.

Now we can define the third outcome of a software project. To make a decision one need the information and the experience. Let's say that third result of a software project is a distributed knowledge base about the software product and the process of development such kind of software. If we consider this project's outcome together with the project's staff, we can speak about a kind of distributed expert system. In most cases the knowledge base cannot be separated from this expert system because large part of it is hidden in the heads of people. Of course this part is subject to lose by time and staff changes.

The quality of this knowledge base and this expert system defines the quality of the current software development process and of the future projects. The second problem must be considered is the cost of them. DoD organisations make complex and expensive things to raise the quality of this knowledge base. However the projects in such organisations fall too. The firms doing ordinary software business are unable to carry analogous cost.

There exist a problem how to get this knowledge base maximal cheap without loss of quality. The second problem is how to make this expert system maximal effective and stabile. In other words there exist a problem how to force good decisions about software and how to prevent bad in the future.

The solution of this problem is not the primary subject of this paper. Let's return to the main path.⁷ Now we are ready to speak about Software Dragons.