Software Development with Linux

First iteration must fail

MON, 01 AUG 2011

Iterating is about getting feedback, fast. It is well known that shorter iterations are better. Everyone should aim at getting an iteration as short as possible for a particular project.

What that means is if your first iteration is a complete success (all the features are complete and the client is all happy about your work), you failed to iterate correctly. You should have had shorter iterations. You should have showed your work with the client sooner.

Now, it is possible that you were lucky or that the task was small and easy enough that no iterations were required. But when things are simple and easy to understand, they are not fun to work on. ;-)

So what if your first prototype or draft document is shut down fast? You should be happy about it; you just had some great feedback to improve your works. That will made the end-result better and more in line with what the client wanted.

Now, do not produce shitty stuff on purpose. You simply don't need to push it too far before showing it, before validating what you have done with the client. Get something workable and get comments immediately. You have no time to lose working on the wrong aspects.

In the end, you will gain confidence in your work, and your client will have exactly (no more, no less) what he wanted. Everyone wins.