'Power to the people" may not arouse much passion as a political slogan, but 25 years ago, the idea of "computer power to the people" helped start a revolution. Rather than being controlled by white-coated technocrats working for governments and giant corporations, computers should be available to everyone - which implied they must also be affordable and easy to use.
It also implied that, if there was a market for market for personal computers, some people could have lots of fun and become very rich supplying them. Thousands of companies were founded to sell the new types of hardware and software the new industry required. Most promised far more than they could deliver, which is not uncommon with revolutions. Even so, it is impossible not to be impressed by the personal computer's growth in power and performance, and its market penetration.
Today, PCs can be found in more than 50% of American homes - a far bigger proportion than any of the pioneers seriously believed. Where California's Silicon Valley has led the way, the rest of the world has followed.
Intel made the revolution possible by designing the first microprocessor - a "computer on a chip" - in 1971, but it really started in Albuquerque, New Mexico. That is where a small company called Micro Instrumentation and Telemetry Systems, MITS, run by Ed Roberts, made backyard model rockets and pocket calculators. MITS probably did not develop the first personal computer, but it put together the first hobbyist construction kit to capture the American imagination. It also came up with a catchy name, Altair 8800, derived from an episode of Star Trek.
The MITS Altair helped set standards that encouraged the commercial use of personal computers, and some of them are still visible today. MITS used two products: an Intel 8080 chip, the ancestor of today's Pentium, and Digital Research's CP/M operating system, which could be used today by anyone familiar with MS DOS (Microsoft Disc Operating System).
The Altair also ran a version of Basic (Beginners all-purpose symbolic instruction code) written by a teenager called Bill Gates and his friend Paul Allen. They were inspired to write it after seeing the Altair featured on the cover of Popular Electronics magazine. They moved to the Sundowner Motel in Albuquerque to be close to their only customer, with Gates dropping out of Harvard and Allen taking a job at MITS. There, in 1975, they co-founded Microsoft.
MITS did not last long, but the Altair inspired dozens of companies to develop similar CP/M machines, most of which could share parts. These came to dominate the small business market for personal computers, but were not really suitable for domestic use.
However, dozens more companies targeted home users with less expandable but better integrated machines. The main ones were the Apple II, Commodore PET (Personal Electronic Transactor) and Tandy TRS-80, all launched in 1977.
In 1978, Science of Cambridge (later Sinclair Computers) launched the MK14 microcomputer kit for £39.95 plus VAT, then the ZX80, ZX81, Spectrum and QL. In 1979, another Cambridge company, Acorn, launched the System 1 microcomputer kit for £65 plus VAT, followed by the Atom, the Proton, the Electron, and the Archimedes. Actually, what would have been the Proton became much better known as the BBC Micro, announced in 1982.
The BBC's adoption of the Acorn as the basis for a government-backed computer literacy programme shows what a grip the personal computer had already gained over the popular imagination.
Unfortunately for Acorn, and dozens of similar start-ups, the future of the computer industry had already been decided. The microcomputer business had grown too big for IBM to ignore, and on August 18, 1981, it launched the IBM Personal Computer.
IBM was by far the biggest computer company ever (with $87.5bn in annual sales, it still is) and part of its job was to set standards. That is what it did with the IBM PC. But for whatever reasons, including earlier failures, "time to market" and pressure from a long-running US government anti-trust case, IBM did not take its usual line of shipping a box with an IBM processor, an IBM operating system, and IBM-everything else. Instead, it based its PC on an Intel 8088 processor and Microsoft's MS-DOS operating system. It was, it turned out, handing its 40-year monopoly of corporate data processing to a couple of upstarts.
In 1980, IBM had a turnover of $26bn and annual profits of $3.6bn, so perhaps it felt no need to fear a tiny company like Microsoft, which had only 40 staff. And while IBM realised that Microsoft and Intel would sell their stuff to all-comers, it perhaps didn't imagine that small, unknown companies would make better, faster IBM PC-compatibles and outsell the mighty IBM. But start-ups such as Compaq and Dell were soon on the way to doing exactly that.
By setting a standard, IBM did the industry a service. Most personal computers produced up to that point had been incompatible with one another, even if they used the same chip. The Acorn BBC B, Apple II, Commodore 64 Atari 800,and others all used the MOS Technology 6502 processor, but software houses had to write a new version of their program for each machine. Given the cost and the time required, they often didn't bother, leaving these proprietary machines struggling for applications. By contrast, any PC program would run on machines from hundreds (later, thousands) of manufacturers. This helped propel the growing market for PCs and PC clones.
The Altair 8800 and IBM's PC and PC AT (Advanced Technology) computers also led to a fundamental change in the way the industry was organised. Until then, the fashion had been for "vertical integration", with each manufacturer producing as much as possible in house. This could provide a better integrated system, but also created "lock-in": customers had few alternatives and could not easily "defect" to a rival supplier.
Many early microcomputer firms, such as Apple and Commodore, followed this proprietary line.But they could not sustain the huge research and development required, and all struggled, were taken over, or went bust.
By constrast, the PC market of the 1980s was based on "horizontal integration". Whatever PC you bought, it had an x86-compatible processor supplied by Intel, AMD, IBM, NEC or whoever, a version of DOS from Microsoft, IBM or Digital Research, and the manufacturer's choice of graphics card, hard drive, keyboard, memory, and screen. Choosing the "best of breed" created intense competition between parts suppliers, which led to rapid innovation, increased capabilities, and (usually) lower prices.
Microsoft was in the thick of the competition in two areas. First, there was the need to replace DOS with a more powerful multi-tasking operating system (one that could do more than one thing at once). Second, there was the long-term aim of introducing a mouse-driven graphical user interface, to make computers easier to use. This had been part of the dream since 1968, when Doug Engelbart of the Stanford Research Institute demonstrated the first mouse, and it continued in the 70s at Xerox's Palo Alto Research Centre (Parc). Xerox's work provided the inspiration for Apple's Lisa and Macintosh computers, Microsoft's Windows, and Digital Research's GEM (Graphical Environment Manager) interface, which also appeared on the Atari ST home computer.
In the 1970s, Xerox Parc was a decade ahead of the rest of the world, with its graphical interfaces and networks of personal computers connected to laser printers, which it sometimes showed off to visitors such as Apple co-founder Steve Jobs. But while it did launch some products, such as the Xerox Star 8010 in May 1981, they were expensive and never achieved commercial success. Instead of making Xerox rich, Parc's ideas enriched the industry, as firms such as Apple and Microsoft hired away their staff.
At first, Microsoft had seen AT&T's Unix as the obvious replacement for DOS, and announced its own version, Xenix. But while powerful, Unix was notoriously hard to use, and required more power than most PC users could afford. It also came in too many incompatible "flavours", of which Xenix was merely one. In 1987, Microsoft and IBM combined to launch OS/2 as a DOS replacement. However, since it didn't control the market and couldn't control IBM, Microsoft also wanted to hedge its bets. It didn't abandon Windows, and it opportunistically hired a group of defectors from Digital Equipment to write another DOS/Unix replacement, which appeared as Windows NT (New Technology).
Some saw Microsoft's actions as treachery. Others recognised that IBM was trying to wrest back its control of the industry from Microsoft, Intel, Compaq and the rest. This plan revolved around OS/2 (later transferred to IBM's sole ownership), a proprietary MCA (Micro Channel Architecture) expansion bus, and Systems Application Architecture - a schema designed to tie IBM's Personal Systems (PS/2) to large IBM minicomputer and mainframe systems. Compaq formed a "Gang of Nine" to support a plausible alternative future, and IBM's plan failed.
Next, IBM formed an alliance with Apple, which adopted IBM's PowerPC chip. They were still the two biggest personal computer suppliers in the US, and thought they could take on Intel. They were wrong, and later, Apple bottled out of the plan to move to a common hardware reference platform (CHRP) agreed between IBM, Apple, and Motorola. IBM then abandoned the launch of its own Power Personal Systems.
The balance of power had already tipped Wintel's way in 1990, when Microsoft shipped Windows 3.0. Unlike OS/2, it was cheap, and easy to adopt. Users could even buy graphical applications such as Microsoft Word for Windows and the Excel spreadsheet (both first developed for the Apple Macintosh) and Corel Draw.
At the time, Microsoft's turnover was only $1bn a year while IBM's was $67bn, so Big Blue was not about to admit defeat. It pumped vast amounts of cash into OS/2, promoting it as "a better Windows than Windows", until the stunning success of Microsoft Windows 95 sealed its fate. From that point on, Microsoft was clearly the computer industry's top dog, and sales of Windows-based PCs exploded. Microsoft's turnover grew from $1bn to $23bn during the decade, despite the continuous attentions of the Federal Trade Commission and the US Justice Department.
Their investigations culminated in a trial in which Microsoft's business practices were comprehensively damned, leading Justice Penfied Jackson to order the break-up of the company - a decision now pending appeal.
What is interesting is that, during the 1990s, despite its monopoly power, Microsoft was no more able to replace the DOS-Windows combination than anybody else.
The Windows NT operating system, introduced in 1993, was supposed to do it. NT was designed to run on a range of different processors (Intel, Mips, DEC Alpha, PowerPC), to use more than one processor at once, and to offer the reliability and stability of rivals such as OS/2 and Unix. Despite various inducements, such as letting companies install NT for the same price as Windows 95, sales struggled. Three times, Microsoft has announced that the latest version of DOS-Windows (Windows 95, Windows 98, Windows Me) would be the last of the line, giving users little choice about moving to NT (now called Windows 2000). So far, it has not made it stick.
Apple Macintosh users have been equally conservative, declining to move to technically superior systems such as Steve Jobs's NeXT or Be's BeOS, launched on the dual-processor BeBox. It remains to be seen if Jobs, the Apple co-founder who retuned to the company at the end of 1977, can shift them to an updated version of NeXT's Unix-based operating system, Mac OS X (see page 12 for a review of this system).
That's the problem with revolutions: they may be exciting, but most users prefer evolution. They don't what to risk losing what they have in some great leap into the unknown. The result may be boring, but it is a sign of the personal computer's success that what started as a revolution has become the status quo.
Which is not to say that progress has stopped. It has, rather, moved to handheld devices such as electronic organisers and mobile phones, to sealed-box "appliances" such as WebTV sets and games consoles, and most importantly, to the internet. And as the growth of the personal computer market has slowed, many PC companies - including Intel and Microsoft - have invested heavily in these new areas in search of the Next Big Thing.
For a couple of decades, Microsoft's mission statement summed up the ambition of the PC industry: to put a computer on every desk and in every home.
That has been changed to take in every type of device used anywhere at any time. The only argument is about whether we are entering the "Post PC" era, as some pundits claim, or the "PC Plus" era of Microsoft's new .Net (dot.net) vision, where the net, rather than the PC, is the software platform.
Whatever happens, I expect companies raised in an industry based on rapid innovation and cut-throat competition will adapt and survive. Old and slow minicomputer and mainframe companies were generally unable to live with PC-based competitors, and old and slow consumer electronics and telecoms companies may be in for a big surprise.
Twenty five years of personal computing
1975
MITS Altair 8800
Microsoft Basic
1977
Apple II
Commodore PET
Tandy TRS-80
1980
Sinclair ZX80
1981
IBM Personal Computer
Microsoft MS-DOS
1982
Acorn BBC B
Commodore 64
1984
Apple Macintosh
IBM PC AT
Atari ST
Commodore Amiga
1987
IBM PS/2 and OS/2
1988
NeXT Cube
1990
Microsoft Windows 3
1991
Apple-IBM deal
1993
Intel Pentium
Microsoft Windows NT
1994
Apple Power Macs
IBM OS/2 Warp 3
1995
Microsoft Windows 95
1997
Apple buys NeXT
2000
Microsoft.net announced
Apple OS X
