This document is an ever growing collection of memorabilia, contemporary documents, and anecdotes recounting the history of UNIVAC 1100 series mainframes. The first computer I ever used was a UNIVAC 1107, and for more than a decade stretching from 1967 through 1978, most of my programming was oriented toward those machines, spanning four generations of hardware: the 1107, 1108, 1110, and 1100/80 (which I used briefly to develop microprocessor software).
As the collection grows, I will organise it more coherently and make it easier to find your way around. In these early days of the archive, it's simply a collection of unrelated documents linked to the items below.
Enjoy, and may minus zero never be (completely) forgotten.
In September of 1967 I wrote my first computer program and ran it on this machine. This illustrated document recalls the characteristics of a supercomputer, 1960s style.
The ultimate fashion statement for assembly language programmers of the Case UNIVAC 1107 wasn't a pocket protector and pens of many colours but these plastic laminated cards documenting the instruction set of that computer.
People who have used a 100 megabyte hard drive that weighed two and a quarter tons and cost more than US$130,000 in 1968 experience a special sense of wonder when tucking one of today's 2.1 gigabyte drives, just purchased for less than US$1000 and weighing less than half a kilo, into their pocket. The FASTRAND Page turns the clock back to the days when mass storage was massive.
In 1968 you could pick up a 1.3 MHz CPU with half a megabyte of RAM and 100 megabyte hard drive for a mere US$1.6 million. Oh, and you want a printer too…?
What UNIVAC programmer can ever forget the day he or she found a way to use an instruction like “Magnitude of Characteristic Difference to Upper”? Here are all the instructions for the 1100 family from the 1107 to 1110, colour-coded to indicate which machine included what instructions, with SLEUTH I and SLEUTH II mnemonics.
UNIVAC 1100 machines used one's complement representation of negative numbers, as opposed to the two's complement form almost universal today. In one's complement, where zero is a signed number, you have not only zero, but minus zero to contend with. This document explores various representations of negative numbers and describes some of the bizarre consequences and applications of minus zero.
Nothing better illustrates what could be accomplished with the rather odd UNIVAC 1100 instruction set than this program, which edits a 36-bit number into a 12 digit octal string in exactly 12 instructions, with no loops.
Programmers of UNIVAC 1100 mainframes in the 1960s through the mid 1970s exalted in discovering clever ways to squeeze the most out of machines which, by today's standards, would be considered hopelessly slow and short of memory. Here's a collection of assembly language and FORTRAN coding tricks for the UNIVAC 1100 collected in 1974 and 1975 by Walter Gilbert of the University of Maryland Administrative Computer Center.
If they can put a man on the Moon, why can't they put a UNIVAC 1108 in your shirt pocket? Well, a few months before a man walked on the Moon, programmers at Chi Corporation in Cleveland did the next best thing—a shirt-pocket reference card for the UNIVAC 1108 packed with essential information for assembly language systems programmers. Can't find yours after all these years? Here's your own personal copy, on the Web.
Back in the '60s, almost every mainframe computer had its own character code. The UNIVAC 1100 machines used a 6-bit variant of FIELDATA code, originally designed by the U.S. Army.
Programmers of the mainframe era tended to be better typists than those who started with timesharing terminals or personal computers. Why? Because the only way to get your program into the computer was to punch it on cards, and as you discovered within seconds after sitting down at the keypunch, “you can't erase a hole”. Since the computer only reads the holes, cards could be any colour and bear any design, and many sites printed their own custom cards, sometimes quite elaborate. Here's a gallery of punch cards from UNIVAC sites, including a rare 90-column card dating from the era when 80-column “IBM cards” were the exclusive property of IBM.
This hilarious mid-1970s poem by Harry Gilbert, in the spirit of Lewis Carroll's Jabberwocky, packs more EXEC-8 acronyms than you can probably recall into a delightful story which can't help but summon memories of the late night “code rage” which occasionally possessed even the most mild-mannered UNIVAC system programmers.
In 1971, John Walker figured out how to convert a multi-million dollar room-sized UNIVAC 1108 mainframe into a Morse code practice oscillator with a stand-alone, bootable operating system dedicated to that dubious application. Here it is, complete with source code.
Ever heard the one about the self-reproducing ANIMAL program that spread throughout systems and colonised new machines? I wrote it. Here's the story, both as it actually happened, and recounted in the press as a “computer urban legend” 10 and 15 years after the fact.
Freeware, 1972-style! FANG was one of the rare applications to fully exploit the multi-tasking, multi-processor, and multiple I/O channel architecture of the UNIVAC 1100 series. A file and tape utility, it allowed simultaneous execution of any number of commands, while guaranteeing results identical to the much slower serial execution employed by UNIVAC's own utilities. FANG was freely available, first released in 1972, and was used by hundreds of UNIVAC sites. It was last modified in 1975, and remains in use today, more than 20 years later. To view this document, you need a browser which supports frames.
UNIVAC has been, over the years, a registered trademark of Eckert-Mauchly Computer Corporation, Remington Rand Corporation, Sperry Rand Corporation, Sperry Corporation, and Unisys Corporation. FASTRAND is a trademark of Sperry Rand Corporation, since merged into Unisys Corporation.