The rewards of disciplined software development February 22, 2008
Posted by Wille in Software Development.add a comment
On my current project, we started out being fairly disciplined in using TDD and applying refactoring whenever it was suitable, after a couple of months however, we started slipping a bit, primarily due to massive external pressure to throw out “quick fixes” for problems in applications that were not yet even finished.
However, in the last month and a half, we have gotten back on track, caught up with our test coverage for code that was hurried out, refactored any code that was duplicated or could be reused, and the results are if not beautiful, then at least pretty darned close to it.
After having finished three out of four applications, the productivity we are now seeing is mindblowing compared to the beginning of the project, and that is simply because we have so much common code that we have been disciplined in refactoring for our purposes so that doing just about anything is now a matter of just fitting a few pieces of code together.
Refactoring that would not have been politically feasible had we tried to sell it explicitly, but that we managed to push through anyway on our own accord are now reaping huge rewards.
The fact that our test coverage is also good (but not perfect) means that we also have a high level of trust in our code, if we build the code base and no tests fail, we are fairly certain everything works and no old bugs have crept back. The time spent in writing the tests (which is questionable if it really is a drag on time spent) has been more than made up for in removing the need for doing infamous “sanity checking” of our application whenever we do a release.
As a team, I would attribute our relative success to three main factors:
- Instead of doing the “quick fixes” and “hacks” at the beginning, we took the bullet of doing things properly at the outset and fixing things in hindsight even if the rewards were delayed, and we are now seeing the rewards.
- Communication within the team has been excellent, we sit together, work together and everyone is eager to learn.
- Everyone has professional pride in what they are doing: everyone in the team actually wants to write good code and deliver successfully. There is both professional pride in the team, as well as loyalty to each other in the team.
Having a good, disciplined team of highly motivated individuals will be better every time than an army of demotivated automatons, even if the highly motivated individuals cost ten times more a head.
Why I’m sick of Fast Company February 11, 2008
Posted by Wille in Media.add a comment
Fast Company is one of the magazines I’ve enjoyed reading for over 8 years - It was originally a magazine for entrepreneurs and business people who were passionate about what they did, and wanted to keep a finger on the pulse of what other similar people were doing, and what worked.
Not so anymore, Fast Company seems to have become a source of politically correct drivel, simply parroting whatever fad or hysteria is going on at the moment. Not an issue seems to go by without the magazine having a feature on “environmentally friendly companies” or “social entrepreneurship”. There is nothing wrong in either of those concepts, but when it is veiled in a slight sense of shame over the fact that Fast Company actually covers a phenomena that is capitalist at it’s core, and has profit as a goal, it gets a little too much: they simply seem to be ashamed over having to cover capitalist pigs who make a profit from creating new products, services and jobs that improve peoples lives and create prosperity.
If Fast Company is ashamed of covering those who create prosperity and need to shroud their coverage in various politically correct veils, maybe they should either change their name and profile entirely, or cease existing?
Political correctness, fad following and shaming of economic creation is tiresome wherever it rears its ugly face, it is doubly so when it comes from a supposed business magazine.
Return of an old friend February 5, 2008
Posted by Wille in Entrepreneurship, Media.add a comment
For those of us who tasted our first “battle” in IT and entrepreneurship during the dizzy heights of the dotcom-bubble, one of the defining publications of the time was The Industry Standard, a magazine that rose to prominense during the bubble and went bust with the demise of that era.
Well, mourn no more, The Industry Standard is now back, with the appropriate Web 2.0 bells and whistles to go. It remains to be seen whether the Standard can yet again climb to prominence, or if it will just be one last horraah for the nostalgics.
The difference between predicting and extrapolating February 1, 2008
Posted by Wille in Human Behaviour, Investing & Economics.add a comment
People love predicting and predictions, yet new turn of events always seem to catch us off guard. Just compare last years predictions with todays realities: UK property was predicted to continue growing at mindboggling rates, today it’s dropping for the fourth month in a row. Growth in 2008 was expected to beat growth in 2007, now we are worrying about a potential recession.
Why are we so sure about our predictions, yet manage to get them so wrong whenever events turn in another direction?
The hint is in the words of the last sentence: direction.
Most predictions aren’t actually predictions, they are extrapolations - They extrapolate the future based on the current direction and trajectory, nothing more. They may take known information into consideration, but they rarely affect the extrapolation in a material way.
The fact is unexpected events and turns of direction are just that, unexpected, people rarely, if ever see them coming. But these events are the events that often dramatically change the current direction and shape the future: Who saw 9/11 coming? Almost no one. Has it shaped history as we know it for the last 6-7 years? Definitely.
The moral is that predicting the future is a fools errand, you’re always going to be wrong. Extrapolating the future from the current direction is fine, as long as you realize that it is just that, and not necessarily an accurate prediction of future events.