The NC is not vaporware. Once serious applets are available, first for corporate intranets, then for everybody else in the world, the use of NCs will be significant. This is the first shot across the bow of the PC.
This is not to say that Oracle will corner the NC market, or even that the NC will always take the form of a near-desktop computer. What is the difference between a PC/TV and a NC, or a set-top box? Ultimately, nobody knows, but various devices of this type will proliferate and become commodities. A common characteristic of successful ones will be ease of use, which is why they will take on so many distinct forms and functions. While Oracle seems to be championing the Java-based NC, actually Oracle is interested in the whole software food chain from the server database and web interface, to Java applets, and ending with the client, or Network OS. Recall that Oracle was one of the first software companies to embrace the set-top concept, and must have been disappointed when its partners postponed implementation time and again. I suspect Ellison finally thinks he can control the outcome by making the interactive set-top box an internet machine instead. The essential communications infrastructure is already in place to begin commercial implementation. And once even moderately higher bandwidth is provided routinely to homes and offices (cable modems, xDSL, ISDN, wireless, or PowerCom) the NC can evolve to a fully-featured multi-media gateway (teleconferencing, phone, video, games, email, internet browsing, applications like word processing, spreadsheet or anything else that a PC can do).
The importance of the Oracle NC announcement is that WIND is now guaranteed to be a significant player on the client side of the internet/intranet business and commodity gadget world, and WIND will be playing with Oracle among others. By providing the leading RTOS for modern developments in network communications, WIND already is becoming a dominate player on the infrastructure side of the internet/intranet.
Buy the way, don’t you wonder if WIND’s involvement with I2O was a major reason for its selection over MWAR, who should have benefited from an established relationship with Oracle through their DAVID OS? I2O is a natural future extension to any multi-media internet device, if only to offload all standards and protocols for communicating all the rich data types soon to be served up by Oracle-residing servers (on Windows NT, Unix or IBM’s venerable mainframes), and to be able to manage updating protocols from servers. The usual justification of additional speed, by reducing the load on the CPU, is simply a plus. But since the $49 i960 RP chip is probably prohibitively expensive as an extra for commodity appliances, one expects soon that I2O processing will be integrated into the CPU as a parallel processor or simply implemented in software. But either of these approaches would require a custom development of the equivalent of IxWorks working in concert with the Network OS. Not an easy undertaking. The safest strategy, given the expected importance of data-rich, flexible, upgradable I/O in a commodity world, and given all the difficult possibilities for implementation, would be to partner with WIND.
Allen, |
