Sorry, I've been meaning ever since to write a little article about my
thoughts on the NBN, and of course haven't been able to find the time,
what with all the slacking off I have to do.
It was pretty much what I expected. Peter Croll from SCU gave a little
preliminary PowerPoint presentation with, as anticipated, a bullet-
pointed list of random "E-" words. He went to great pains to establish
his geek credentials by waving around his iPhone 4 at every
opportunity, then spoke in some detail about the possibilities for
broadband "E-Health". Curiously, he thought the future was medical
practitioners wandering about entering patient data into iPads, and
patients then voluntarily putting that data into Microsoft HealthVault
(
http://en.wikipedia.org/wiki/Microsoft_HealthVault), saying that
thereby patients would retain complete control of their medical
records. I'm not entirely sure how giving your data to Apple and
Microsoft equates to retaining complete control, but I let that
reservation slide.
He showed a promotional video from some vendor of a proprietary health
monitoring system where a telegenic American couple spoke of how
grandpa (it's not made clear whose - if anybody's - father he is), a
beguiling silver-haired twinkle-eyed character, is sadly not as with-
it as he once was, so now they and his GP have to be able to spy on
him 24/7. Grandpa, for his part, is delighted that the system comes
with a game of solitaire, which he can play on his TV set from the
comfort of his La-Z-Boy. It was all very heart warming.
He then handed over to NBNCo's Glenn Holdstock, who waved around a bit
of cable and showed some isometric diagrams of little Monopoly houses
with black lines, representing these cables, running past them. At no
point did these black lines intersect, except at junction boxes, where
the thin black lines merged into thicker black lines on their way to a
thing called "backhaul", which leads to a capital city. This is the
detail I was, in my total ignorance, mostly concerned about. The
presentation, I think, was mostly aimed at mollifying any community
concern over unsightly overhead cables or streets being dug up; those
with long memories will remember the hubbub about the Optus cable
rollout. However none of the half a dozen or so people who turned up
to the community session were worried about that; most questions were
"I live out at such-and-such; will I get fibre, and when?"
I voiced my concern over who would ultimately be in control of the
network infrastructure and what were the technical and legal
constraints over what the community could do with it. I gave the
examples of say a geographically disbursed educational institution
like SCU or TAFE, or a local FPS gamers club. Both of these would have
a desire for as direct and low-latency a connection as possible, and
less concern for a fast connection to Sydney or a server farm in
California. I made the point that the people in the room were every
bit as much a part of the internet as Facebook or Google (as Douglas
Adams said "the Internet is just us"), and that a network optimised
for carrying traffic to and from a trans-pacific cable via an ISP in
Sydney will be de-optimised for carrying traffic between me and my
friend who lives round the corner.
If I were to rephrase my question for a tech audience it would be
"Bugger faster broadband; do we get packet-switching to the premises,
or can we use the hardware to implement a local packet-switching
network ourselves, which may or may not have a gateway to the
Internet?" Glenn Holdstock briefly looked like a deer caught in
headlights, apologised for being an economist by training rather than
a geek (though in my experience, you don't get geekier than
economists), and gave me the email address of somebody who could
answer my question (the short answer, for those who are interested,
turns out to be "no"). After adjournment, one of the bigwigs present
sidled up to me and said that of course the elephant in the room was
Telstra, but I think the problem goes even deeper than that.
As Paul Templeton pointed out to me after the evening business
session, it's not just a political problem of mollifying this one big
company whose copper wire network will shortly be redundant. If every
premises in the country has an IP address from NBNCo, who needs ISPs?
If you were given tens of millions of dollars to solve "the last mile
problem", or as I think it is more properly formulated, "the _first_
mile problem", how would you do it? I naively think the no-brainer
answer is to extend TCP/IP to the premises. NBNCo however is spending
this money to more or less exactly duplicate the existing
infrastructure we have with dial-up and ADSL, only faster, in order to
avoid disrupting the status quo and upsetting politically powerful
existing players. It's a network designed by economists for industry,
rather than a network designed by technologists for the community.
You are now invited to say "Well, DUH!"