Why MV-DBMS Is So Affordable

Several studies show that the annual total cost of corporate computers averages about $6,000 per user. Comparably, most MV-DBMS systems run only a small fraction of this cost. Why is there such a difference?

When you analyze the costs of computer ownership, one fact stands out. Equipment and basic software is but a small component of the total cost, representing but 10% to 20% of the cost. Not much more goes into system maintenance. The bulk goes into either general administrative costs or software development.

Historically, MV-DBMS has had two core properties which cut these cost components dramatically: simplified administration, and grossly reduced programming costs.

The first was simple administration, due to a simple-to-use operating system. This operating environment, tailored to the needs of multi-user businesses, was called "the Pick Operating System." Historically, this operating system was so simple that its core functions could be learned in a few days of training and practice, and practically required only 10 to 15 minutes a day of operator use to assure reliable service. This contrasts strongly with "open systems" such as UNIX and Windows NT which can require 6 months or more of training before they are mastered, and where ongoing work was likely to require 10 to 20 times the skill of the Pick Operating System.

The Operating System portion of the MV-DBMS model has been largely abandoned because while it creates a simple to use environment, it also places major limits on the adaptability of the MV-DBMS database model to new and compelling computerization techniques. Thus, more than 90% of new MV-DBMS systems run on open multi-user environments such as Windows NT, UNIX, and Linux.

Never the less, there remains one traditional model product: Mentor-Pro, supplied by General Automation. In addition, Pick Systems preserves the simplicity of "Native" in its D3/ProPlus product, which maintains the appearance and use of the native concept, though using a Linux operating system to drive the functionality. In striving to grossly simplify use, they sacrifice a lot of the possibilities inherent in Linux, but some people are better off without a lot of bells and whistles.

Our company follows a different approach. We set as a goal expansion of the range of functionalities of small business computers, while reducing the associated administrative burden. We have two basic ways we obtain this. The first is EZ-Open and EZ-Lite, which are administrative tools designed to preserve the functionality and properties of open systems while reducing the time burdens associated with greater functionality. Our goal is to obtain either a 10 fold increase in effective functionality without time increase, or a corresponding 10 fold reduction in the time cost of individual functions.

The second of our administrative approaches is Easy-TimeShare, which gets you out of the administration business entirely, and obtains very high levels of administrative efficiency through a combination of scale and automation.

The second is programming efficiency. The MV-DBMS database model is designed to cut programming costs dramatically by providing tools which allow programmers to rapidly collect, transform, and manipulate data. Historically, a 5 to 10 man MV-DBMS programming shop has been able to obtain the same work-output as a 100 (or more) person COBOL shop.

Historically, programming cost savings have not manifested themselves in radical price reductions, but rather in the form of more thoroughly thought out and specialized products. For instance, while a non-MV-DBMS software house might produce a package for physicians, or for booksellers, the MV-DBMS house tended to produce applications for neurologists or for technical booksellers, each of which is but a small subset of the larger universe. Thus, productivity in the MV-DBMS world manifests itself in a larger number of applications tailored to very specific industry needs.

Never the less, the MV-DBMS paradigm has lead, through a combination of administrative and software savings, to environments which often cost their users only $1,000 to $3,000 per user year, rather than the industry average of $6,000.