Here's a slight ammendment to Fabrizio's comments on pre-historic SAS.

The developement started around 1967 at NCSU under a Federal grant. The original developer was Jim Barr. Jim Goodnight came on board a year or so later. I started grad school at NCSU in the fall of 1969 and the packaged existed on the mainframe in a rudimentary form. In the very early years you submitted a mainframe job that printed a manual for you. I seem to recall an actual printed manual that was perhaps a half inch thick. As far as I know, the package was being offered on the market for a nominal fee. SAS 72 was released while the project was still housed on the NCSU campus. I finally left school in 1973 and at that time, a copy sold to a business cost $100 or so (government prices were cheaper due to the grant). It was fairly soon after this that the package was deemed to be a viable product and the grant stopped. John Sall had joined by this point. The group moved accross the street from NCSU and the institute was born. Jim Barr left around 1980.

Nat Wooding