Monthly Archives: November 2008

Madmen In The Halls

I read Tom De Marco’s book Slack a few months ago, the first chapter titled “Madmen in the Halls” starts by suggesting that:

“The legacy of the nineties has been a dangerous corporate delusion: the idea that organizations are effective ony to the extent that all their workers are totally and eternally busy. Anyone who’s not overworked (sweating, staying late, racing from one task to the next, working Saturdays, unable to squeeze time for even the briefest meeting till two weeks after next) is looked on with suspicion. People with a little idle time on their hands may not even be safe. As [one of the authors friends] at Digital Equipment Corporation told [him] during the company’s darkest days, “There are madmen in the halls looking for someone to ax.” Of course, the ones they were looking to ax were the folks who weren’t all that busy.”

It’s sad to be able to draw parallels with some of the rash actions that were being taken then and the drastic ones people are being forced to take now.

Inevitably, as people look to make their respective companys’ more efficient and trim their bottom lines, Development teams are called upon to get the latest cost / time saving feature delivered as soon as possible.

It’s at times like this, when we’re being asked to go as fast as we can it becomes increasingly difficult to sell the fact that we need to deliver a quality product rather than one in which we’ve cut corners, as Jason suggests in his post, perhaps quality will suffer as part of the credit crunch. But now is precisely the time that we should be pushing harder to see the quality of our product improved.

I’m a believer in the Lean principle that you should decide as late as possible and that development teams should seek to defer decisions until the last responsible moment. Should we defer that decision to include unit tests, to tackle that particularly naughty piece of the code that is always causing us problems or to automate our build and deployment process. Definitely not.

Choosing not to pursue quality from the outset is almost like behaving like a student, leave doing it until the last minute and you will either hand in something that’s rushed, doesn’t really meet the expectations and will therefore need to be reworked or having to ask for an extension to the deadline.