: Your message of Tue, 07 Jan 92 21:01:01 +0000.
Date: Tue, 07 Jan 92 14:17:16 PST
From: "John Gong" <jgong@us.oracle.com>
> In message <9201071921.AA23425@world.std.com>
> bzs@world.std.com (Barry Shein) writes:
> >If the end-user is not-for-profit, but the vendor providing him with
> >access is for-profit, is access for-profit or not-for-profit?
> [ and ]
> >Universities often sell "spare" computing time to for-profit
> >organizations. ... Would
> >anyone argue that such a machine's attachment to the internet provides
> >added-value to those for-profit customers?
> In message <9201072022.AA04264@psi.com>
> bill@tuatara.uofs.edu (Bill Gunshannon) writes:
> >Or, how about this little potential real-world example:
> > My daughter needs info for a school term paper. She connects to DIALOG
> >or some other Commercial Online Encyclopedia and starts extracting informati
on.
> >
> > Is it commercial or not?? Is it education?? After all, my daughter is
> >only in 8th grade. If this is commercial and would be billed as such, there
> >goes any concept of residential connection right out the window.
This is commercial, IMHO, since the information that is accessed is the
propriety of Dialog. If your daughter had the option of accessing the same
information from an R&E source, say a school or public library, then I'd
consider it R&E.
> It seems increasi> ngly clear to me that attempting to maintain a
> distinction between commercial traffic, which is paid for by end node
> subscribers, and R&E traffic, which is paid for centrally by NSF, is an
> endless rathole.
Amen. There was some discussion about changing this model about a year
ago. What happened to that effort?
Any effort to enforce this distinction is unworkable. The self-describing
by a customer to be commercial or acceptable use and labeling its network
as such is very flawed. I worked for a company who used the same network
number for both categories, both with substantial traffic to justify one
label or the other.
Enforcing a TOS option might be easily implementable at code level, but it would
take years for this enhancement to be used throughout the existing and
future TCP/IP community. Look how long it's taking to get BSD 4.3
universally implemented.
> I am beginning to think that it would be simpler to make the Internet
> backbone and mid-levels completely "commercial".
Hear, hear. Put the money into providing more accessible services and
information. Then customers will decide whether they're willing to pay
for it (they will) and how fast/how much they want (pipe size-oriented fee
schedule).
> If the present NSF subsidy were divided between the subscriber
> institutions according to their present usage, and each subscriber were
> free to select the service provider with the best value for the money,
> the market would sort itself out pretty fast.
> --
It seems to me that much of this discussion has centered around how NSF
spends its money. From a free market standpoint, the CIX affiliates have
demonstrated some downward pricing from competition or economies of scale,
as pointed out by another writer. The EFF and other folk interested in
enlightening the masses seem to want to see the NSF monies diverted to their
causes. Can someone describe how they would like to see a grassroots
effort evolve? $19/mo. for dialing into a CIX affiliate for email and
USENET (tm) and other future services sounds pretty cheap to me, already.
I wouldn't advocate government subsidies in order to provide the same
service for, say half the cost to the general public. This government
gives out a lot already. I don't want an information welfare system
put in place. And really, I'm not a high tech elitist. I just think
there's a limit to how low valuable services should be priced.
> / Lars Poulsen, SMTS Software Engineer
> CMC Rockwell lars@CMC.COM
John Gong jgong@us.oracle.com
