One thing I have found to be a very valuable habbit is to first think about what your software has to do on paper and draw some shitty flow charts, lists and other things, without too much care about whether you will do it (especially if it isn't software that you strictly need to do for some reason).
Whether an idea is good or not can often only be judged when it becomes more concrete. The actual finished project is as concrete as it gets, but it takes time and work to get there. So the next best thing is to flesh it out as much as possible ahead and decide based on that whether it is worth doing it that way.
Most people have the bad habit of being too attached to their own ideas. Kill your darlings. Ideas are meant to be either done, shelved or thrown into the bin. It doesn't do any good to roll them around in your head forever.
Whether an idea is good or not can often only be judged when it becomes more concrete. The actual finished project is as concrete as it gets, but it takes time and work to get there. So the next best thing is to flesh it out as much as possible ahead and decide based on that whether it is worth doing it that way.
Most people have the bad habit of being too attached to their own ideas. Kill your darlings. Ideas are meant to be either done, shelved or thrown into the bin. It doesn't do any good to roll them around in your head forever.