[2295] in Commercialization & Privatization of the Internet
Re: Connecting to the Internet outside of the US?
daemon@ATHENA.MIT.EDU (Jeffrey Shapard)
Wed Feb 5 15:45:48 1992
Date: Thu, 6 Feb 92 05:39:36 JST
From: Jeffrey Shapard <kddlab!twics.co.jp!jefu@uunet.UU.NET>
To: com-priv@psi.com
X-Vms-Mail-To: UUCP%"com-priv@psi.com"
I have been lurking in the background and following the discussions
in this list with great interest for, oh, I guess a month and a
half or more, and I have been rather disturbed at what I feel is
a provincial and narrow view of the internet as a domestic
American affair. Yes, I know that this list was spurred on by
the NREN initiative as well as by what is perceived as rather
unbalanced competition between ANS CO+RE and the independents
and regionals collaborating in CIX, both of which are pretty
much American issues politically. But the implications are far
more global, as what happens, and what does not happen, in that
large sector of the _world_ internet that is geographically and
politically American affects people throughout the networked
realms of the planet.
For example, I am an American living and working in Japan, where
I operate one of only two public access sites in this country
that have any sort of internet email or netnews links. (For
reference, if you are interested, we subscribe to the DASnet
Connection forwarding service of Campbell CA US, to the InetClub
gateway operated by KDD Labs in Japan, and are a registered
domain on JUNET in Japan. We use uucp links as we cannot justify
what are now rather high costs for "buying into" the WIDE Internet
project and establishing a proper IP link, given the "normal"
restrictions on the nature of traffic, as is currently being
discussed in this thread.) Now, while most of you probably
enjoy the feeling of the cost for internet resources being just
that of the cost of your bandwidth channel, which is probably
paid for as an operating expense by your organizations, the
costs for all traffic TO AND FROM sites in Japan must be borne
by those sites. Those sites which are in the ac.jp and go.jp
top-level domains and are subsidized to some extent by
grovernment grants can make use of the high-speed channels that
are also subsdized. Those sites which buy into the WIDE Internet
project and have IP links there can also use the subsidized
high-speed channels, for "research" traffic only, as buying
into WIDE means providing a big chunk of the subsidization,
which is not an unreasonable policy at all. BUT, those of use
who fall between the cracks and make use of the "oficial"
gateway for everything else get billed for traffic TO AND FROM
our sites, based on the amount of time it takes to squeeze
the data through the Trailblazers on the dialup uucp links
between KDD Labs and UUNET or KDD Labs and EUnet.
My complaint is not that KDD Labs are ripping anyone off. Quite
the contrary. They provide a necessary service and only bill
for telephone time, plus a small monthly overheard (6000 yen
or a little under US$50) for administration. They contribute
personnel and resources to the service, and provide very clean
monthly itemizations on a per message basis to each site. My
complaint is that THERE ARE NO MECHANISMS FOR INTER-SYSTEM COST-
SHARING AND BILLING on any of the major hub connections in the
US or Europe, or at least none that I know of. So, we have to
bear an unfair burden and inequal share of the costs. Such is
the price of connectivity in the current State of the Art.
When I sent you a <gasp> telex, I pay. When you send one to
me, you pay. When I call you on the telephone, I pay. When you
call me, you pay. Or you can call collect. When I access your
host computer through an X.25 circuit of a public data network,
I pay. When you access my host computer in the same way, you
pay. When I send you email through a commercial email service
like MCI Mail, AT&T Mail, Dialcom, etc., I pay. And when you
do likewise, you pay. Telex carriers, telcos, PDNs and even
a lot of the commercial email providers have extensive cost-
sharing and inter-system/network billing arrangements, and the
mechanisms to support this are even built into most of the
protocols used.
So, why is this overlooked in the RFCs? Why are there such
assumptions that the whole internet show is a national affair
or something that will extend no further than the realm of
subsidized research activities? Why is there so little work
being done in accounting mechanisms, or cost-sharing and
billing structures, to make it easier to administrate things?
Why should I have to pay for email that you send me, or bear
part of the costs of the "COMBITS" on sessions that you
originate? Why are the models, such as they are, of telex
carriers, telcos, PDNs and commercial email providers just
ignored in the discussions here so far (or have I missed
something)?
And... how can you really imagine that NREN and ANS CO+RE and
CIX are only relevant to internet traffic within the USA,
when most of the world internet traffic today passes through
some segment in America at some point along its journey?
Please think globally. Computer networks transcend political
boundaries and those little abstract lines drawn in between
the different colored areas on maps.
Jack Haverty, thank you for starting this thread and spurring
me into unleashing what I have been stifling so far. (I hope
you don't feel you have created a monster. <smile>) And since
I am paying for our traffic across the ocean, feel free to
talk to be about business or research. We are a commercial
domain, but American networks being what they are, they
probably only recognize ".com" as "true" commercial domains,
and any of those other pesky two-letter domains are ignored.
Glenn Kowack, thank you for your perspectives from the EUnet
realm. So far I have seen nothing from your associates who
administrate JUNET or WIDE, and I wonder if anyone else in
our '.jp' domain is on this list, or even aware of it. (Ah,
but so much electronic insularism lingers on in the virtual
seas of our network world.)
Bob Sutterfield, thank you for bringing in the issue of
applications which fit neither into the RE or the CO categories,
such as they are, but which could bring benefits to many
through the potential offered by this and related technologies.
You may be aware of some of the work being done by the Institute
of Global Communications (igc.org) and their partners in the
APC (apc.org) network in various parts of the world to spread
this medium, especially into those regions which can gain the
most benefit. I believe I have seen notes on this list by
folks there, and they are doing good work indeed.
Sean Donelan, thank you for your perspectives from Canada
about the confusion caused by all these RE vs CO distinctions.
And thank you everyone else if you made it all the way through
my textual rampage. The discussions on the list are relevant
and critical. I have learned a lot.
One little request: If you reply to anything I have said in
this rather long note, please drop the quotes or edit severely.
We get charged an average of 45 yen (c.US$0.30) per 1K, and
all that extra bandwidth gets real expensive. Thank you!
--jefu
* Jeffrey Shapard <jefu@twics.co.jp> "Connectivity *
* Connect-activist and Operations Director is our biz..." *
* TWICS/IEC, 1-21 Yotsuya, Shinjuku-ku, Tokyo 160 JAPAN *
* tel: +81-3-3351-5977 -- fax: +81-3-3353-6096 -- DASnet: [DCTWCS]Jefu *
