1)	Parse the Host: header that browsers provide. This should result
	in recovering the real host name.

	A simple implementation of the above that I have done of the above
	has resulted in about 90% of accesses to not have to use the DNS at
	all! The implementation still needs some work.

2)	Convert it into a non-forking server that handles multiple connections
	in a single process. Either by a threads library, or the more
	conventional state machine approach.
