Adam Gifford: IBM's problem child still hasn't reached its potential

IBM'S US$1.25 billion ($1.75 billion) sale of its PC division to China's Lenovo may be the last and largest of the strategic stumbles Big Blue has made with the personal computer.

It sends a message the PC is mature technology, there is no more serious innovation to do, that the world's biggest computer company no longer cares about the computers most people use.

But the PC is less than a quarter-century old and, despite analyst firm International Data estimating that almost 200 million will be sold this year, it is a shambling, crippled, flawed thing.

It doesn't have to be that way.

In 1980, IBM engineer Don Estridge assembled a skunk works team - a small group who would be free to develop a product outside IBM's usual management constraints in Boca Raton, Florida, to develop a personal computer.

There were plenty of microcomputers about - Apple, Atari, Commodore, Kaypro, Osborne, Radio Shack and others - but users were in the main hobbyists and game players. When IBM entered the field, Apple ran an ad thanking it for bringing legitimacy to the area.

Estridge decided to make IBM's personal computer from off-the-shelf parts, and to make the design specification public - revolutionary concepts at the time.

He predicted the machine, which went on sale in August 1981, could sell 250,000 units over three years but this turned out to be a gross underestimate. Senior management scoffed - IBM had never made a computer that sold more than 25,000 units over that period.

Because it failed to understand what it was about to unleash, IBM fumbled the acquisition of the software needed to drive its new PC.

It ended up giving the task of writing the operating system to Bill Gates' Microsoft.

Microsoft had never written a full operating system, so it bought the rights to Seattle Computer Products' Qdos (for Quick and Dirty Operating System) for US$50,000 and tweaked it into the Microsoft Disk Operating System or MS-Dos.

Qdos was based on Gary Kildall's CP/M (Control Program for Microcomputers), the most commonly used microcomputer operating system, but Kildall missed his chance to sell to IBM.

The story goes that a meeting was not set up either because he was out gliding or because his wife baulked at IBM's insistence on a non-disclosure agreement.

Gates' masterstroke was to get IBM to allow Microsoft to retain all rights and market MS-Dos separately - the basis for his fortune.

If IBM had owned the software, it could have published the spec along with the hardware. Instead, it ceded the evolution of the PC to Microsoft and chip maker Intel.

We are all familiar with that evolution. There is a metal box containing a power supply, a fan or three, a hard disk, and something to import and export data (originally a floppy disc, now a CD or DVD player and burner).

There is a screen, which has grown bigger and thinner. There is an input device - mostly a keyboard. The commercial emergence of the graphical user interface, first in the Apple Lisa and Macintosh, led to the machine growing a tail ending at a mouse.

It also started to connect itself to other machines through standards such as Ethernet.

While Macintosh and Unix-based operating systems were built with the presumption any machine could be part of a network, the core Windows operating system had no such presumption. It was full of doors and windows that allowed outside users to walk in and take up residence.

The inherent vulnerability spawned a multi-billion-dollar industry in software to block or remove viruses, spyware and other nasties. Many users now feel the battle is unwinnable.

In a recent Salon interview, Andy Hertzfeld, one of the original Mac engineers, said personal computing was a broken business.

"There's a poison in the computer industry, and that is the fact that the common software base is controlled by a predatory software company with a lack of ethics," said Hertzfeld, who post-Apple has made significant contributions to the free and open source software movements.

"Microsoft is not a good steward of the standards."

Fortunately there are alternatives. Despite its market share retreating steadily to stabilise at between 2 and 3 per cent, Apple continued to innovate, providing benchmarks for software usability and hardware design.

The PC industry still hasn't caught up with the design breakthroughs of the iMac, and the Mac Mini makes buying that next Windows PC a decision which can't be justified on price or usability grounds. The free software industry is also hitting its straps. Distributions such as Mandrake and the new Ubuntu, and application suites such as Open Office, have made it easier for home users to throw away their Windows operating system and its uninvited cargo of malware and switch to a more secure and stable system that allows them to do pretty much everything they need.

Or does it? With the exception of Apple and some of the specialist multimedia application vendors, most of the industry has been in me-too mode. That includes Linux.

Sure, the qwerty keyboard has proved a better interface with the machine than tablets or voice recognition, but is that all there is? Why is email such a hassle? Where did I put that file?

Instead of dumping a low-margin part of its business for a bargain-basement price, IBM might have done better to set up another skunk works to take the PC to the next level.

Don Estridge can't help. He died in the crash of Delta 191 at Dallas-Fort Worth airport in 1985, after building the IBM PC division into a US$4.5 billion business.