The following text is copyright 1995 by Network World, permission is hearby given for reproduction, as long as attribution is given and this notice is included.

Contraindications for Fragmentation

By: Scott Bradner

From the unrealized fears department. For a while it looked like just about every major company thought they could do the Internet better than the Internet could do the Internet. Ziff Davis, Apple and others announced grandiose plans for the creation of private networks with national scope but only tenuous descriptions of how they were going to interact with the existing Internet.

I wrote about the Ziff Davis announcement a while back (NW 3/7/94 pp 13) and it turns out that the problem was in the telling not in the planning. I have now been told that they had plans from the start to be well connected to the rest of the Internet. I’ve not been told what those plans are (or were, that part of the Ziff Davis holdings was purchased by AT&T last year) but I’ve been assured that it was only a glitch in the PR department that kept the truth from the world.

For quite a while it has been known that Microsoft was going to do something with networks and on-line services in conjunction with, or at least in temporal proximity to, Windows '95. There was a great deal of speculation over what form these efforts would take. In particular, there was a real worry that Microsoft would decide that it was big enough to build its own international network which would compete with the existing Internet.

This was not a totally irrational fear. Many other big companies have felt that way in the past, with the best example being IBM and SEARS with Prodigy. With projected sales of Windows '95 in the tens of millions of copies and with the resources it has at hand (now that it is not buying the Catholic Church), Microsoft would quickly have the scale and reach of the existing Internet.

But last week’s announcements by Microsoft and Uunet Technologies painted an entirely different picture. There will be "The Microsoft Network" but it will use the expanding Internet as its structure. It will not be a collection of entirely proprietary applications running over a private network but instead it will be Mosaic running over an integrated Internet. I would not be surprised if there were some proprietary additional applications but the basic thrust seems to be that Microsoft does not need to invent everything anew but can make use of existing technology. Although it was not clear from the announcements, it appears that Microsoft will deploy, or help get deployed, an array of new services on the Internet and that it is this collection of services that will be "The Microsoft Network".

The attitude of the other big players seems to be changing. CompuServe and Prodigy started out thinking that they could go it alone without dealing with the Internet. After much prodding, and in some cases private gateways, they finally accepted the idea of exchanging email with the Internet but did not think much of the idea of actually supporting interactive applications. These and most of the other services of this type are now rushing to figure out how to let their users "surf the Internet." I expect that they are not doing this because of some inner light but because they fear losing customers to competitors who do a better job of putting their users on the Internet itself.

MCI is pushing its InternetMCI service with millions of dollars of advertising, advertising which is a lot clearer and more to the point than what they started with. Speaking of advertising, I don’t know about you but selecting the name "Warp" for an operating systems does not convey an image of maturity and stability to me. It sounds more like an over sugared breakfast cereal. Well, IBM has now gone from pop art to classical (from the ridiculous to the not quite sublime) with its latest ads of old French men and Italian nuns talking about the marvels of IBM technology including the lure of the Internet.

The common theme for all of these is that the Internet is better, or at least more popular, than what a company can do by itself, thus the fragmentation of future data services which many feared is not in the current cards.

Disclaimer: Harvard, like all users of the Internet, has an interest in seeing that future data services be universal but these words express only my own opinions.