North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: Peering versus Transit

  • From: Vadim Antonov
  • Date: Mon Sep 30 22:07:16 1996

Sanjay Dani(maillists) <[email protected]> wrote:
> 
> The Web page owners pay for that, ok?  If they do not want other people
> accessing their data they wouldn't put it there for open and pay for it, too.

>I didn't know access service providers are not paid by their
>web browsing clients.

The cost of transaction is split between content providers and
clients.  Is this that hard to understand?

Everybody placing information on Web _pays_, like those people
who place ads in magazines or newspapers.  That does not mean
magazines must be free.

People actually pay more for ads in popular places, ok?

>I've a public telephone number, that doesn't mean you can keep
>calling the number as often you want.

Heh.  That's an example of how stupid technology is being
preserved by stupid legistlation.  The laws regarding POTS are
generally legal fixes for techincal problems (like inability
to block annoying calls).

>You have conveniently ignored to address the main issue of how
>bandwidth and traffic usage should be charged as the nature
>of network usage changes.

The issue of per-connection charges on Internet is dead.  It is
not technically feasible due to fundamental scaling problems.
The volume of connections on Internet surpassed what multimillion
telco boxes can handle years ago.

>We are not arguing technical issues here.

Yawn.  Sure.  Let's argue the benefits of perpetuum mobile.
"Technical issues" are the hard boundaries between possible
and just blabbering.

>I do not like the model where established players will
>contiue to pay only for telco costs, but not IP traffic
>costs to anyone else whatsoever even though their clients'
>usage is affecting someone else's network and their customers
>pay them for the ability to browse contents elsewhere, and
>therefore are effectively subsidized to some extent by what
>a relatively small ISP content provider pays for IP traffic.

Cool.  Why won't you write a business plan and go ahead with
your new pricing model.  You should be able to convince more
than few VCs, it's a hot stuff nowadays.

FYI -- i had a more than an engineering role in service providers
covering together all time zones on this planet.  One of them
was the first ISP in 16-odd countries.  Another was the first
commercial DS-3 backbone in the world.

>There might very well be a caller-pays-for-the-session model
>or settlement charges based on who initiates how much traffic.
>I don't know.

So why you just don't learn why the present system is here,
and why it delivers while others only promise.

>I look forward to competitive offerings from
>big providers of the kind Sean Doran envisioned in his
>essay.

They are already here.  Just ask somebody more clueful than a
regular salesdroid who doesn't know Internet from Istanbul.

>Sure, start labeling when you run out of arguments. A certain
>losing politician could use your help.

It is not labeling.  It is diagnosis.

--vadim
- - - - - - - - - - - - - - - - -