Thursday, June 21, 2007

An Experiment

My head is still swimming from all the meetings which have taken place this week. It's all part of the "Getting in Over My Head" plan for the summer, where I play the part of "Befuddled and Confused" in this bizarre Morality Play.

For the past three years, this program has managed to survive despite the fact that we've been ignoring the proscribed software development processes, instead doing things on a panic-of-the-week basis.

In the normal software development process, we take a set of requirements which have been negotiated with the customer and design the software on paper, get approval from the customer on the design, transfer the paper design to an actual design, match the design against the requirements by running tests to verify that the requirements have been met, then release the software - and all the documentation that goes along with it (design documents, requirements documents, test documents, test reports).

In our "modified" process, the managers negotiate a delivery date with the customer, along with an 'incentive' payment that guarantees (by appealing to executive greed) that we'll make that date; then they turn around and demand that the engineers work 50-hour weeks (minimum) in order to meet that date; then the engineers, in order to meet the delivery date, ignore all the paperwork; and after much pain and torture, they make the delivery date, but have no idea what it is that they have delivered.

So why are they so surprised when it doesn't work as advertised?

**

The "mandate" for my new position is to ensure that the aforementioned 'correct' process is followed. This is what is known as "bucking the trend". I'm not sure if success is even possible at this late stage of the game, but it'll be interesting to find out.

After all, the worst they can do is to fire me.

No comments: