DOS Ain't Dead.

Software developers have declared the death of PC-DOS. Most of the development tool vendors, who showed their wares at the Spring Software Development 1994 Conference in San Jose, California did not have DOS targeting capabilities.

Many of these development companies, who provide software development tools, ignored the fact that over 10% of the computers in business cannot run Windows or OS/2. Even if these computers could run Windows or OS/2, they cannot run it very well. Either they have an 80826 processor or less, lack sufficient ram, or lack the disk storage required by these environments.

Last August a Symantec representative stated that bedrock, a multi-platform development foundation library, could not include PC-DOS support. He further explained that supporting PC-DOS in Bedrock would be impossible. An unnamed source from Novell, put forth a similar view at this year's conference.

BULL! Innovative Data Concepts, XVT, and Zinc have developed these types of libraries. They all support DOS, Windows, and OS/2 using the same code base when you write your programs to their libraries. Watcom, which has one of the best compilers available, have compilers and tools that support all of the aforementioned platforms. Knowledge Dynamics writes and sells a program installer, Install Pro, which Works with DOS, Windows, and OS/2. It is true that under DOS you may have to make some compromises, but that does not mean that you completely drop support for that platform.

What these companies forget, is that there are still several million machines in service that can only run DOS. When I develop my programs, I want them to run on as many machines as possible so I have more potential users. When you write your programs, don't you want to sell to all platforms? In most companies, there are many machines that cannot run Windows well, but run your programs perfectly, Why spend the money and time to upgrade them?
According to the Software Publishers Association (SPA), there have been 6.81 billion dollars in North American software sales for 1993. The SPA figures also state that 1.93 billion dollars, 29%, of these sales was PC-DOS software. Last year $3.43 billion, or 51% of sales, was spent on Windows software. Scott Benson, Vice President of Marketing for Microsystems Software predicted that, that 15% of new software sales this year will be for be for PC-DOS systems. A vice president of a local software company, who requested anonymity, stated that they expect between five and ten million dollars in additional sales this year due to the release of a Macintosh version of software.
This additional money is not from the sales of the Macintosh version, but is generated by Macintosh availability. Prior to the release of the Macintosh version, they had PC-DOS and Windows versions. There are several companies refused to buy the DOS and Windows versions since some of the companies employees work on Macintosh. One expected sale is from a company where everyone runs on Windows, but the president has a Macintosh on his desk. This situation applies to PC-DOS as well as the Macintosh. There are still machines in companies that use machines that cannot run Windows well.

Using tools that support multiple platforms and following some programming guidelines one can develop software that runs well on multiple platforms. If more people can use a program, a vendor can sell more. As a software developer, I want a share of this $6.8 billion dollar market, not a share of a portion if this market. Would you want to lose millions of dollars of potential sales?

In your own organization, do you want to upgrade all the machines to run one program? If you use a program that is available under DOS, Windows, and OS/2, you can avoid, or at least defer, upgrading many of the users. Costs of upgrading are enormous, not only in hardware and software costs, but in the user retraining costs. Multiple platform software allows one can migrate people when there is sufficient benefit justifying the cost. This allows a better justification of the upgrade costs without the risk of less risk.

I would not want to ask my CFO for the funds to upgrade all the machines in the company, especially if most employees can use PC-DOS versions instead of the Windows version. I would not want to explain to the CEO of my company that a program that works on his secretary's computers would not work on his computer. To avoid this situation, develop and buy products that will work on all the platforms that your organization uses.


Go Home


Copyright 1998 William Silverstein. All rights reserved.
Revised: April 29, 2004.