Monday, 12 November 2007

Two talks of Joel Spolsky in Dublin

Both of them took place on November 7th. The first one was in the morning and Joel Spolsky was presenting FogBugz. Explaining the product on feature by feature basis Joel managed to tell the audience a little bit about how he thinks the software development process should look like. And you know that people like him are right in 99% of cases and you can follow them blindly ;).
FogBugz is a great piece of software that provides many desired by every software house functionalities like bug tracking, project tracking, Wiki, discussion groups, evidence-based scheduling and even a small help desk system. All these features are seamlessly integrated together in a way that the user is not aware when he/she moves from one part of the system to the other. Every activity that requires user attention is as simple as possible and what is very important as natural as possible. It's enough to say that whenever you need to specify an end date you can simply type word "week" and the system will calculate the date based on the current date. Adding new tasks to your to-do list is simpler then firing notepad (actually I do this every morning to keep my 5 most important task always in front of me). When you start working on a task you just hit start button and when you are finished you hit stop. The system calculates everything for you taking into account lunch break, scheduled meetings, your working hours, etc.  Those are small things but they show how the whole application has been build. Basically, FogBugz tries to be as transparent as possible to let you focus on your real tasks.
The next thing that impressed me a lot was the evidence-based scheduling. If you work with FogBugz you will never get just the end date of your project. Instead, you will get a set of dates and probability assigned to every of them. Taking into account how many IT projects are completed on time and on budget then this must be a better approach than what we have now. Needles to say that it does make sense especially when you take into account Quantum Physics.
The second presentation was part of the IJTC launch and Joe was talking about a topic that is as incomprehensible to software developers as the fact that they should not spend half of their life in front of a computer :). He explained how Companies take advantage of misattribution to sell us more of their products. His example was based on comparison of iPhone and a phone X of company Y. It's X and Y because I don't remember their names which is the best evidence that misattribution works :). Basically iPhone has fewer features, it's not extensible for the time being, it doesn't have a replaceable battery, it's 4 times more expensive and the features that are present in both phones are better implemented in the phone X. Why Apple has sold millions of iPhones and the company Y can only dream of something like that? iPhone looks way better and it's much more user friendly. People then attach the UI experience to the rest of it and that's how iPhone as the whole seems to be perfect in their minds. That's what Joel calls misattribution. Although I suppose that the marketing success of iPhone is a little bit more complicated beast I still think that his point is valid and it was certainly refreshing to me.

No comments:

Post a Comment