[11080] in Commercialization & Privatization of the Internet
Re: ANS and the CIX - have they really connected?
daemon@ATHENA.MIT.EDU (George Herbert)
Mon Mar 21 19:41:01 1994
To: karl@mcs.com (Karl Denninger)
Cc: matthew@echo.com (Matthew Kaufman), com-priv@psi.com, cook@path.net,
fair@apple.com, stpeters@dawn.crd.ge.com, washburn@cix.org,
gwh@crl.com
In-Reply-To: Your message of "Fri, 18 Mar 1994 19:31:59 CST."
<m0phptP-000BbPC@mercury.mcs.com>
Date: Sat, 19 Mar 1994 12:11:43 -0800
From: George Herbert <gwh@crl.com>
>From: karl@mcs.com (Karl Denninger)
>Note: There's some heat in here. My blood boils when I read posts from
>people who are continually trying to devise a way around a voluntary
>associative agreement!
This is in some ways the crux of the problem. From the top down,
CIX as it is was intended to be and appears to be a volountary
associative agreement to keep settlements down. From the bottom up,
as Matthew keeps pointing out (and Dick seconded) CIX has set
itself up to be a cartel, with a buy-in level designed to prevent
any small fry competitors from setting up in an entirely legitimate
(i.e. guaranteed routing) manner.
>Either you're for us or again' us. If the CIX goes away tomorrow you can
>expect to be billed BY THE BYTE for your FTPs and email. DO YOU WANT THAT
>to be the model of the Internet for the future? The RBOCs and ANS would
>LOVE to be able to impose this. The rest of us think that idea sucks.
Small providers who are barred entry based on a fixed membership fee
but will have low traffic would benefit from such a settlement arrangement,
Karl, as offensive as settlements are to most of us. Do you really want
to form a crowd of anti-CIX small fry competitors who start lobbying
congress to help ANS's model out?
>The CIX <DOES NOT> block this backdoor traffic, as MANY people can attest
>to today who ARE cheating. However, INDIVIDUAL providers can choose to
>block that traffic UNLESS you have a separate business arrangement with
>each and every one of them.
So why is it that the CIX has to officially stand its ground and not
figure out some accomidation which will legitimize the "backdoor"
small-provider traffic (which the backdoor people would probably
love, even if their bills go up some) rather than insisting it's
not officially covered or ok and then looking the other way?
>The CIX is the best thing to ever happen to interconnectivity. Flat-rate,
>no settlement connectivity to over THIRTY other providers. Grouse all you
>want, but nobody is going to do this for free, and nobody has YET come up
>with a more efficient model to interconnect with that many network providers
>on a peer-to-peer, everyone's an equal, basis.
The CIX is the best thing to ever happen to national-level provider
interconnectivity. It also is a nightmare for local area providers
in relatively small areas. It essentially is saying that a certain
sized market area (say Santa Cruz, CA) doesn't deserve to get relatively
cheap internet access. Mr Kaufman can probably go national at about his
current pricing levels and a little venture capital; despite everone's
bitching about it, buying long distance leased lines (or frame entry
points or...) is not very expensive and is getting cheaper. It's quite
a bit cheaper than even most backdoor pipelines, so the costs will
go down a bit as he gets bigger. The CIX model forces him to do that
to be legitimate. Are you sure you want that? Does that make any sense?
The question is not whether CIX is the right model for large provider
connectivity, and I don't think it ever was. The question is whether
large providers are the whole market (obviously not) and whether CIX
and its members pretending otherwise is being a problem (apparently).
In addition to the question of how small providers can fit in, there are
minor little issues like antitrust laws and things like that, which
CIX is probably backing itself into a corner with. Which I don't want
to see, for one, as I think CIX is basically a good thing.
-george william herbert
Speaking strictly for myself
