Andre Poenitz wrote: > There is always a choice.
Yes, the choice is there. It's only that We take it. ;) No, I'm joking. Only that for learning something, I would like to have explanations on these decisions. For instance, is not that the boost graph code fits like a glove. It doesn't have f.ex. shortpath searches with multiple sources. Nothing that cannot be overcomed, but I don't expect the code using boost to be extraordinary shorter than the one I sent. Also, how much testing does the boost graph library has? How widely used is? (I'm asking because I don't know) > My approach to boost is something like > > if ((I can implement it in 20 lines myself) && > (it is portable) && > (it is needed only once or twice)) > do_it_myself > else > use_boost > > Andre' > Hmm... the pseudo code you posted for short path was 75 lines long... Alfredo
