Archive for March, 2014

Parting Words on Being Parted Out

After 50 years or so in the LLNL harness, last week Staff Relations called me into their inner sanctum and asked me to resign my position; they said I’d get a ‘buy out’ if I granted them their request. “This is all strictly voluntary,” they said, “You don’t have to do it, but be aware that a ‘buy out’ may not be offered in the future if you decide to resign at a latter time.” It didn’t look like there was any wiggle room here, so I played the ‘compliant employee’ card and resigned, and in the process composed the following prose:

The Barn Book

Fresh out of grad school, I showed up that first post-cooler day in 1964 carrying my very own copy of the Barn Book (neutron cross sections), with my trusty K&E buckled to my belt—I was 27 years old, neutron savvy and a whiz on the slide-rule.  Johnny Foster was LLNL Director and Sid Fernback was the head of its Computations Department. My only experience with computerized numerical solutions to differential equations had been recently gained in a but one semester class devoted to the subject; the largest program I had ever seen was the one I wrote in FORTRAN, ~100 lines, for a term assignment. Needless to say, I was not prepared for what happened next; I was assigned to one of A Division’s weapons codes, one that consisted of over 22,000 lines of assembly code. Not only that, no self respecting weapons designer would be caught dead without a Marchand electro-mechanical calculator capable of calculating to 15 significant digits. I stowed my Barn Book on my bookshelf, locked my K&E in a desk drawer and sat down wondering just exactly what had I got myself into—I felt so overwhelmed. Fast forward 50 years—still overwhelmed, still wondering.

When I look back over my 50 years of stuff, what leaps out at me is not my accomplishments and I’m almost positive there were some. No, it is the bugs I encountered, the ones that brought me to grief, sturm und drang, and made me doubt myself. Yet it is these I slew, each in its turn leaving the road stretching out behind me littered with their bloody corpses. No milestones here; gravestones is more like it. I could regale you with a recounting of the blow by blow battles I’ve fought over the years, but I won’t—except for one: George Bing, A Div physicist, was using my code to calculate the yield of a very modest device. The expected yield was just a few kilotons, but in trying to get the most detail possible out of his code runs, he had divided his design into a 100 zones. This was computationally the largest size design my code could handle. Up to this point, for most designs, the usual number of zones was significantly less than this to keep the cost of the computation down. Well, unbeknownst to me, the real maximum size computation was 99 zones; not 100. So? Well, at some point during the simulation, once fission was in full bloom, neutrons would begin to reach the surface of the device­—what was supposedly the hundredth zone,—at which point they would magically reappear in the first zone. The yield went through the roof and so did George who was paying for such ridiculous results. This was my very first bug and my very first experience with the confusion that existed just about everywhere over zero-based indexing. Back then, physicists (and FORTRAN programmers) were all wired for 1-based indexing, but assembly languages all used 0-based indexing. But that’s enough about me, for now; next I want to talk about Egbert Gittens.

Egbert Gittens

I think I Egbert’s and my paths crossed sometime during my first year at this place; he was a conspicuous figure, and was so for many reasons, not the least of which was his attire. Female employees were subjected to a strict dress code, i.e., figure hugging attire was verboten and nothing could jiggle., There was no dress code for male employees. Livermore, locally referred to as ‘South Hell’, in the summer time is quite hot, so the physicist who wagged the tail of the LRL dog back then, decided that suits and ties were out, and Bermuda shorts and sandals were in, especially in the summer. And not only that, titles were out, too, and we’d all be on a first name basis with each other. Now while the shorts were strictly hot weather attire, the sandals (always worn with socks) were more or less worn year round. Sid Fernbach was a bit of an anomaly in this regard; he wore expensive suits and bow ties to work every day, but with sandals and socks-white, cotton gym socks.  Egbert wore expensive suits and ties to work every day too, but with proper shoes and socks, and that was one of the first things that brought him to my attention. That, and that he had a very scholarly air about him, a professorial presence. He was not a technical person, but had received a classical education in Jamaica where he was born.  He spoke with a Jamaican accent in a resonant voice that would’ve rivaled that of Paul Robson had Robson been a tenor and not a baritone. So, to sum up, he was a striking person, a resourceful person, and energetic and full of ideas—altogether engaging. It was Sid who hired Egbert. He belonged here.

Somewhere along in the early history of this place, Egbert decided he wanted to write physics codes too, since that was after all the raison d’être of this place. With the help of people like Dick White, he learned enough physics and FORTRAN to be able to come up with a program of some kind, and this he’d carry with him as a large deck of IBM cards in a box. He always had it with him, it seemed. To my knowledge he not only never got any computer to compile his program, it was a rare day that he even got one to read in his deck of cards. (He didn’t understand that it was not enough to  just enter a sequence number in columns 73-80 of each card; the cards had to be in the right order when read in, in order for the FORTRAN compiler to make sense of them.)

Similarly he didn’t understand that the various steps to be performed in any procedure involving a computer had to be performed in a prescribed order, that it wasn’t good enough to create and follow a procedure of his own design, i.e., one that made sense to him. Consistency was not his forte. So, over the years Egbert would be moved from assignment to assignment, and then when found wanting, moved to yet another. Sid was his guardian angel, and wouldn’t hear of him being let go. I eventually left A Division, but before I did so, I was there during the time that Egbert was the CalComp operator. The CalComp was a computer controlled plotter, one that Lab physicists would use and rely on in their design work. It could produce high-resolution plots in multiple colors of ink on a roll of graph paper that was a meter wide. Its computer would inform the operator when it was time to unload the ink cartridge from the plotter head and insert one of a different color. These ink “cartridges” were basically designed along the lines of a Rapidograph ink pen, i.e., somewhere within the business end of a cartridge could be found a cylindrical lead weight attached to a short length of ridged wire about the thickness of a human hair. By shaking the cartridge up and down vigorously, this wire would slide up and down inside of the tube that delivered the ink to the graph paper. The cartridges had to be shaken up every so often, especially if they’d been inactive for a while in order to get the ink to flow properly again as they had a tendency to clog up. Otherwise the plotted lines would be full of blank spaces or faint gaps when the ink would have temporarily stopped flowing, or blobs and gushes of ink once it resumed flowing again. The CalComp was temperamental, to say the least, but in Egbert’s hands it was positively demonic. It wasn’t good enough for Egbert to just shakeup these cartridges; he had to dismantle them completely and give each one a good cleaning before inserting it into the plotter head. Because he was so conscientious, he would methodically clean all of the cartridges at the beginning of his shift. That doesn’t sound so bad, now does it? Actually, it was the responsible thing to do. However, as Egbert cleaned all these cartridge, smudges of indelible ink of various colors would end up on his fingers, the front of his fancy, starched dress shirt, his expensive tie, and his face, especially on and around the lips. (Some of the ink pens would be so clogged he’d have to resort to blowing through them.) Unfortunately and invariably the finished plots would be festooned with Egbert’s Technicolor fingerprints.  There for a while, each morning would begin for me with a serenade of howls, roars and screams from the physicists on the third floor of building 111 as they unfurled and viewed their ‘decorated’ plots from the previous night’s CalComp production run. By the end of Egbert’s tour of duty at the CalComp, they didn’t just want him replaced, they wanted him dead—most likely drawn and quartered. But Sid was deaf to their laments so Egbert stayed where he was…for a while.

For those of you who never knew him, Sid, more than any other person at this Lab, as the head of Computations, was responsible for ushering in the age of the super computer, because it was he, using Lab funds and Lab computer scientists, who spec’d and then financed their creation and acquisition early on.

The Lab as Sanctuary

Okay, this by way of introduction to what I really wanted to say: the management at our Lab is compassionate, empathic and caring, and committed to seeing that each of us gets a fair shake at what s/he needs in order to thrive in his/her various assignments. I have been here 50+ years and what I am saying here is true, for during these 50+ years I have been the recipient of such largesse on many occasions: once when my wife was being treated for cancer and I was her principal caregiver, management looked the other way when I shifted my hours from day shift ones to night shift ones; once when the DOE revoked my clearance having discovered I am a manic depressive, the very next day NSED management found me an office in an open area and an unclassified assignment for the 9 months during which I had to fight like hell, in court, to have my DOE clearance restored; and once after 7 years in A Division when I came down with a case of scruples and had been asked to resign from the Lab, Sid Fernback came to my rescue and gave me an office and assignment in Computations.

Okay, I’ve dragged my departure out as long as I dare; it’s time to go. The pleasure has been all mine I assure you. I love you all and will miss this place like hell.

No Comments