Original URL: http://www.theregister.co.uk/2012/11/01/a_history_of_computing_in_20_objects_part_one/

A history of personal computing in 20 objects part 1

From the 17th Century to the 1970s

By Tony Smith

Posted in Vintage, 1st November 2012 12:00 GMT

Archaeologic Personal computing. Personal. Computing. We take both aspects so completely for granted these days, it's almost impossible to think of a time when computing wasn't personal - or when there was no electronic or mechanical computing.

To get from there to here, we've gone from a time when 'computers' were people able to do perform complex calculations themselves, through mechanical systems intended to do the work for them and then to powered machines able to automate the process. These led to systems that could be programmed to perform not only mathematical tasks but to store and retrieve other forms of data, taking us right up to desktop devices for a one-on-one interaction with computing power.

Apple 1

I, Apple I

Since then, that power has been compressed into smaller, more convenient packages: laptops, tablets and smartphones.

What a trip. In memory of the many people who have help us along, here then are some of the key stages of that journey, represented by the 20 objects that, to us, most embody the steps that brought us to where we are today.

It's not a comprehensive list - and feel free to comment with the devices you think we should have included - but here are the first ten of our 20 items, from the early days up to the end of the 1970s. Part two will bring us from the 1980s to the present day.

Computing's Prehistory...

Napier’s Bones

Napier's Bones
Reg Hardware retro numbers

As if introducing logarithms and evangelising the use of the decimal point weren’t enough, 16th Century Scottish mathematician John Napier (1550-1617) also devised a calculating machine able to multiply and divide. It could even work out square roots. There was initially no mechanism as such: just a series of rods - the ‘bones’ - marked with two columns of numbers. To calculate a multiplication, line the rods up side by side so the figures at the top of each show the digits in the multiplicand. A column of numbers on the left side of the apparatus provides the multipliers: read across the appropriate row, right to left, adding numbers grouped by diagonal lines, to generate the digits of the answer. Division is performed in a similar way. The rods were soon replaced by cylinders, each of which replicated all nine rods and could be rotated to ‘set’ the a given digit in the multiplicand.

Babbage’s Difference Engine

Babbage Difference Engine No. 1
Reg Hardware retro numbers

Charles Babbage (1791-1871) first came up with the notion of a calculating machine capable of printing out perfect logarithm tables in 1812 at the tender age of 21. Seven years later, he set about producing such a device and, in 1822, he and his tame mechanic had put together the first Difference Engine, a clock-like assembly of interlocking cogs able to calculate squares and solve quadratic equations. It worked sufficiently well for the government and other institutions to invest in the production of a large, more capable version. It proved too complex and was never completed - the project was effectively abandoned in 1834. By this point, despite many personal and professional setbacks, Babbage had his eye on the Difference Engine’s successor: a multi-function, programmable ‘Analytical Engine’. It too was never built, but nonetheless prompted the writing of first computer program.

The War Years...

Zuse Z3

Zuse Z3 reconstruction at Deutsches Museum, München

Source: Venusianer/WikiMedia

Reg Hardware retro numbers

Devised by Konrad Zuse (1910–1995) in 1941, the Z3 was the world's first programmable electromechanical digital computer. It was built in Berlin and was used to analyse aircraft wing designs. Nominally a general-purpose machine, it nonetheless featured an instruction set heavily oriented to solving engineering problems. The Z3 comprised 2000 electric relay switches to store and operate on binary numbers. Programs and data were stored on punched film, and it was able to crunch five to ten numbers a second. The Z3 continued in operation until 1943 when it was destroyed in an Allied air raid. Zuse went on to develop the Z4, a more advanced version of the Z3, and to pioneer computer design on through the 1950s and 1960s.

Colossus

Colossus

Source: Timitrius

Reg Hardware retro numbers

Colossus, the first electronic programmable digital computer, was created in 1943 by Post Office engineer Tommy Flowers (1905-1998) and others as a faster, more reliable alternative to the electro-mechanical rigs then being used to compare German coded military messages with the entire possible output of the Nazis’ Lorentz cypher machine. Flowers believed his electronic machine could don the job more quickly and more reliably, and when it was put to work in 1944 it quickly proved itself a success. It was able to compare message at 5000 characters per second, almost three times the speed of the electro-mechanical system. Flowers immediately began work on an improved model which formed the basis for nine more machines installed in code-breaking centre Bletchley Park throughout the remainder of the War. Flowers’ work remained a secret until the 1970s.

ENIAC

ENIAC programmers, photo: US Army
Reg Hardware retro numbers

ENIAC (Electronic Numerical Integrator And Computer) began life in 1943 as a University of Pennsylvania project to devise a machine capable of calculating ballistic firing solutions for the US Army, though it was first used for hydrogen bomb design calculations. Conceived by John Mauchly and J Presper Eckert, ENIAC was completed after World War II in 1946 and so its existence 1946 was made public in a way Colossus and the Z3 never could be. Designed as a general purpose computing machine, ENIAC contained almost 17,500 vacuum tubes, and used IBM punch cards to present results and to take in input data. Data was stored not in binary form but as decimal numbers. Like Colossus, ENIAC was programmed by setting switches and wiring.

Computing goes Commercial...

LEO

LEO I, credit Leo Computing Society

Source: Leo Computing Society

Reg Hardware retro numbers

The world’s first business computer, LEO 1, was developed by bakery and restaurant chain Lyons in the late 1940s. It grew out of Cambridge University’s EDSAC (Electronic Delay Storage Automatic Calculator) project, itself inspired by ENIAC. The development of EDSAC was accelerated by Lyons managers, led by David Caminer (1916-2008) seeking to put the nascent computer technology to use making their post-War business operate more efficiently. EDSAC had the ability to store programs, though no high-level language to allow their creation. Lyons engineers nonetheless were able to use its 31 basic instructions to create a coded program - fed in on card rather than wired in place, as was the case with earlier computers - that was able to track and cost the labour and material of cakes, biscuits and bread moving through Lyons various profit centres.

DEC PDP-8

DEC PDP-8
Reg Hardware retro numbers

Digital Equipment Corporation (DEC) introduced the 12-bit PDP-8 - also known by the company as the ‘Straight 8’ - in March 1965. At a time when existing computers were large, room-sized systems, the PDP-8’s compact, fridge-sized design made it the first co-called “minicomputer”, a machine designed to provide computing power to small groups of people. It contained memory enough to hold 4096 12-bit words, while the unit’s processor ran to two registers. The CPU was operated using just eight instructions. DEC would go on to sell more than 300,000 PDP-8s and variants before discontinuing it in the mid-to-late 1970s, pushed out by the growing demand for truly personal computers.

Onto the Desktop...

Xerox Alto

Xerox Alto
Reg Hardware retro numbers

The Alto was an experimental machine built by boffins in Xerox’s Palo Alto Research Centre (PARC) in the early 1970s to explore new thinking in user interface design, and while never made available commercially - Xerox would sell the Star, a version of the Alto, in 1981 - a couple of thousand were made for use by Xerox staff and some were donated to universities and research facilities. Arguably the first personal computer - though some historians consider it a minicomputer - it was also the first to feature a graphical interface controlled by a mouse and to incorporate networking.

MITS Altair 8800

MITS Altair 8800
Reg Hardware retro numbers

The Altair 8800 may not have been the world’s first personal computer aimed at hobbyists rather the scientists and engineers - that was 1973’s Scelbi-8H - but it was the first of the switches’n’lights home micros to gain wide appeal. Manufacturer MITS (Micro Instrumentation and Telemetry Systems) had been founded in 1969 to build electronics for tracking model rockets, though it soon turned its attention to programmable-calculator kits. A single-user computer was the logical next step, and work began in 1974 on an Intel 8080-based machine to be sold in kit form. A crucial tie-in with US magazine Popular Electronics helped put the Altair in front of thousands of potential users in December 1974. The assembled unit was a pain to program: instructions were fed in by flipping switches on the front. That changed when a small, two-man company called Microsoft was hired to write a Basic interpreter for the machine, fed in on paper tape.

Apple II

Apple II with monitor and floppy-disk drives
Reg Hardware retro numbers

Launched in April 1977 and shipped two months later, the Apple II is probably the member of a small group of first personal computers launched that year - others include the Commodore Pet and Tandy’s TRS-80 - that did the most to take the technology beyond the reach of hobbyists and into the mainstream. The work of Steve Wozniak (1950-) and Steve Jobs (1955-2011), the Apple II was equipped with a 6502 CPU and 4KB of Ram, colour graphics and - an interesting innovation - expansion slots, the Apple II proved hugely popular, despite being priced well above the competition, largely because it quickly established an extensive library of third-party software users could run to get things done.

In Part Two, Tomorrow: Objects 11-20 - the 1980s and beyond