In this 1969 photo of the O.R. faculty at Stanford, George Dantzig is on the far left and Richard Cottle is on the far right. Others include, l-r from Dantzig: Alan Manne, Frederick Hillier, Donald Iglehart, Arthur Veinott Jr., Rudolf E. Kalman, Gerald Lieberman and Kenneth Arrow.

To gain background on the latter topics, I bought a copy of Saul Gass' 1958 textbook, "Linear Programming." From it I learned of George B. Dantzig and his towering role in the development of the subject. Fortunately, I also learned that he had joined the Department of Industrial Engineering at Cal and was looking for graduate students to work at the Operations Research Center. I went for an interview with Professor Dantzig and was hired — as a programmer. I left the debugging job to my successor at BRL.

What impressed me in the interview, as it did forever after, was George Dantzig's kindness. In going to meet this famous man, I had been expecting someone with a tough, Prussian-type personality. Instead, I met a gentle soul who welcomed me into his research world.

My first assignment concerned the famous Hirsch conjecture. I worked on a pivot selection rule Dantzig hoped would achieve the minimum. It didn't, and I got nowhere with the Hirsch conjecture. Rather than insist that I struggle on with that difficult topic, he kindly suggested another problem: to extend the quadratic programming algorithm he included as Section 24-4 of his book, "Linear Programming and Extensions" (LP&E). This I did.

I tossed my write-up on his messy desk, just as he had done with the problems he mistook for homework while studying with Jerzy Neyman decades earlier. Except in my case, the paper was never seen again. This was not a great loss to the scientific world, especially since Whinston and van de Panne had simultaneously discovered nearly the same algorithm. But as a by-product of this study, I was lucky enough to stumble across duality in quadratic and nonlinear programming. The latter was the subject of our first joint work. Further, it was the composition of primal and dual quadratic programs (analogous to what Albert Tucker had done with primal and dual linear programs) that led us to what was later to be called the linear complementarity problem.

In those days, computer programming was done with punched cards, and manuscripts were typewritten and ultimately checked on long galley sheets. Textbooks on operations research topics were rare. Those of us who studied LP under George Dantzig in the very early 1960s were exposed to his book "LP&E" in the form of galley sheets. A group of us became proofreaders or indexers of these sheets; part of my job was to coordinate the lists of typos we found and present them to George. This could at times be tedious. I remember well how painstakingly he traversed the minefield of L.V. Kantorovich's contributions on the classical transportation problem. We spent half an hour wording the last paragraph of page 299 of "LP&E."

Most of all, I remember with admiration and gratitude the many opportunities George Dantzig provided to his students, myself included. These included invitations to meet leading figures in the mathematical programming world, travel to professional conferences and social events. This generous approach was repeated for the benefit of dozens of students and is only one of the reasons George Dantzig was loved so much.

Reminiscences of George B. Dantzig The Life and Times of the 'Father of LP' By Saul I. Gass   My 'George Stories' Now Cherished Memories By Saul I. Gass, University of Maryland   He Made Powerful Ideas Seem Simple By Thomas L. Magnanti   Monumental Contributions Met the Test By Arthur F. Veinott Jr.   Never a Dull Moment with Brilliant GBD By Mukund Thapa

Table of Contents

OR/MS Today Home Page


OR/MS Today copyright © 2005 by the Institute for Operations Research and the Management Sciences. All rights reserved.


Lionheart Publishing, Inc.
506 Roswell Rd., Suite 220, Marietta, GA 30060 USA
Phone: 770-431-0867 | Fax: 770-432-6969
E-mail: lpi@lionhrtpub.com
URL: http://www.lionhrtpub.com


Web Site © Copyright 2005 by Lionheart Publishing, Inc. All rights reserved.