If the download cap fits...

Whoever said "Hell hath no fury like a woman scorned" never met a geek deprived of their inalienable right to infinite free data. Ask NTL, whose move to restrict their broadband users to a gigabyte of downloading a day has produced more spleen than an abdominal surgeon sees in a lifetime. To be fair, NTL did it all wrong -- they sneaked the change in without properly telling anyone, they blustered and bumbled when caught out, and then said that they didn't mind unless you were being really silly.

But the reaction of the users was splendid. Huge Web pages, dripping more invective than you'll find on a Bin Laden media training course, sprang up almost instantly. Strikes were called, boycotts proposed, newsgroups filled up with outrage: you'd be forgiven for thinking that imminent war, exploding spacecraft, plague, famine and pestilence were party games compared with this final Horseman of the Digital Apocalypse.

What this proves is that naivety exists on both sides of the commercial fence. It's hard to credit NTL with the stupidity that informed the initial non-announcement, but equally difficult to believe that the users expect leased-line guarantees for a fraction of the commercial price for such things. Simple maths proves the impossibility of this: 100 1-megabit users maxing out their lines will require a dedicated 100-megabit backbone and will bring in £3,500 a month. That won't pay for the raw bandwidth, let alone the business of supplying the modems, connection and support while running a company. Anyone who thinks otherwise is welcome to make the business case and revolutionise the industry.

So, what was NTL doing selling such a service? Like any company, it assumes that most people won't use all of what they buy. No airline could survive assuming that everyone who books a seat actually turns up: it's the Coleman's Mustard principle, that you make your profit on the stuff the customer leaves in the pot. But broadband is a new and unknown market, and NTL had to pitch its offering so that it looks attractive and competitive but leaves enough wiggle room for the company to adjust its service in the light of what people actually do. What some people actually do, it transpires, is get really keen on amassing gigabytes of stuff. Movies, music, ripped-off software -- far more than any sane person could possibly actually consume -- these are the honeypots of broadband.

File-sharing is without doubt the crack cocaine of broadband -- outrageously attractive to users, but with a distinct link to criminality, fairly unpleasant side effects and when withdrawn provokes the user into outbursts of aggression and badly designed Web rant. Providers know this, and have no desire to negate the real reason so many people sign up. And most people use it sensibly -- we've been around the legal and moral aspects enough times to make a priest sick, but there's a great deal of pleasure in hearing an old song on the telly and grabbing it off the Net for a little nostalgia. It's the hard core willy-wavers, who aren't happy unless they fill a 40GB disk a month with their finds, who make the equation fall apart. If the action of one person upsets the life of 100 others, where does the problem lie?

The answer has to be in what is euphemistically known as attitude readjustment. It's happened in the airline industry, where people have learned that if you pay peanuts you get a ragged service -- with no peanuts. Ryanair has dispensed with all the slack that makes normal airlines run on time: every aircraft is worked to the bone, and if there's a delay then tough. There's no spare capacity to help make it up. There's next to no customer service, no guarantees and no flexibility, and while the passengers are vocal in their dislike of the effects they're queuing up to pay their money. That's what matters. And if you want to get better service and better reliability, then find yourself a more expensive airline.

So, NTL could have said "no caps on download, but no guarantees on throughput" from the outset -- in fact, that's effectively what they and everyone else offers, although it's never explicitly stated. Load balancing would then take care of the few hogs, who'd be free to go elsewhere if they didn't like it, while leaving most people better off, and the company would always be free to offer premium services to those who really did need their bits shifted in quantity and still wanted to use cable.

There's no doubt that Telewest and the DSL providers suffer from the same problems, and they're watching the NTL experience with considerable interest. If, as looks likely, NTL and Telewest merge, there's every chance that they'll take the opportunity to introduce a more sophisticated approach to service provision than the greediest gets the most. Pigs might fly, but they won't fit down the pipe for much longer.

For a weekly round-up of the latest broadband-related news, sign up for The Broadband Informer newsletter.

Let the editors know what you think in the Mailroom.