I very much like the idea of patch based version control; I am not a dumb person but find it hard to imagine what git is doing, when I enter git commands such as found from the web (non-intuitive commands). My question here, is even more basic as the difference between patch based and snapshot based version control : Suppose one has a source code file that contains a loop and a counter in that loop has been forgotten to be incremented, and it does not really matter where in the loop body that counter gets incremented when that gets corrected, as long as, say it is the second half of that loop body (for sake of this discussion). Now suppose person A develops a patch AA in which counter gets incremented right in middle of loop body and person B develops a patch BB in which counter gets incremented at end of loop body. Now as there is no conflict of lines it appears as if both patches can be merged without conflict. However, then, the counter gets incremented twice. This is a fundamental problem of automatic merging.