Government regulation forms one bubble, business interests another and technology the third. You can argue about any one of them with regard to telecommunications but you'll miss a huge chunk of the picture if you don't include the other two.
Telecommunications is possibly the most interesting area in IT today because of that confluence, broadband sitting slap-bang in the middle.
It's a political football at the moment. Well, it would be if any of the non-governmental parties had anything to say on the matter. As it is, they're deeply opposed to government intervention, but oddly quiet on alternative strategies. Regulations have been introduced and are being imposed even as we speak -- for better or worse. I guess they're just all waiting to see what happens next. Meanwhile, the government has tens of millions of dollars put aside to ensure rural and regional New Zealand gets some kind of broadband offering.
Technology is just as interesting an arena. DSL v ISDN v fibre v wireless. Should we be calling 128kbit/s broadband (short answer: no) and will 512kbit/s be enough for schools for the future?
When you mix in business interests, you get a huge bun fight. Telecom in the blue corner, TelstraClear in the red. Walker Wireless prowling on the edge along with Vodafone. An ISP market that seems at times to be caught in the headlights of change.
Ericsson's regional product manager for multi-service networks, Colin Goodwin, was in town recently for the Telecommunication Users Association (TUANZ) broadband conference in Nelson. We talked about the demand curve for bandwidth. He pointed out that if you draw a line through the "comfort" zones of bandwidth you get a logarithmic progression. People were happy enough once with 9600 baud for bulletin boards and the like. Then came the joyous 14.4 modem, followed by the 33.6. Today most would be reasonably happy with 1Mbit/s, says Goodwin, which sounds about right.
Draw a line through those years and you'll see that every decade the demand for bandwidth to the desktop increases tenfold. So, 1Mbit/s today, 10Mbit/s in 2012, 100Mbit/s in 2022. We've got 30 years to get the network up to speed to cope with demand around the 1Gbit/s speed. We're not going to do that on copper, I don't imagine.
If you think this is a tad far-fetched, you might be right. I'm leaning towards the opinion that it might happen sooner, having spoken with Chris O'Connell from Wellington-based Radar Guidance. He presented findings from a survey of customers who are using City Link's metropolitan area network. It showed most are happy with 10Mbit/s but would really like 100Mbit/s. If that's the case, all bets are off as to how long it takes us to reach the 1Gbit/s mark.
I can hear the scoffing now. What would you do with such speed? What's the point? If the server at the other end of the connection is still slow or congested, it doesn't matter how fast you go, you'll still hit the wall.
That's not the point. The point is that when those kinds of speeds are available people will find a use for them.
O'Connell talked about one Linux geek on the service who regularly boots his PC from the latest build of GNU. The kind of capability that lets you do what you want when you want changes everything: storage, application ownership, content provision, you name it.
O'Connell says this could be the end of the ISP as we know it. Broadband could be forcing this interesting proposition on us. The ISP's original function was to provide that always-on connection for users to dial into. If the customer is always on as well, what does the ISP do? Can't I just buy my bandwidth directly from the cheapest vendor on the day? I could host my own mail and web server and whatever else I need, so why not?