Thursday, June 10, 1971

Flying to Brookhaven with box-fulls of cards

I had, in 1971 or so, written one of the ur-algebraic symbol manipulation programs, my own, described in my first physics paper, of no interest whatsoever today (not sure I have a copy of the paper anywhere, and certainly not the code)

Our electron magnetic moment calculation was implemented by entering integrands on IBM punch cards, flying on a 4-person plane to Brookhaven with box-fulls of cards, and feeding batch jobs into the CDC computer there. 

We had no computer, but Cornell Synchrotron had several state-of-the-art DEC PDP-10s, so advanced that they had keyboards and screen terminals attached. I was allowed to use them after 1AM, only to edit, but absolutely not compute. So I wrote an symbolic algebra language for evaluation of Dirac gamma-matrix traces entirely as macros in TECO, the DEC text editor. Every time they checked on me, I was in editor mode, so they never caught me. 

Then Kinoshita went to CERN for a summer, and broke their computing budget - it was the first large physics calculation, as expensive as an experiment. I think my TECO code produced some errors, so he switched to SCHOONSCHIP and reentered all our integrands into it, without me, the amazing man that he is. 

I (Tom believes he) invented a wholly independent way of computing sets of integrals corresponding to self-energy insertions, and that enabled us to catch a single faulty Jacobian in a single integral, so our calculation had no errors. I have a ton of things written about that and this in my diaries. If it becomes of any interest to anyone, ever. For many years there was a large metal cabinet in a Newman Hall hallway, containing our entire calculation as punch cards - it might still be there. Only Tom knows. 

I suffer big father-son guilt there, because I used to call him up every few years, but then he exploded a car tire in his home’s garage, and went totally deaf, so I do not call him, and I never write letters (to anyone). But I assume he’s as frisky as always. 

We have a running thing going, because I have a crank theory that perturbation series for g-2 (and all on-shell physical observables) is not an asymptotic series which he doesn’t think much of. So 3 years ago I returned to it  - put maybe 1/2 year into it again, failed again, with nothing to show. But psychologically it must be that I wanted to show my (academic) father that I can do it. Psychology - I do not recognize any form of paternalistic authority - I only listen to women - and still...

The smart graduate student ("hrra-hrra T-uh-u-ft") I had always liked a lot. His adviser -it seems- I found grumpy, so he only features on the back cover of my ChaosBook.org/FieldTheory as "El Noble Cigaro", channeling Pauli, but I do not remember why I did not think all that highly of him.