A Moral Dilemma

The process of task listing also points out an interesting effect of the spectacular growth in technology and technology options.

In the 1960s and 1970s, the choice of development options (i.e., what technology was going to be used to implement the system) was extremely limited. Usually the choice was binary: COBOL ^[5] on the mainframe or manual. As shown in Figure 13.5, in the 1990s, development options are virtually unlimited.

^[5] For purists , Assembler and FORTRAN were other language options.

Figure 13.5. Task listing dilemma

In Figure 13.5, the team is concurrently analyzing the requirements for the system while they are evaluating alternative package solutions. The system requirements are then compared with the initial evaluation of package system options.

Matching the requirements against the package software functionality provides at least three development options. Scenario 1 is that there are no packages that fit the requirements and the team will develop a tailored in-house solution (there are a number of suboptions within this option). Scenario 2 is that the package "fits" the requirements perfectly and the development is a package installation. Scenario 3 is to modify the package to fit the requirements.

As you can see, each scenario has different tasks involved; hence the dilemma.

Until the team members have both understood the requirements and evaluated alternative packages, they cannot professionally eliminate any of the scenarios. In addition, until they have eliminated all unsuitable scenarios, they cannot estimate how long the project will take! This dilemma is becoming even more complex as development scenarios proliferate through new technology and languages.

There are a number of additional implications of this dilemma. First, the process of task listing does not make much sense beyond the "if" test.

Of course, you should recognize this as an example of the real-time or scenario planning approach we introduced in Chapter 3. The solution to the dilemma is to estimate the worst case scenario (usually tailored development) using the techniques we cover in the next chapter with a more accurate estimate to the "if" test. Then, once you have decided which option is the most viable , you can provide a second estimate for the remainder of the project. This is often called two-stage quotation.

Where Is the Moral Dilemma?

The moral dilemma is clear. The easiest route to take in solving the existence of multiple development options is to eliminate them early, choose the option that is of most interest to you, or hide them from your project sponsor and stakeholders. You choose whether you go with the good side or the dark side.