Il 08/03/2010 22:16, Andrew Sutton ha scritto:
Hopefully, you don't take from what I just wrote that I think that BGL are more important than your thesis. I'm just referring to how you might write the proposal to make it more competitive.
Sure, don't worry about! ;-) What's the timeline for BGL-algorithms proposal? My idea is to submit two kind of algorithms as solution of the same problem, two-graphs common-tree problem, where the latter is the refinement of the former (better performance, about 300%!). Authors are Grimbleby for the former, and Schach for the latter. MRT-based (Schach) implementation is ready, generic as BGL algorithms need to be, and it works fine (I use it in my degree thesis). I've not found since January/February a little bit of time to develop the other algorithm, I'm very sorry. However, I can write down not-generalized pseudo-code for the Grimbleby algorithm (to drive to C++ code, of course), and submit the one by Schach, as-is. What about this way? Michele