I'm a Critical embedded systems developer with a very honest passion for software engineering. Currently serving as Manager and Lead of a software architecture team. Although my profession is in embedded development I cut my teeth early in desktop applications, mostly Windows based and enjoy continuing to execute those skills in my hobby projects (only I only have interest in supporting Linux or mobile apps now). Truly appreciate good software design documentation and being able to complete it efficiently so I'm terribly disappointed with the direction UML tools have gone, and the lack of good use of them in the industry. UMLet is an interesting path to me and I'm very interested to see where it goes, but I have a different vision in mind and would like to get some time to pursue it. The rules I have in mind for that pursuit are as follows:

1. A tool meant to support a diagram based language absolutely must be a great diagramming tool
2. Generating diagrams from code is not capturing design, its capturing implementation, which your code already does
3. Generating complete code from diagrams just means you skipped design and put all your implementation into a language that is really inefficient to implement in
4. If its of critical importance to you that your implementation looks like your design, you've missed the point of design
5. The mouse as and input device is fine for those not interested in getting done with their work in a timely manner