 |
Taming the Monster
by Mike Morris
December 21, 2000
Those Were the Days
Mainframe computer programmers today are 10 to 20 times more productive than they were in the 1950s and 1960s. This is due to a number of factors, involving advances in hardware, software, accumulated expertise and human nature. In the early days, a mainframe computer was an awesome multi-million dollar marvel, and computer programming was a difficult business - far too serious for us aristocrats of technology to rush.
INTO THE MINEFIELD
In those days, programmers carefully flow-charted every little nuance of the code, erasing rectangles and diamonds, and re-routing flow lines until they were satisfied or exhausted. And then, after brushing off the eraser flakes, they would start a whole new battle with the coding sheet.
After laboriously coding a fairly simple program, we had to get it keypunched, and when boxes of cards were dumped on our desk we had to check that the keypunchers had exactly transcribed our gibberish code into holes in the punched cards. All this took time and reinforced our perception that we were working in a minefield and had to tread carefully.
It seems strange now, but in those days we had no idea of how to design a program. We knew code, and individual instructions. We wrote them down and coaxed them through the many steps required to get them into the guts of the computer. We wrote spaghetti code that you wouldn’t believe today.
We tortured the computer and it tortured us back, and after many weeks or months we had something that worked most of the time and was marginally useful. And all this hard work reinforced our belief that computers were very difficult to master.
A single test took days. After flowcharting, coding, and checking keypunch cards we had to wait for an audience with the computer. We guarded our boxes of cards jealously. A dropped box of cards meant hours of sorting.
We had to prepare JCL for a compile run first, and then our programs were offered to the clattering card reader in the manner of a human sacrifice. If it didn’t chew up the cards, the mainframe would usually print out compile errors, and we’d have to go back to the source code. When we finally got a clean compile we were ready for a real test on a real computer.
THE DINOSAUR
The early machines are difficult to visualize today, but I worked on one, years ago, and have never forgotten. It was a LEO, only used for odd jobs.
It was old even then, living in its own building, which looked like a holiday chalet. In fact, it filled the building. Trainee programmers filed through the door into the belly of the beast. We shuffled forward, single file, between ceiling-high banks of flashing vacuum tubes, clutching small decks of punched cards. It was like being in an amusement arcade on a hot summer day. We waited for our turns, sweating. A vacuum tube went ‘pop’, and a bored operator unscrewed it.
Our programs took a few minutes each to run, and were, I think, supposed to print our names and addresses on a dusty printer in the corner. Mostly, nothing happened, but the guy in front of me put his deck in the card reader and about a minute later the console lights froze. The head operator let out a howl of rage.
“They’ve done it again,” he said to no one in particular. “We’ll have to shut it down and re-IPL. No more tests today!”
We all left the building.
INTO THE FUTURE
As time went by, of course, the software got better, and the methodology got better, and we learned top-down coding, and how to incorporate standard routines. The kids coming up behind us began to treat computers like cars, to be souped up and tinkered with, and kicked a little when they didn’t perform, and eventually we got the message.
Still, in the early days, the hardware didn’t help us at all. When online processing via the CRT became available, we all began to pump out more code.
Coming to the States meant a quantum leap from the old way of thinking to the new for me. In Britain in 1968, the computer was still a scientific marvel to be treated with awe.
Americans had already gotten over their sense of wonder.
I learned that the computer was a machine to be used, just like a car, or a TV, or a washing machine. Programmers tended to grab the computer by the throat and force it to do their will. Pretty soon, I was doing the same, and producing much more.
SHE’S NOT INTIMIDATED
Today, mainframes and PCs chatter to each other constantly, and new methodologies and procedures and hardware have transformed computers - and the way we use them - beyond recognition. We make them do things that were impossible a few years ago. I watch my granddaughter, who has just learned to read as she bangs away at her word processor. Soon she’ll be hitting the chat rooms, and zipping around the world on the Internet.
She’s not awed. She handles her computer the way I’d handle a book. One day, she, or the kid at the next desk will come up with an idea that will revolutionize the way we use computers. It will seem so simple to them. They grew up with the computer. Their minds were transformed by it.
Computers, for me, were monsters to be tamed. To my granddaughter, they’re cute little puppies. She’ll make them do tricks I could never imagine.
|
