Eh, my knowledge belongs to me :) And as long as you don't violate patents, you can redo stuff from scratch as often as you want.
In general it's not smart to look at code whose copyright may conflict with yours (or your intended copyright) and lift ideas directly. That was probably people's main concern back then.
A knowledge management package done from scratch would look very different anyway: it has to be based on the content repository, its TCL could be roughly based on the survey and/or the cms package, and (I hope) it will be developed for Postgres first and then (maybe) ported to Oracle.
Philip Greenspun's Problem Set 4 would provide a good starting point for the philosophy of such a package. Is that still available?
