Re: An answer to metadata, complete.

[James Smith <j-smith physics tamu edu>]

There's already a way to determine file types - Apache uses a mixture of
magic numbers and file extensions.  All we need now is a database
telling us how to implement the different methods for each mime type. 
Something like

		text/plain.read less $@
		text/plain.edit vi $@
		image/jpeg.read ee $@

or whatever variables we would want - could borrow from Makefiles.