[ilaksh]

To me there is an inherent contradiction between the concept o a hackathon and any actual effective effort produced during one. The idea is that you only work on your project during that day. However in the current era just about any worthwhile programming effort will draw heavily upon years if not decades of previous work. I would even guess that the most interesting efforts at hackathons are the ones most directly related to an individuals current area of expertise.

Many good programmers today are quite adept at modularization or other related efforts like framework building. By utilizing existing frameworks and modules one can leverage an enormous amount of existing code. I think though that many of the most worthy efforts at hackathons actually leverage existing modules or frameworks that were actually built or co ntributed to by the particular hackathon participant who is using them.

I think because of this contradiction and because it is difficult to distinguish between a completely new project and one that is a continuation of existing work that we need to alter the concept somewhat.

Getting together to hack on existing projects and describe them is probably a more honest and generally applicable idea. If you really want to limit people to work done on that day you are going to need to be very specific about what types of existing modules and frameworks can be used and whether the individuals participating may have created those. I'm not sure I really see a great point to that sort of exercise though unless it is just for developers to demonstrate how many interesting techniques and algorithms they can carry with them in their mind. That seems kind of a primitive thing to test for when we have so many great ways of reusing code.