---------------------------------------------

NOTE: This article is an archived copy for portfolio purposes only, and may refer to obsolete products or technologies. Old articles are not maintained for continued relevance and accuracy.
October 25, 1996

Network Computing Will Lead to Free Word Processors

Technologies typically follow a four-phase market evolution cycle of hardware, software, service, and commodity. For example, the network protocols used with the original Internet were embedded in hardware-based communication controllers. Then IP was developed, and the protocols became software. As the number and variety of IP implementations flourished, the focus moved away from software-vs-hardware towards service-related issues, like who had the best performance, features, and customer support. Now we see that IP and the related services have been moved directly into the operating system on almost all platforms, resulting in a truly commodity technology that is essentially free.

Word processors have also followed this same evolutionary path. The original word processors were typewriters and Wang systems, both of which were definitely hardware-oriented. As the PC market began to bloom, software-based word processors became the norm. As the number and variety of implementations flourished, the focus has moved towards the service aspect, with features, pricing and distribution determining the winners. The next step in the life cycle will be where word processors are free commodities.

Furthermore, the transition between these four phases is almost always initiated by transitions in the technology's arena. In effect, as the networking arena shifted from hardware to system-level software, TCP/IP was also moved from dedicated hardware to the OS. And as the state of keyboard-oriented technologies moved from typewriters to electronics, word processors were also moved. In essence, the shift in the paradigm also triggered a shift in the related technologies.

This is why we'll soon have free word processors. As the computing paradigm shifts towards one of network-centric services, all of the related technologies - including word processors - will also get shifted at the same time. Since the next logical step for word processing technology is to the commodity phase, this means that for all practical purposes word processors will become "free."

For evidence, I point to Corel's new WordPerfect for Java. Using less than 700k, it has an almost exact feature match for its Windows counterpart. Since it's a server-based, client-executed application, it carries tremendous benefits for network administrators. You install it (once) onto your network server, set the appropriate user permissions, and then clients execute it on their local PCs. When a new version is released, you just update the server's store, and all clients are immediately on the next rev. No more desktop management nightmares.

The problem - and also one of Java's selling points - is that Java applications are not pre-compiled. While this allows the application to run on any Java-enabled system, it also means that anybody who can write an HTTP client can have full, unhindered access to your Java application's raw source code. So, if Corel were to release WordPerfect for Java, anybody that wanted to could effectively steal the code and sell their own Java-based word processing application.

Obviously, this is a big problem, and until it's solved we're not likely to see many of these applications on a commercial basis. Indeed, the only applications that we are likely to see distributed in this manner will be those that are tied to some sort of back-end system. For example, Oracle could (and does) publish Java-based clients that relied on Oracle back-end databases. Similarly, Open Connect could (and does) publish Java-based clients that only speak to Open Connect's SNA gateways. If the client source code was copied, then so what? The revenue from back-end systems would be the cash cow in this model anyway.

However, Corel doesn't sell a back-end system for WordPerfect, so it makes it very difficult for them to release these products without fearing the potential cuts made to their existing revenue stream, meaning they aren't likely to start giving away their products for free anyway. This is unfortunate, because the transition to the new phase in this technology will require it to be a commodity.

In fact, the only companies that can give away applications such as these are the companies that have a motivation for giving these applications away, and do not depend on the revenue stream generated by these applications, or who are willing to do without one if it already exists.

Who might do this, then? The immediate thought that jumps to mind is that Netscape might. If, for example, Netscape were to bundle Java-based versions of these applications with their SuiteSpot servers, they could accomplish several things. First of all, they would be able to further legitimize the position of Navigator as a self-contained platform. Second, they would be able to link the applications to the use of SuiteSpot directly, thereby adding more value to their server lines. Finally, since they have no current revenue stream to depend on, they could implement this without any loss of revenue, while simultaneously crippling Microsoft's desktop application revenue stream. With the recent announcement of AppFoundry, we've seen that Netscape is willing to bundle workgroup applications with their servers, so it seems reasonable that desktop applications might not be that big of a deal, either.

Another possible candidate is Novell. I know, I know, they just got rid of WordPerfect, but that was probably a premature mistake in light of WordPerfect's new-found Java capabilities. If Novell could bundle the WordPerfect for Java suite with every copy of IntranetWare, then they could add value outside of their file-and-print stronghold, while simultaneously achieving Ray Noorda's dream of "beating" Microsoft. After all, the users of these products will still need to store the documents somewhere, and it might as well be on a NetWare server, or in an NDS tree.

Yet another possible candidate is SunSoft, who could stand to add some of this functionality to their solutions. These folks have also shown that they're not afraid to fight it out either, and are willing to give stuff away simply to secure a strong position.

Note that all of this is pure conjecture, and that there is absolutely no evidence that any of this will happen. And whether or not Netscape, Novell or SunSoft brings any of this to pass or not is irrelevant in the bigger picture; the new focus on network-centric computing will eventually force the commoditization of the application technologies, regardless of who the players are behind the effort.

-- 30 --
Copyright © 2010-2011 Eric A. Hall.
Portions copyright © 1996 CMP Media, Inc. Used with permission.
---------------------------------------------