25 Cents to Stream a DVD Quality Film > You can't compare the two reports!

You can't compare the two reports!

ISP infrastructure has multiple uses. While corps like to say that the cost of, say, a single phone call includes the price of building the transatlantic cable, the fact is that a large part of the infra was built for other uses and is already paid for.

In Bruce Sterling's Hacker Crackdown, he cites a legal case where a company claimed that a document cost the _entire value of the minicomputer it was written on_ but the defense showed that the "confidential" document was available from the company's own store for $10.

Of course, I haven't read the reports either, just saying. . .

Well, that's true to a point. But real work vs. investment vs. revenue is all a very tangible, real world cost.

Like your computer example. If during the entire period that computer was in use, it only was used to produce that one document, their argument would have been just.

Broadband roll-outs cost their owners billions of dollars to deploy, any way you cut it. It's true that the vast majority of the infrastructure was in place, to some extent, but it is still very, very costly to adapt that infrastructure to support broadband. And then we have other situations, like FIOS, where the entire infrastructure is gutted and re-engineered. Now, each piece of equipment involved has a real cost. A DSLAM, for example, has a finite tangible cost. We can buy one and know what we paid for it. It will have a usable life span before it is outmoded, replaced, upgraded, etc. During that life cycle, it served X customers who paid X dollars for their service. We can look at those amounts and know what the real cost of that device was to each subscriber. Then we can look at how much data on average each subscriber used over that period and know how much it cost for that DSLAM to deliver a single packet to each customer. If we apply that same logic across the entire industry, to the bandwidth purchased by the ISP, to the equipment constantly rotated through the ISP, to the original cost of the original deployment of the BB infrastructure, and so on and so forth, we can put a real world cost on every aspect of Internet service. From each city, to each RT/node, to each block, to each customer, to each gig, to each packet, all the way done to the zeros and ones.

But what most people fail to realize is that with DSL, or with Cable, a part of the cost of providing Internet service is subsidized by other, more profitable services sold in conjunction with BB. Even if you come to the conclusion that a DSL service is sold at a loss, it is still profitable because it is sold along side POTS.

In the IT world, every part of our budget making and job costing must include life-cycle planning. The BB business is the same in that regard, but in many ways, much worse. In IT, we are not forced to upgrade our servers and networks every few months to keep up with the Jone's like ISP's do. We are not constantly undercutting ourselves, buying customers, just to get them under us. Consequently, Internet service provision is a revolving door of costs.

You could consider how much the ISP pays to push that data down the pipe to your computer, but that's not really relevant. You have to consider the ongoing cost of the entire infrastructure to find any kind of relevant real-world cost. When you look at that and see the big picture, it's a very large number. Much the same way that driving 20 miles doesn't just cost you in fuel, but also in car payments, service, tires, etc. putting the real-world cost of that 20 miles a lot higher than you'd think.
--
Never argue with a fool, people might not know the difference.