The ultimate effect of over planning is that the team and the management spend more time in planning and re-planning rather than reaching the goal of delivering the project, notes
Dr R Srinivasan

"In the case of Detailitis plan, over-planning still continues in the various stages of the project development, with a wrong belief that it will lead to effective completion of the project "

Having seen the antipatterns in software development and software architecture, the next few antipatterns to be studied are the ones that occur in software project management. It is one of the most difficult activities in any engineering discipline and it is more challenging to manage a software development project.

A software development to be successful calls for a balance among various considerations. Typical ones are formalised construction, using well defined and reusable components adhering to the processes to have good control on the development, adapting new and current technologies, pattern-oriented developments to solve known problems and above all leading the project team in an effective manner so that human aspects are also taken into consideration. A project manager in software development should take care of the software process management, resource management that includes not only hardware and software but also the members in the team, and also establishing and maintaining relationship with customers and partners. Each one of us who has been involved with the software industry goes through different kinds of stresses at different times.

In this regard it has always been emphasised that a good manager should be able to moderate the stress levels of the organisation, particularly when they have to face serious consequences of the project not meeting the planned goals. We will be discussing the management antipatterns to illustrate how a software development project can be affected by issues related to members in the team, non-conformity to processes, inadequacies in infrastructure and also lack of experience in handling the customer and external people.

The first antipattern that we will be seeing in this direction is 'Death by Planning'. In any project development, planning is a very vital aspect. Of course, planning has to take place only after the given problem is understood and thoroughly analysed. It is absolutely essential to take good care of the planning process; but in some cases the team and the manager take it too seriously that it may result in over-planning. William Brown illustrates two types of over planning, called as, 'Glass Case Plan' and 'Detailitis Plan'. The Glass Case Plan is like something which is assumed to be the best and put in a glass for exhibition.

In some organisations this happens, thinking that the plan which was decided before starting the project is the best and so it does not even go through the process of reviewing. The reasons for such a decision may be many like inability and inadequate experience to plan at a pragmatic level or overconfidence on the plan, saying that there is a plan of the project, which will save it from the customer, or too much of commitment as long as the project is funded etc. But, it will result in unforeseen consequences resulting in continued increase in cost during development of the project, frustration among the team members, pressure from the customer, crisis management of the project and as an extreme case cancellation of the project.

Glass Case Plan is a subset of Detailitis plan, which means that the act of over-planning will cease as soon as the project is started. In the case of Detailitis plan, over-planning still continues during several stages of the project development with a wrong belief that it will result in an effective completion of the project and delivery to the customer. The ultimate effect is that the team and the management spend more time in planning and re-planning, rather than reaching the goal of delivering the project.

Both these cases of over planning are clear examples of the violation of successful software development practices and lack of project management processes. The refactored solution for either of these is to revamp and go to following the software development processes like: good planning and tracking using a Gantt chart, identifying the deliverables and milestones at the planning stage. The plan must be base lined and certified by the quality department which will enable an effective tracking to identify slippages in efforts and schedule. The most important part in the development cycle is to have reviews at each stage to get the approval of not only the plan but also design approvals at the conceptual as well as specification levels, test plans and maintenance plans.

( To be continued)

(The author is Chief Technology Officer, Internet Component Management Group, Bangalore and can be contacted at: r.srinivasan@iCMGworld.com)