>I get most of the reasons that java overcomes limitations with bandwidth,
>but I was hoping you could explain a little more about why unlimited
>bandwidth makes java obsolete. Also, could you please explain a little
>bit about Teledesic and why satellites promise unlimited bandwidth.
There is a growing urgency that desktop computers become easier to manage and increasingly network-centric. Networks serve dual roles as (1) the enabler of the transfer of information between nodes and (2) the storehouse of application software and parameters. With decreasing system costs of desktop computers, unit demand will expand into corporate WANs and the consumer space (home, cars, hotels)-as long as they can be easily network and managed.
It should be obvious that Bill Gates views low-bandwidth connectivity exactly like he viewed desktop computers in the x286 days. Recall he ignored critics completely when he built a bulky, blunderbuss of a GUI called Windows and insisted that the world overlook limited processor speed, high costs of disk storage and crappy code. We now know what Bill Gates learned during the limited time he spent at Harvard: doubling speed every 18 months, and reducing storage costs by half frequently, would provide cost-effective computing power about the time his lethargic Windows programs could be made adequate. This stroke of genius and doggedness made him the richest man on earth.
Today Bill faces a different problem. He has to grow his monopoly into less receptive areas, namely the consumer and corporate WAN, while also defending his monopoly against a "balance of power" assemblage of technology giants. The giants are slowly realizing that the NC solves the problem of managing desktop computers even with current-day low-speed connectivity available to most consumers and corporate WAN users. In other words, the NC is perfectly designed to gain a foothold and spread rapidly in the very space Microsoft needs to continue growing. Even worse, if the NC were to succeed in these market segments, there would be little to stop the NC from penetrating deeply the existing monopoly, corporate PC networks and high-end users.
Bill is approaching his current problem exactly like he approached windows. Solve the problem under the assumption of infinite bandwidth. And get it done by the time higher bandwidth becomes generally available.
The difficulty managing PCs, in the office, on the road, or in the home, can be solved through software by the company that owns the desktop OS, and increasingly owns the OS on network servers. The solution will be difficult, and it will take time, but it can be done at least adequately within the timeframe broadband networking becomes available to consumers. This timeframe will be about the same as the time it took the x286 to evolve into an adequate processor, say five years or so. Broadband communications will be plentiful within that timeframe even if Bill Gates has to provide the service to the world, or even if he has to completely rejuvenate a dead-beat cable industry to speed the introduction of cable modems.
Think of Microsoft's WebTV and digital set-top box incursions as testing grounds for the management of consumer PCs, which may take the form of Windows CE or who knows what. All we know is that Bill wants the solution to be Microsoft's, not the technology giants pushing the NC.
Teledesic involves a huge number of LEO satellites and enough frequency and satellite smarts apparently to enable broadband communications anywhere on earth. The fact that the satellites will be close to earth means that signal latency will be negligible and ground antennas can be diminutive. Teledesic is distinguished from existing and other planned LEO satellite systems by its scope, measured by the number of satellites and the capacity of the system.
Allen |
