A crack in the PC paradigm? Microsoft panicked? The Wall Street Journal reported today that "Microsoft Sets New Direction On Terminals". What does this mean?
Microsoft has been playing defensive ball for quite awhile now, first waging war on startup Netscape, then challenging AOL. Of course these challenges were just fodder, for the larger issue: the NC challenge by Oracle/Sun/IBM/Novell, with the stelth winner being WIND. Even this issue is minor compared to what we know is the mother of all challenges, the ubiquitous computing/communications attack on the PC paradigm itself - but no one yet recognizes this.
Intel is absolutely ambivalent about the NC: the numbers are attractive (billions), but the threat of a loss of monopoly power in the choice of processor is abhorrent. Microsoft has never been ambivalent; it despises any mention of the NC. Microsoft is on record, over and over again, for ridiculing the notion of the NC. The PC gives the power of the computer to the end-user. Who would ever voluntarily sacrifice it to a centralized server?
NC advocates respond by pointing out that no one is suggesting servers take over local processing; the NC can duplicate anything that a PC can do, but at a much reduced price, and at a vastly reduced amount of administration.
Intel, being paranoid, decided to develop the world's best worthless NC. "We don't think the NC is viable, but Intel will develop the world's best one, just in case we are wrong."
Microsoft intervened and convinced Intel to join forces and counter the specious NC with the NetPC, which is a reduced PC in some unspecified fashion that is easy to administer in some unspecified way.
Apparently corporations have made it clear to WinTel that enough is enough, the PC is too difficult administratively - something serious needs to be done. Whether corporations talked directly to Microsoft or whether they voted with their noted interest in the NC, the result is the same.
Well, today we got the answer from Microsoft: Windows Terminal.
What is a Windows Terminal? Apparently, unbelievably, a Windows Terminal is exactly the same as an X-Terminal in Windows space. A Windows Terminal is simply a dumb display that depends on the server for all processing - exactly like an X-Terminal. In case you don't know, X-Terminals have been around for years, and are neat extensions of Unix boxes. Take one Unix workstation, add a few X-Terminals and you get multiple workstations for the cost of one plus relatively low-priced X-Terminals. The problem with X-Terminals is that they don't scale up very well. They are neat, inexpensive ways to double or triple a workstation, but they fail miserably as a means departmental computing. Normal servers cannot handle the memory and processing requirements of hundreds or thousands of X-Terminals. Similarly, the Windows Terminal is not an answer to the NC. It is particularly not an answer when zillions of consumers need to be connected to servers. (By the way, just in case I am wrong, take pleasure in the knowledge that VxWorks is used by HP to power its competitive X-Terminal.)
X-Terminals are demanding of high network bandwidth, since bit maps must moved from the server to the terminal for display - constantly. As a consumer tool, therefore, Windows Terminals are a very bad idea - even with infinite server resources.
This cannot be happening. Life isn't any fun if it is this easy. This idea is so stupid that it represents absolutely no challenge whatsoever to NC, while completely validating the NC as bona fide alternative to the traditional PC paradigm.
The embarrassing thing on this thread is that we have been dancing around the threat of Microsoft to embedded systems RTOS vendors. Not only is Microsoft not a threat to embedded systems vendors, it is now questionable how much longer Microsoft will even dominate the PC paradigm. I think I would put my money on Larry Ellison, where the bet is that Oracle becomes number one ISD within, say, five years.
What does this mean for WIND? Two things: (1) WIND will not face a serious threat from Microsoft in the embedded systems space - because Microsoft is very, very busy trying to refigure the PC paradigm, and (2) WIND will be an enormous winner underpinning Oracle's NC with VxWorks.
Please note that Oracle owns the NC. As we suggested earlier on this thread, the key to the NC is the server, and the absolute, dominate NC server will be Oracle. This means that the Oracle NC really will sell in significant numbers, bringing riches to WIND stockholders.
Unlike all the WIND analysts, I do include I2O projections in my estimates, but like them I still do not include NC projections. Revenues for WIND from the NC should start in CY 1998, as Oracle implements the 2nd generation NC OS, and should quickly grow significantly.
Allen |
