If you were about to launch a new consumer electronics device, would you give away the plans so that anyone - anyone at all - could make a copy of it and sell it for themselves?
Furthermore, would you make the device you were going to sell out of off-the-shelf components, so that anyone else could buy and use exactly the same component in building their copies of your product?
The IBM PC
This is exactly what IBM did when they launched the IBM PC. The hardware design was made public, and utilised commodity components that were readily available.
Third party software developers were encouraged to write software for the new platform. The platform, that combination of hardware and software tools that provided a sound basis on which other people could build added value products and and service, was open. Almost.
Gluing together the hardware and the software and was a specialised piece of software know as the BIOS - the Basic Input-Output System. If you’re familiar with that philosophy of mind conundrum known as the mind-body problem, you might consider the BIOS to be play a similar role to the mysterious bit in the middle that connects body and mind, or in this case, hardware and software.
A philosophical sidethought
Serious philosophers might take issue with the previous claim, as might technologists who very well know how to convert things between analog and digital worlds - and so might you? To start exploring whether this comparison is a valid one, why not listen to the Philosophy Bites interview with Tim Crane on Mind and Body? or visit this LearningSpace unit Crossing the boundary - analogue universe, digital worlds.
In order to keep control over the platform, whilst still apparently giving most of it away, IBM sought to keep control over the BIOS. If you wanted to manufacture your own PCs according to IBM’s design, you could have the hardware and the software designs for free, but the glue that would allow them to work together? That, you’d have to pay for.
To protect the BIOS further, IBM then played openness at its own game. They published the specification and the code that implemented the BIOS where everybody could read it, as part of the PC repair manual.
So what was to stop every other manufacturer in the world just copying the BIOS too? The answer lies in copyright. As far as the really open parts of the IBM specification went, IBM provided an open license to third party manufacturers so they could use the designs freely.
But no such license was attached to the BIOS. And by making the BIOS public, IBM thought they’d be able to protect their ownership of the BIOS. That is, they would have the right to either prevent or charge people making use of it, or other programmes that did the same thing, through being able to claim that anyone who did produce a compatible BIOS must have been influenced by exposure to IBM’s BIOS, and the intellectual property it contained.
It wasn’t to be though. As Van Lindberg describes in Intellectual Property and Open Source: A Practical Guide to Protecting Code:
Two companies in particular were able to reverse-engineer the PC BIOS: Compaq and Phoenix. Compaq was first, hiring a team of 'virgin' engineers who could prove that they had never been exposed to IBM’s copyrighted code. Compaq set up a clean-room reverse-engineering team, with the 'dirty' team creating a specification from IBM’s code and the 'clean' team creating new code to implement the specification.
However, Compaq decided to keep its BIOS as a trade secret - it was a competitive advantage over other computer makers - so another company had to duplicate the work and create a second IBM-compatible implementation.
The second company to successfully clone the PC BIOS was Phooenix. Like Compaq, Phoenix was careful to recruit engineers that had never seen or worked on the 8088/8086 chips used in the IBM PC. That alone might not have stopped a lawsuit by IBM, but Phoenix also took the unusual step of getting a legal insurance policy through Lloyd’s of London.
This legal insurance policy gave Phoenix the legal resources of a company many times its size - enough to fight and survive through a long legal battle. Ultimately, IBM was unwilling to fight a company armed with both a reverse-engineered BIOS and an insurance-funded war chest. Phoenix decided to sell copies of its BIOS instead of keeping it a proprietary secret, thus creating the massive clone PC market of the 1980s.
(You can also read more about the story here: Mac In Touch Reader Reports - PC History.)
If we now turn to the Microsoft Windows operating system, this can be argued to be an open software development platform. It provides all the third party developer needs in order to be able to write - and sell - whatever programs they like to run on the platform.
However, Microsoft Windows is not in itself open software. Whilst a third party developer is free to write software that will run on Windows, they cannot make a copy of Windows, or make changes directly to it.
In contrast, an operating system like Linux does allow a developer to make copies of, distribute and/or modify the operating system largely at will.
To the extent that the Apple iPad - running the iOS operating system, can run programmes developed by third party developers, whilst at the same time not allowing those developers the freedom to tinker with the operating system itself - is as open as Microsoft Windows.
However, the installation route for software on the iPad is through an Apple controlled App Store, within which software listings are subject to approval by Apple. The iPad is thus a closed system to the extent that software developers are not allowed to distribute software that is not approved of by Apple (though they are free to develop whatever application they want as long as it can run on iOS), and iPad owners are not able to install apps on their iPad if they aren’t Apple approved.
Openness, it seems, has many levels, even if it’s only open platforms we’re talking about.