[ NNSquad ] Re: The Once and Future King: Multicast looks to (finally) be the future of television.

[Brett Glass <brett@lariat.net>]

At 10:57 PM 12/21/2007, Bob Frankston wrote:
 
> I've long argued against
> multicast -- it's based on the assumption that we are broadcasting in the
> same way the phone network reveled in circuits.

I agree with this point, Bob, though my arguments are a bit
different.

 From an economic standpoint, by far the most cost-effective and
efficient way to broadcast data to a large number of receive-only nodes
is terrestrial or satellite-based wireless. It covers the largest number
of subscribers; there's plenty of bandwidth; and the technology is now
extremely well developed (though it wasn't bad even in the days of black
and white TV). 

Multicast over IP networks is nowhere near as efficient. IP, by its very
nature, was designed for two-way, interactive traffic consisting of
relatively large data packets. It's very non-optimal for broadcasting.
(It also wasn't designed to carry telephone conversations. VoIP is a very
high overhead way of conducting a conversation compared to ATM, which was
designed for voice and is suboptimal for data.)

> The OSI stack is just a
> vocabulary ? it isn?t really but the ITU is so besotted with protocolism
> that they give fictions like spectrum allocation all the weight of a ton
> of lead.  Who else would think bandwidth was a useful measure ?
> what?s the bandwidth of my copper at home. That?s a stupid question ? the
> answer is whatever I can achieve and that?s now a gigabit at $10/port
> when they can?t get their costs below $1000! 

Bandwidth becomes an important concern when someone needs to be paid to
engineer the network, hook equipment up to the fiber or copper, climb up
radio towers and mount the radios, and then maintain the network and take
responsibility when it goes down.

> You can?t make a profit for
> selling transport any more than you can make money operating a canal
> across the ocean 

That's funny; shipping lines make pretty good money. Why? Not because
they're "operating a canal across the ocean" but because they
know how to build, maintain, sail, manage, organize, load, and unload
big, efficient boats. And that does take fuel, experienced captains (the
Exxon Valdez being a great example of what happens when the captain is
NOT good), labor, organizational skills, and smarts. Just as individuals
don't sail their own small boats to Taiwan every time they need an RJ-45
connector, they need folks who know how to build and maintain the
telecommunications infrastructure. There's tremendous social value in
that. We're just a small WISP, and the appreciation and loyalty we earn
from our customers is amazing. 

That's what motivates me to go out early on a frigid Saturday morning (as
I am today), driving down Colorado's Front Range to buy used and surplus
equipment from another WISP. I'll recondition it, sell it at a slight
markup, and save my customers $30 per antenna and $100 to $150 per radio
unit. None of my competitors can touch the prices I can offer my
customers. Could they get the bargains I get for them by buying large
lots of used gear (at auction or via private deals) and then using 15
years of experience and expensive instruments (such as microwave VSWR
bridges) to test and recondition it? Nope.

> The solution is not to fix the
> network nor V6
> (
> http://www.frankston.com/?name=IPGeni2 and
>
> http://www.frankston.com/?name=InternetDynamic). It is to give us the
> ability to do our own networking rather than having to pay for a ride
> from the today?s robber barons.

While there are some robber barons in the telecomm industry --
particularly the large monopolies-- there are also an awful lot of people
(like our WISP) who are not. It's not responsible or reasonable to lump
us all into the same basket. If you do, you'll make poor policy
recommendations and decisions.

--Brett Glass