Thursday, November 17, 2011

The Next Internet: What's Holding Us Back?


MCI logo (original).gif
Tim Berners-Lee, to his credit, did not invent the Internet. He did have one good idea. He was not the first person or even the twelfth with the same idea, but he did make it work. Yet most of the underlying work - the bringing together of dozens of communications systems with slightly or wildly varying protocols - was done before him. He just plugged it in, and for that, he gets most of the credit.
What made the Internet, and thus the Web, possible - the thing that, without which, Tim Berners-Lee would still be watching reruns of "Eastenders" - was a decision. The major carriers of electronic mail, whose business it had become to route messages to each other's members, collectively reached a truce. They decided that the long, endless fight over who has the biggest volume, the longest distance, the fastest network, so that one could charge the others more postage than it was being charged, was too expensive and was stifling progress. They decided to call off the war. I know. I was on the phone with them the moment it happened.

The boldest stroke of all

The date was April 25, 1990. My assignment was a story on the subject of universal e-mail addresses - the notion that someone could have an address that is independent of CompuServe, GEnie, Delphi, or Prodigy. The protocol these addresses were based upon was an international standard called X.400. That standard not only specified what we now call the "domain" for each e-mail recipient, but also the manifest for the settlements of any charges between one carrier, such as GE Information services, and another to which it might hand off the message, such as MCI Mail.
I asked one of GE's lead engineers, Kenneth Murphy, how X.400 made sense of all the various exchanges an e-mail had to make on its way to its recipient. To my astonishment, Murphy explained that, if the meeting going on that day was successful, X.400 would not even be necessary.
Here is how I presented the story for the May 1990 issue of one of those printed computer magazines that is now the stuff of landfills:
The way the structure is being developed, an E-mail message is, to some extent, a property of the service on which it was composed; it will then be bought, under contractual arrangement, by whatever service the recipient chooses for reading that message. The legal issues are rather complex, so I'll let Kenneth Murphy elaborate: "Let's put some real names to this - I'm going to use AT&T [as merely one example]. AT&T Mail and GE Information Services have negotiated an interconnection agreement, which was announced last Monday [23 April]; it means all of their subscribers can send and receive messages to and from all of our subscribers. I haven't a clue what an AT&T Mail user enters at the 'To?>' prompt, in whatever front-end they've got; but presumably, they should be able to address me or any of our other subscribers without too much difficulty, as long as they know the correct way to format the address. They will pay AT&T Mail charges, whatever they may be. I, as GE Information Services, will deliver what's handed off to me, to my subscriber free of charge.
"Today, there's a concept of Settlement of Accounts. The interconnect agreement is a legal, contractual document between the two of us. It covers a lot of stuff, but one thing it specifically does not cover - and this is temporary - is the issue of settlement. Settlement means, I want reimbursement for my cost of delivering a message to my subscriber on your behalf, you being AT&T. Therefore, we need to finish this process and agree on some monetary values for messages, and then we just keep track of those accounts, and at the end of a period of months/quarter/year, we will settle the difference. That is a process that has long been in place for the interconnected postal services, telex networks, telephone networks, public data networks."
...
Since there will only be a handful of E-mail services, they may all decide to settle accounts with each other just as GE has with AT&T; so every E-mail provider will be connected to every other - what then? X.25 [the international standard at the time for packet-form communication] will make it possible for a user to make connections with more than one service; but if all of them are interconnected anyway, why should she want to? These are unresolved questions.

In fact, at the very minute Ken Murphy and I were talking, an international meeting was taking place regarding these very issues. Murphy told me, "There's a meeting going on today [25 April 1990] up in New Jersey where the service providers of North America are trying to grapple with the problems, both technical and commercial, of implementing interconnected directory services using the X.500 standard. It's not going to be easy. Also, there is going to be a meeting in two weeks of the North American service providers, working with the Electronic Mail Association. It's a working meeting of editors to sit down and publish a user guide that can be distributed throughout the marketplace, on how the heck you get from one system to another, now that these systems are interconnected. What I see at the 'To?>' prompt is going to be very different from whatever you see, unless we're on the identical system. If we're on the identical system, we don't need X.400."
Nothing we do on the Internet today would have been conceivable, let alone feasible, had these corporations never decided amongst themselves, "You know what, the heck with it." HTTP protocol could be developed because there was no longer any need for a settlement of accounts, or a protocol to keep track of such a settlement.
Nobody thanks GE, MCI, Sprintnet, UUnet, and the other names from another era buried in the rubble of our selective history, for the one bold stroke of sensibility that inevitably transformed our world. Every innovation brought forth since that time, including the one with the hyperlink thingie, resulted not so much from inventors or business titans or shrewd negotiators but from battle-weary pragmatists. They knew the real revenue would eventually come from the cavalcade of new services enabled by a unified digital communications platform, and dickering about how the seams should be sewn was trite. It was the best decision to quit ever made.

The long, grueling road to par

Today, we stand at a similar precipice of history. There are a handful of wireless carriers, a handful of Internet search providers, a handful of device manufacturers, a handful of technology portfolio managers. Each of them has the keys to something about the Internet that all of them require to run their business. The potential in front of them is an all-inclusive service of unlimited media from multiple sources: the world stage in one's pocket.
Yet rather than decide it'll all come out in the wash, each player tries to amass more intellectual property than another, driving up the value of their patents. Google buys Motorola because it couldn't buy Nortel, and someone may yet buy InterDigital, maybe to keep it away from Qualcomm, etc. The objective of this game is for one player to be paid the most penance and crowned king - an endgame that never comes.
An agreement between these parties on the order of the e-mail services agreement that forged the modern Internet, is impossible. The reasons why are in front of our faces:
1. Egos. The major players have become like sports teams. They're in front of the spotlight all the time, and we, the people, follow their moves and even place bets on them. Like pro wrestling sometimes, certain players wear the villain's hood, and others the gold belt of champions. We consumers have our emotions vested in this battle to a greater extent than before. In so doing, we've inflated the egos of the people in charge. They don't want to let us down. We've seen what happens when competitors reach covenants with one another (Microsoft + Novell, Microsoft + Sun). Covenants, some claim, are cloaked conspiracies to defraud consumers. (Everything, some claim, is a cloaked conspiracy to defraud consumers.)
2. Lack of faith. No one is certain any more of the potential value of the future services that the expansion of wireless Internet platforms may yet provide. Today's services appear to be undermining the foundations of many once-formidable industries - recording, publishing, news-gathering, advertising, porn. Not only is there no guarantee of a brighter future for content any more, there's no clear sign of it. Those whose futures are completely tied up in content are forging business models based on sci-fi constructs like content farms and self-repurposing. So there's no incentive for corporations to forfeit their revenue from licensing, in hopes of competing in media.
3. Debt. The cost of having built the telecommunications infrastructure that's already in place, will probably never be recouped from consumer services alone. It costs carriers such as Verizon as much as $4,000 per customer to deploy services from which they may expect to see, over their service lifetime, about $1,500. Consumers are not willing to pay what the Internet actually costs. Thus carriers must recoup their investment elsewhere, and licensing - with the litigation that comes with it - is the most convenient alternative.
Assuming that every merger and acquisition that could ever be made has been made, and every intellectual property infringement and licensing disagreement lawsuit that can be pursued has been pursued, the eventual outcome is exactly the same as the one the e-mail engineers foresaw back in 1990. It can be described with one of the simplest words ever given to English by the Latin language:par. That these unknown and, as yet, unheralded people could foresee this simple truth and their successors are blind to it, says everything about the history of technology in the last 21 years that anyone needs to know.

No comments:

Post a Comment