Yup, that does really make sense. But apparently, when I start implementing in one approach and been through say 25% of implementation, I usually face some scenarios or hit blockers, at which point I kind of start thinking should I have been implemented in the alternate way to prevent the blocker. I spend sometime over there rethinking both the approaches. Only way, that I could think of to handle this to write down the whole approach, with pros & cons, potential roadblocks before even starting to write a single line of code to prevent this?