Thanks, a lot of helpful and interesting ideas here!
I often seem to get stuck in my projects because of overanalyzing and inability to decide the best way to tackle a problem, for example, what classes, modules or libraries should I break the software into. I might have a couple of options in mind, but none of those feels just right. I may be falling into the trap of trying to find a too perfect solution. Another thing I notice, a bit like the poster above, is that at some point I can't seem to fit all the code or architecture in my head, and at that point the project starts to feel "foggy" and overwhelming.