Information Technology Reference
In-Depth Information
capacity (time, money, resources) for creating a functional and manageable
software system. One could:
1.
Focus only on certain areas of user needs, either ignoring other
areas or else leaving them un-automated.
2.
Develop policies and rules that reduce the scope of the application
to be delivered.
3.
Phase the delivery.
For the second approach, one could increase the information processing
capacity to match the outside variety. For example,
Employ more or better qualified resources.
Train the resources specifically for the project.
Use more effective technologies.
Contract with other firms to integrate their product offerings,
instead of building them oneself.
Get team members to deal with more problems themselves, through
self-management and employee empowerment. This reduces the
variety required of managers on the project to manage individual
team members.
Choices and Constraints in Software
If one asks whether destroying variety is one of the objectives of regulation,
the answer is yes. Regulation requires constraints on a system. Although
the term “constraints” sounds negative or restrictive, constraints are actually
necessary. Most “laws” of science are constraints on the way nature
behaves. Constraints help us in at least two ways: (1) as an organizing
influence and (2) as an enabler of predictions. They are an organizing
influence through the reduction of variety. Ashby gives the example of a
chair whose four legs, when unconnected and separate, have six degrees
of freedom, each for a total of 24 degrees of freedom. When they are
joined, however, they have only six degrees of freedom, which are the
six degrees of freedom of the chair. The reduction of 24 to 6 reflects the
creation of a chair — an object that is its own entity, which is a whole
and is more than the sum of its parts — a basic premise of systems thinking.
Constraints serve two purposes:
1.
They enable us to make predictions.
2.
Without constraints, there would be chaos.
