March 14, 1997 6:00 PM ET
Requiem for OpenDoc
By Norvin Leach

  At the June 1995 PC Expo in New York, Apple Computer Inc.'s Michael Spindler, Novell Inc.'s Robert Frankenberg and IBM's John M. Thompson stood together on a stage asserting their commitment to OpenDoc and proclaiming it was key to the future of their companies.

The presentation, full of warmth and promise, was supposed to feature a demonstration of interoperable OpenDoc parts across Windows, OS/2 and the Macintosh.

Unfortunately, the setup wasn't working correctly.

The demo's intractability, combined with the boiling heat of a New York summer and the accumulated pressure of preparing for the show, was too much for one of the demonstrators. She fainted on stage. The rest of the crew used that as a convenient excuse, claiming that the demo didn't work because she'd pulled out a power cord.

But the crash was a bad omen. OpenDoc had peaked and was on its way down, culminating in today's decision by Apple to discontinue the product past Version 8.0 of the Macintosh operating system.

It didn't seem that way on the surface. The OpenDoc consortium appeared to be just as solid as it had been back in the fall of 1993, when IBM, Apple and Novell first announced the compound-document architecture.

Granted, a few of the key partners had failed to follow through with their announced support-Oracle Corp. was cautious and quiet, Lotus Development Corp. supported only pieces of the technology, and Borland International Inc. was nowhere to be seen.

Still, a series of seminars and road shows during 1994 had whipped up enthusiasm among hundreds of small developers-the same customer base that had made Microsoft's VBX controls so popular.

OpenDoc was a neat, elegant architecture. In head-to-head debates, the audiences clearly favored the non-Microsoft solution. And the demonstrations grew more and more sophisticated and flashy each month.

But underneath, the development effort was resting on shaky ground. Novell, overburdened with technology from its corporate acquisitions, refused to commit significant resources to building its part of the architecture.

And the Novell part, the link to Microsoft's OLE, was vital for OpenDoc's future. Without OLE interoperability, OpenDoc would remain a marginal technology.

Meanwhile, OS/2, which was to spearhead the growth of OpenDoc, was not succeeding as well as IBM had hoped. A brief, flashy attempt at pushing OS/2 into the consumer market failed, and IBM fell back on selling the product to its more stable corporate base.

By the time of the PC Expo demonstration, a year and a half after the unveiling, the code wasn't finished, true Windows interoperability was many months away and, perhaps worst of all, there were no simple, readily available tools to build OpenDoc applets or assemble them into compound documents. By comparison, the success of the VBX can be attributed to the simplicity and availability of Visual Basic.

In short, the development of OpenDoc had not kept up with its marketing.

A few months after PC Expo, Novell dropped its OpenDoc efforts. IBM took over the development, only to discover that Novell had barely started working on the code. In the fall of 1995, after reviewing what Novell had turned over, IBM announced that it would take more than a year to release a final version.

At that point, developers began to look elsewhere. Many of the hard-core proponents of OpenDoc were small companies that could no longer afford to wait for the technology. And larger corporate customers, such as U.S. Trust, gave up in frustration when IBM failed to deliver technology on a consistent schedule.

In 1996, the development world shifted its attention to Java and the Web. OpenDoc's concept of compound documents had not taken the Web into consideration, and it took some scrambling on the part of the marketing teams to reposition the technology.

Although Apple showed OpenDoc's flexibility by building the CyberDog browser, that browser never caught on. It looked flashy, but Netscape Communications Corp. was too well entrenched.

By the fall of 1996, the end was obvious. In October, John M. Thompson repudiated the technology, saying that OpenDoc had not ridden the Internet wave effectively. He suggested that the technology should be cannibalized and included in the JavaBeans object framework.

Copyright(c) 1997 Ziff-Davis Publishing Company. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of Ziff-Davis Publishing Company is prohibited. PC Week and the PC Week logo are trademarks of Ziff-Davis Publishing Company. PC Week Online and the PC Week Online logo are trademarks of Ziff-Davis Publishing Company.

Send mail to PC Week