Your heroes

When we asked readers to reveal their unsung IT heroes, you told us tales of code-breaking and 1Kbyte PCs. John Kavanagh reports

When we asked readers to reveal their unsung IT heroes, you told us tales of code-breaking and 1Kbyte PCs. John Kavanagh reports.

These five heroes of computing - from a Second World War hero whose developments had to stay secret at the time, to later heroes who invented technologies that have become industry mainstays - are at the forefront of innovators nominated by Computer Weekly readers. The nominations came in response to our original "Unsung heroes" article (Computer Weekly, 28 June), looking at those who have made outstanding technical or business contributions to IT and deserve a bigger share of the spotlight.

Tommy Flowers (1905-1998)

Tommy Flowers helped to win the Second World War by developing the Colossus computer to replace electro-mechanical machines used to crack coded German messages, even though experts thought his plan was impossible.

Flowers was an engineer at the General Post Office, which also ran the telephone service, and he developed arguably the first fully electronic digital computer, although it was dedicated to just the one purpose of code breaking.

In just nine months, the first Colossus was designed and installed at the secret Bletchley Park code breaking centre. It cut deciphering time from weeks to hours.

"The Colossus contribution included providing vital information in the run-up to D-Day, showing that Hitler had fallen for the Allies' deceptions and believed any invasion would be farther north," says Tony Sale, who is leading a Colossus rebuild at Bletchley Park, now a museum.

"The Colossus needed 1,500 electronic valves but Flowers was confident it would work. He had already designed Post Office repeaters using valves and knew they were reliable - as long as they were never switched on and off. Nobody believed him, because of their own experiences with valve radios.

"It worked so well that when I run a Colossus emulator on my 800MHz PC it takes the same time as the original machine."

Yet this historic computer was so secret that information about it did not start to emerge until 1970. After the war, eight of the 10 machines used at Bletchley Park were dismantled. The other two went to the GCHQ defence communications centre and were dismantled around 1960 - and the design drawings were burnt.

Flowers returned to the Post Office and worked on telephone exchanges.

Flowers had more nominations from Computer Weekly readers than any other unsung hero, even though in effect his work came to a dead end in terms of the development of computing because of the secrecy.

His contribution is summed up by reader Graham Tapper, principal consultant at Computer Associates. "The Colossus was critical to breaking the German code and helping to shorten the war and save countless lives."

Jack Kilby (1923-2005)

Jack Kilby is nominated as an unsung hero by Computer Weekly reader Alan Fineberg of Fujitsu Services for the invention of the integrated circuit.

The significance of that invention at Texas Instruments in 1958 is reflected in the fact that it won Kilby a Nobel prize and the two highest US science awards: the National Medal of Science and the National Medal of Technology; only 13 people hold both.

Electronics had evolved from temperamental glass vacuum tubes to transistors in the late 1940s, but still thousands of these had to be soldered together on circuit boards.

In 1958 Kilby conceived, designed and made the first electronic circuit in which all the components were in a single piece of semiconductor material smaller than a thumbnail. The first circuits contained just a handful of components but they were the start of a development that led to today's chips, containing millions of components.

Texas Instruments got orders from the US military for specially designed chips for missiles, but techniques for mass production are credited to John Hoerni at Fairchild Semiconductor, in 1959. Hoerni's work was developed further by Fairchild's research director, Robert Noyce - who later set up Intel with Gordon Moore.

Meanwhile Kilby moved through several technical management jobs at Texas Instruments, notching up more than 60 patents for a variety of inventions. He is credited with co-inventing the pocket calculator and the thermal printer used in portable data terminals. He became distinguished professor of electrical engineering at Texas A&M University in 1978 but also continued as a consultant to Texas Instruments until he retired in 1984.

The company's obituary of Kilby says, "From Jack Kilby's first simple circuit has grown a worldwide integrated circuit market with sales last year totalling $179bn. Such is the power of one idea to change the world."

Things could have been very different if the UK government had persevered with work by Ministry of Defence engineer Geoffrey Dummer, who came up with the same idea as Kilby  six years earlier, in 1952.

Kilby was unaware of Dummer's work at the time, but years later he said, "Had we known about it, it might have accelerated the timing of things."

Donald Davies (1924-2000)

"Donald Davies is one of the most overlooked pioneers of the internet," says Computer Weekly reader Liam Madden. Davies is one of three people credited with developing the internet's communications method of packet switching in the mid-1960s. They each apparently worked with little or no knowledge of the others' efforts.

While Davies worked at the UK's National Physical Laboratory, the others, Paul Baran and Leonard Kleinrock, were in the US and got involved with Arpanet, which ultimately evolved into the internet. But the head of the Arpanet project, Larry Roberts, was very familiar with the UK developments through conferences and meetings and later acknowledged the importance of Davies' work on packet switching.

US and UK historians differ on the various influences of the three packet switching pioneers. Davies barely gets a mention in some US histories of the internet, whereas the National Physical Laboratory says his work "provided a much needed steer to the development of the Arpanet" and therefore "heavily influenced the development of the internet".

There is at least no argument over the fact that Davies coined the term "packet switching" for this idea of splitting data transmissions into blocks and sending them separately, possibly by different routes, and thus catering for disruption or heavy use of certain routes.

"Donald Davies was a remarkable man," says Madden, who teaches software engineering at London University's Imperial College. "He got a first-class degree in physics at 19 and immediately followed this with a first in mathematics."

Davies joined the National Physical Laboratory in 1947, working with computing pioneer Alan Turing on the Pilot Ace, one of the world's first digital computers to store a program in memory. Davies got interested in applications and led developments including systems to simulate road traffic and to translate Russian documents.

After his work on packet switching, Davies saw the security risks in the growing field of data networking, and focused on this area. His team was consulted by organisations including many UK banks and government agencies.

Madden concludes, "Another British innovator, Isaac Newton, wrote, 'If I have seen further it is by standing on the shoulders of giants,' and the same can be said of subsequent internet technologies, which could not have become reality without the ground-breaking work on packet switching. Donald Davies never received the acclaim that his work deserved."

Ted Codd (1923-2003)

Ted Codd presented a paper in 1970 that changed thinking about databases and in effect led to the formation of Oracle, Sybase and other companies in this field. His impact is summed up by Computer Weekly reader Loyd Gravitt's brief response to the appeal for nominations of unsung heroes: "Ted Codd, of course."

Codd was born in Dorset, gained degrees in mathematics and chemistry at Oxford University, flew with the RAF in the Second World War then joined IBM in the US as a programmer.

He invented an early multiprocessing system and later moved to an IBM research laboratory, where he developed the idea of a relational database.

At the time databases used either a rigid hierarchy or a complex system of pointers to the physical locations of data. Such databases were efficient at handling specific queries they were set up for, but new types of queries demanded extra programming, and adding new types of data could mean a total database redesign.

The opening sentences of Codd's landmark paper highlight the problems. "Future users of large databanks must be protected from having to know how the data is organised in the machine. Users at terminals and most application programs should remain unaffected when the representation of data is changed."

The paper "A relational model of data for large shared databanks" proposed databases using simple tables containing rows and columns, with relationships between data items based on the items themselves rather than specified links or hierarchies.

Codd met some scepticism because of concerns over performance, but progress in processor and disc technology was taking off and this, plus software innovations, helped make Codd's ideas reality.

IBM announced its first relational product in 1981 and its mainstay DB2 in 1983. Meanwhile software engineers Larry Ellison, Bob Miner and Ed Oates, running their own consultancy, had been inspired by Codd's paper to go this way themselves.

Their first product, Oracle, was launched in 1979. It was the first of many products from a host of companies which turned Codd's ideas into a database industry put at more than £7bn.

Clive Sinclair (1940-present)

Clive Sinclair is hailed by several readers "for bringing computing to the masses", as Paul Bokor, IT manager at forensic investigators Hawkins & Associates, puts it.

Mick Moses, support co-ordinator at greeting card company Regent Group, says, "The ZX80, ZX81 and Spectrum may have been limited but they brought the computer into many homes for the first time. Even today you can get a Spectrum simulator for your mega-powerful PC so you can play the games."

Sinclair has had a rollercoaster career since leaving school with A-levels in physics and maths and a view that university would be a waste of time.

He started at the magazine Practical Wireless and also designed circuits, forming Sinclair Radionics in 1961. Over the next 18 years the company developed everything from pocket calculators to kit radios, handheld TV sets, digital watches and wristwatch radios, with varying degrees of success.

By the late 1970s, Apple, Commodore and Radio Shack had launched home computers costing hundreds of pounds - and in 1980 Sinclair Research launched its ZX80 at just £99.50, or £79 in kit form. It sold 20,000 in nine months at a time when home computing was virtually unheard of.

The ZX81 followed at just £69.95; well over two million were sold. This little keyboard box, with just 1Kbyte of memory, plugged into a TV for screen; programs such as games could be entered from cassette tape players. A 16Kbyte memory plug-in came later.

Other computers followed but problems with some models - and the launch of the disastrous C5 battery-powered one-seat vehicle - led to the computer business being sold in 1986 to Amstrad, which was also making a mark in home and business computing.

Since then Sinclair has got involved in microelectronics, gone back to his roots to develop a radio the size of a modern hearing aid, produced a satellite dish marketed by Sky, and persevered with electric power with the development of drive units for wheelchairs.

His record suggests he may still spring some surprises but beating the development of a £69.95 computer that introduced computing to a mass audience will take some doing.

Read more on IT jobs and recruitment