You like command-line parsing, but you hate all of the bloat. Why should you have to create a Hash, then create a parser, fill the Hash out then throw the parser away (unless you want to print out a usage message) and deal with a Hash? Why, for Pete's sake, should the parser and the parsed values be handled by two different objects?