Mitch Wheat - Treat the cause, not the symptoms!

Roy Osherove has posted an excellent blog entry Interface Naming - Anything But Java's Standard, Please, on his thoughts on the use of ‘I’ for interfaces. I agree with Roy that using ‘I’ to prefix interfaces is a good idea because it makes code clearer. This is in line with the .Net Framework Design Guidelines by Brad Abrams and Krzysztof Cwalina. You should name interfaces to describe the behaviours they bestow on the implementing class, for example, IPersistable.

I personally believe that the ultimate aim of a programmer is to write code that reads like prose and should be clearly understandable by the reader. Any naming convention or coding standard that leads directly to more understandable code has to be a good thing.

As a young, naïve programmer writing C code over 20 years ago, I, like many others, took delight in writing convoluted, hard to understand code! Writing code with others in mind is not only more productive, but also a sign of maturity:

Any fool can write code that a computer can understand.
Good programmers write code that humans can understand. – Martin Fowler

Another interesting point that Roy mentions, is the reliance on an IDE to understand code through the use of ‘hover’ tooltips. Do you think it is inevitable that we should require tools to understand code, or should language syntax and the printed page be sufficient?