> 5. Test Early, Test Often, Then Test Again I'm not convinced that the kind of Usability Testing(TM) presented is actually the best way. There is still a phys

> 5. Test Early, Test Often, Then Test Again I'm not convinced that the kind of Usability Testing(TM) presented is actually the best way. There is still a physical barrier (a wall) and a communication barrier (the usability tester) between the developers and the users. Why is that a good idea? Despite saying it's "not rocket science", the requirements listed are still a lot heavier than it need be. We aren't doing impartial science here. We are trying to understand users. Why not just _talk_ to them and interact together?