Googling gives me this:
Command A & Command B
Execute Command A, then execute Command B (no evaluation of anything)
Command A | Command B
Execute Command A, and redirect all its output into the input of Command B
Command A && Command B
Execute Command A, evaluate the errorlevel after running and if the exit code (errorlevel) is 0, only then execute Command B
Command A || Command B
Execute Command A, evaluate the exit code of this command and if it's anything but 0, only then execute Command B