Third, in growth markets such as the Web, Microsoft cannot afford to be as leisurely as it was in the 1980s when it faced slow-moving competitors with flawed business models such as IBM, Lotus, Novell, Apple, and Sun.  Now it is chasing Google and Yahoo, who are growing much faster than Microsoft, and whose combined revenues in 2005 will exceed $10 billion.

In my article on Google, I therefore may have given Microsoft too much credit.  I still think that Google does indeed have the vulnerabilities that I discussed, and that if Microsoft attacked forcefully, Google would be in serious trouble.  I also still believe that there will be increasing demand for common, interoperable search functions across many platforms, and that widely available software with open APIs would be valuable for both the search industry and for users. 

But if Microsoft is too paralyzed and complacent to act, Google will succeed whether or not it provides an open platform.  Unlike the Linux case, in the search arena there do not presently seem to be other powerful actors likely to step in to discipline the dominant player. 

As a result, if Google chooses not to release an API for its search engine, the search industry may not evolve the standards and open interfaces that would provide interoperability across PCs, corporate servers, the Web, proprietary content providers, and portable consumer devices.

Alternatively, Google might provide such interoperability itself, but only within its own closed universe.  Thus either short term profitability or internal corporate politics could lead Google to decide against the deployment of truly open search architectures.

This would be a shame, but it could happen, and things like it have happened before.  Much of the success of Linux, for example, derives from the fact that in the 1980s, the UNIX operating system split into multiple, incompatible dialects, each one controlled by a different hardware vendor (Sun Microsystems, IBM, DEC, Silicon Graphics, Hewlett Packard). 

This brought high profits for over a decade, followed by disaster when an open, industry-wide alternative finally appeared. Most industry analysts now believe that Sun is in irreversible decline at the hands of Linux. 

In the search industry, too, market forces would probably eventually bring regime change, but it could take decades.  Only time will tell what Google will do, and whether users will be well served; my own hope is that Google will conclude that enlightened self-interest leads it in the direction of an open architecture.