"After the Johnniac had a memory change (to core planes) Willis froze the design and worked at Rand on other subjects.  The Johnniac then was used as a production machine for a number of years and finally got new software which turned it into a small online service: JOHNNIAC Open Shop System (JOSS).  It supported several systems over the years, including Newell, Shaw and Simon’s Information Processing Language (IPL).

Meanwhile, Willis got involved in starting the profession.  He and Keith Uncapher organized AFIPS and both sequentially chaired it for a while.  It sponsored big conferences to make money to use for good purposes.  One of their projects was to find volunteers to write "best practices" manuals to guide the developing field.  They sponsored one on security, the author they chose fouled out, and I took over and published the first manual on computer security in 1962 or so.

Willis had a quiet non-threating personality.  He worked well with high military officials who wore their egos on their sleeves.  I worked on his team in trying to fix the Air Force logistics system development, which was as big as IBM's OS/360 but without as much talent.  It failed and the USAF kept reviving it under another name until Congress made it a line item in the USAF Budget so they could kill it.

He helped me push through a formal Rand Recommendation to the AF Chief of Staff to establish a series of skill codes within the officer corps so officers with computer skills could be identified and assigned correctly.  I could not have done it without him.

He and I had several intelligence assignments separately and together.  We both held TS clearances.

He pushed through the Department of Defense Computer Institute (DODCI) which is still going under another name.  We both were initial lecturers.  The first class of students were all Generals and Admirals.”