Video Excerpt

IRV
As you look through the history of the field, what project or event has made you proud to be part of it?

PHIL
One bit of fun I had was deciding I really wanted to get into the computation of linear programming. All we had at that time was a pretty elaborate linear programming routine written in machine language by Bill Orchard-Hayes for the IBM 704.

I wanted to use it, get into its innards, but I absolutely couldn't understand it. It was way over my head. Maybe I was reluctant to learn that much machine language at that time to find out how it worked.

Bill had just quit in 1957 to join CEIR (Corporation for Economic and Industrial Research), which Lester Ford joined later. I was to carry on the linear programming software effort that George Dantzig was interested in carrying on. I worked on the code for a while and finally gave up.

I heard about Fortran and it seemed like a wonderful idea: a higher-level language. You could write something like mathematics and have it translated. I nudged the Rand Corporation into installing it, and worked with it for a long time amidst the quiet sneers of my computing colleagues. They thought it was a very effete and ineffective way to go about writing computer programs.

I persisted anyway. And one of my great moments came after putting together a linear programming code that I felt was pretty good. Some friends at Standard Oil Company of California in San Francisco were bragging about their assembly language code. And I said, "Why don't you try Fortran?"

I sent them the results of the test problem I'd run. They refused to believe that I was getting a better time than their program. Finally, they checked it all out and, indeed I was! Winning a race like that, outside of my field as a mathematician, was a real pleasure.

Previous

Next