Seed Idea

I believe that one must have some seed idea upon which to build considerable amount of work and produce something concrete at the end. I should stick to my first and only idea, ePet and build upon it whatever is needed in the best possible way to deliver its potentials. For example what can be the mathematical formulation of the idea and what are the practical limitations, can be a line of research. For a start, I should build a working application of ePet in Symbian C++ and test performance based on various criteria.