Karl Auerbach wrote:
> There is a need for coordination of things like TCP port numbers and other
> things that go on at the transport level and below. The IETF, W3C, and
> IANA have done a fine job of that. All they need is a bit of money to pay
> for a couple of people at IANA to keep track of this. These are
> non-contentious issues.
Assuming that the warm and cuddly 'openness' commitment and business models
of the most influential players don't change...
<snip>
> The issue that may require coordination is one with some really tough
> technical, economic, and political issues - inter ISP
> peering/transit/billing policies. It is unclear whether the current
> inter-ISP sitution isn't one being used by the "big guys" to squeeze the
> "little guys". On the other hand, the small guys aren't screaming that
> loudly (or I'm not hearing 'em.)
and here's a good example of how they could change. I agree, I don't hear
them screaming loudly either, but I'm concerned that we're in something of a
calm before the storm for two reasons:
(1) The implmentation of QoS is going to separate the ISPs into those which
can make the necessary investment and play with the big boys, and those who
are just quaint dial-ups offering 'plain old internet service.'
Standard-grade operators would likely die out for lack of customers
(particularly business customers), but even if they could play in the QoS
leagues, they would get killed on upstream settlement payments (if you
believe Geoff Huston, Rob Frieden, and Ken Cukier).
(2) AOL and @Home are showing that in the consumer market, brands count, and
the Internet on training wheels is actually something that is desired among
the general public. "The Internet" is a good enough brand for me, but there
are a whole lot of people out there who don't even know there are addresses
outside of .com, .net. and .org, much less .tj or .web. I agree that if the
.car domain is something that its customers want access too, then any ISP,
including @Home, would resolve .car addresses.
My concern is that with near-complete control over local access, @Home has
no incentive to tell the blinking-VCR-clock crowd that .car exists, but if
they worked out a marketing deal with .gm, .gm would be plastered all over
the start page. Think about what percentage of Web users probably don't
even know that you can change your start page. Think about how many AOL
members actually think AOL is the Internet. I know it isn't and everybody
on this list knows it isn't, but we're an infinitesimally small percentage
of the 'Internet market.' When I see Canada's largest bank and its largest
telco falling over each other to buy equity stakes in AOL Canada, I get a
bit worried.
Issues like peering/transiting/settlements and instant messaging suggest to
me that the current Internet model may be something of a relic of a
different past, which I'm less confident than others can survive its very
different future. So I'm trying to think about what makes the Internet the
way it is, what threatens that, and on what basis we might insist that it be
preserved. An Internet community which instantly devours anyone who
suggests there is anything more than autonomous, private action going here
may be dooming itself to domination by better-organized and better-funded
forces. Sound familiar?
Craig McTaggart
Graduate Student
Faculty of Law
University of Toronto
[EMAIL PROTECTED]
