Re: An Attempt at Economically Rational Pricing: Time Warner Trial

From: Mark Newton (no email)
Date: Sun Jan 20 2008 - 18:06:12 EST

  • Next message: Andy Davidson: "Re: Lessons from the AU model (was: An Attempt at Economically Rational Pricing: Time Warner Trial)"

    On 21/01/2008, at 7:53 AM, Jeff Johnstone wrote:

    > All of these discussions ignore the developments taking place in the
    > consumer electronics marketplace. A quick glance at this years
    > consumer electronics show in Vegas shows a HUGE variety of home,
    > mobile and automobile consumer devices using IP services. These
    > devices, AppleTV as an example, will require large bandwidth
    > commitments.

    Sure. But it isn't an ISP's job to provide below-cost infrastructure
    to subsidize the bandwidth requirements of Sony and Apple.

    Pricing should be set without reference to the application developers.
    If the application developers end up building applications which are
    too expensive to use, then that's their loss. Nobody in the service-
    provider industry should be going out of business because they
    can't afford the infrastructure needed to give their AppleTV users
    20 Gbit/sec ports when Cupertino comes out with HHHHHD Video
    in a few years time.

    > Add this to IP based telephony and you can't just "shut off" a users
    > service after they reach a cap, you would be removing their
    > emergency services access.

    Why can't you? We do it all the time. You shape to 64kbps, which is
    more than enough bandwidth to run a SIP session with a low-end
    CODEC. Phone calls still work, hardly anything else does. (performance
    on a Virtual-Access interface with a rate-limit on it is way worse than
    performance on a BRI interface without a rate-limit because rate-limits
    lead to tail-drops, which spooks TCP very, very badly. So 64kbps is
    actually worse than it sounds for TCP applications, but constant bitrate
    CODECs deal with it just fine)

    > Hopefully we won't be seeing "basic" internet services of a couple
    > of gig per month and "channel" offerings of AppleTV, all you can eat
    > as "tier 2 plan", or "other service" as "teir 3 plan".

    You guys seem to be behaving as if this stuff hasn't happened before.

    No, you won't see "basic" internet on a couple of gig per month. You'll
    see "basic" internet on 40 - 60 Gbytes per month, which is more than
    most mortals use in any given 30 day period (like, ferchrissakes, who
    needs 2 Gbytes per day, day in day out? Grandma certainly doesn't
    use that much when she's checking her email...)

    > Alongside this discussion is AT&T's direction of content censorship
    > and its impact on end users.

    No, the two issues are completely orthogonal.

    You're supposed to be a network operator, stop thinking like an end
    user.

    > Our help desks are going to take a huge hit in the future as we
    > start trying to troubleshoot issues where the general rule of "I'll
    > pass any packet I get" becomes "I will pass any packet I am payed
    > for and have dissected for content, and after I have determined that
    > the rest of the stream won't push my end user over his network cap
    > this month".

    No, metering means the network neutrality debate undergoes gravitational
    collapse, and caring about what's inside a packet turns into a total
    waste
    of time.

    You don't need to care about whether a packet has been paid for because
    you know that every packet has been paid for. That's what metering
    delivers.

    Why do you think the DPI vendors haven't had much traction outside of
    Europe and North America? It isn't because the rest of the world can't
    afford them. It's because Europe and North America are where all the
    "unlimited" access services are sold, so they're the places where DPI is
    actually needed. Would you need to spend millions with Ellacoya or
    Sandvine if your customers imposed their own self-created backpressure
    against P2P usage?

    Again: Some of the significant economies on the Internet have done this
    already. The TWC paper isn't trailblazing, and every issue you can
    think
    of to explain why it'll be horrible and won't work has already been
    demolished
    by real-life day-to-day business in other countries. You guys who are
    behaving as if the sky will fall are going to have to explain why the
    Internet
    industry hasn't ceased to function in .au and .nz before you get on to
    explaining why its collapse would be inevitable in the USA.

       - mark

    --
    Mark Newton                               Email:   
      (W)
    Network Engineer                          Email:   
      (H)
    Internode Systems Pty Ltd                 Desk:   +61-8-82282999
    "Network Man" - Anagram of "Mark Newton"  Mobile: +61-416-202-223
    

  • Next message: Andy Davidson: "Re: Lessons from the AU model (was: An Attempt at Economically Rational Pricing: Time Warner Trial)"





    Hosted Email Solutions

    Invaluement Anti-Spam DNSBLs



    Powered By FreeBSD   Powered By FreeBSD