Houghton, John --- "Computer Software Protection" [1993] JlLawInfoSci 17; (1993) 4(2) Journal of Law, Information and Science 242

• Computer Software Protection
• Abstract
• References

Computer Software Protection

by JOHN HOUGHTON^[*]

Abstract

In this article the author considers that a distinction needs to be drawn between computers and other information processing devices such as telephone exchanges, motor car ignition systems and anti-lock braking systems. He considers that the subject matter to be protected are the designs and specifications and that the protection of program code is only a means to this end.

Andrew Christie's paper has come to what I think is a most important question. Namely, what is the intellectual property; what is the subject matter. But, before I take up that question, I would like raise another about the range of protection. I will begin at the beginning. What is a computer?

The Copyright Law Review Committee (CLRC) draft report on the protection of computer software spends a number of pages defining computer program, but little attention is given to defining a computer. From my reading of the CLRC's draft report a computer is "a device having information processing capabilities" (CLRC 1993, 56). One could debate the meaning of the words 'information' and 'processing', but the point I want to make is that the class of devices with information processing capabilities is far larger than the class of devices we usually call computers.

As an IT industry watcher I follow the computer industry, and, to a lesser extent, the component industries that underpin it. I assume its fair to say that microprocessors are the most basic component of information processing devices. Now, in fact, only about one third of the microprocessors produced during 1992 were sold to computer manufactures, as components for computers. The remaining two thirds were sold into what is called the embedded devices market. Sold, that is, as components for robots, machine tools, a host of communications equipment, such as telephone exchanges, household appliances like washing machines, TVs, VCRs, etc., motor car ignition systems, anti-lock braking systems, and so on.

If the extensive protective coverage of the programs in all these devices is intended, implied and/or desired, then I think it would be better to use another name. A name that makes it clear that the coverage is more extensive than computers alone. I would suggest stored program controlled device. If such extensive protection is not intended, implied and/or desired, then a somewhat tighter definition of computer is required.

What is a computer? What's the difference between devices like VCRs and telephone exchanges, and the things we normally call computers? Well, I think it is this. A computer is a multifunctional stored program controlled device. All the other information processing devices are mono-functional stored program controlled devices. This gets us right back to the question; what is the intellectual property to be protected? It seems to me that the function of a mono-functional stored program controlled device is defined by the nature of the device. By the hardware, not the software. Whereas, the function of a multifunctional stored program controlled device (a computer) is defined by the software, not the hardware.

A telephone exchange switches telephone calls, a VCR records and plays video tapes, an anti-lock braking system monitors wheel speed and regulates brake pressure. A computer, on the other hand, can do all sorts of things. I believe that there is an important distinction between a program that controls functioning, and a program that defines the function.

The question is do we intend to treat the program code in all stored program controlled devices in the same way? Should we? Should we protect the programs in mono-functional devices, those things that are not computers, in the same way as computer programs, or are they worthy of only a lower level of protection? If fact it seems that embedded code may already be subject to a different level of protection within copyright and through possible coverage under patent and/or circuit layouts. In any event it looks to an outsider like an area of confusing anomaly.

I want to get back this functional distinction in a minute, but for the moment I would like to take another tack. So far I have dealt with the first word in the title - computer. I would now like to deconstruct the second - software. What is software?

Software is not hardware. Software is made up of at least two sub-categories - programs and data. To start with the first sub-category, a program is "a set of statements or instructions used directly or indirectly in a computer in order to bring about a certain result" (CLRC 1993, 4).

Leaving aside the meaning of the word certain, I want to examine the point of focussing on the results. It seems to me that the results of executing a program have as much to do with the information going in, the data, as they do with the instructions and statements, the program. You know the old GIGO rule - garbage in, garbage out. It may just be semantics, but I would prefer to see more emphasis on what I think is the real point. Namely, as I have suggested, the function.

A program is "a set of statements or instructions..." What is a set? How can one possibly define the bounds of a program on this basis? What sense can one possibly attach to substantially copy, if there is no way to know how many instructions or statements make up a set. Or, for that matter, what the categorising criteria are for establishing the in-the-set versus not-in-the-set boundary?

Again, I would suggest that focusing on the function might help. If a statement or instruction contributes, and is an elemental part of the process of performing a specific function, then it is part of the set. If not, it is not. The same test could be applied to data, and to any other element.

Such a functional definition might require some specific inclusions, for program locks for example, but that should not present any major problems. It would, however, be necessary to think about the levels of abstraction problem, and internal versus external functions.

Taking the latter first, there is an obvious difference between functions integral to being a computer - input/output, addressing, device handling, etc. - and the business function that the creator of an applications program is automating. This is the distinction between operating systems software and applications software. Again, there should be no major problems with a functional perspective. The levels of abstraction problem is much more difficult, and I would like to approach it gradually. I want to focus on the basic question; what is the intellectual property?

What do programmers and systems analysts do? And what does this tell us about what software is? Well, typically, a systems analyst analyses the operations of a department, or one of the functional divisions of an organisation. A purchasing department for example. The department's function, purchasing, involves a number of interrelated processes, and interacts with other parts of the organisation in a number of specific ways. Put simplistically, the point of systems analysis is to (re)describe the functioning of the department, the business process. Hence, systems specifications are, in effect, a business process or function design.

Program specifications are a more detailed version of this design. They expand on the steps necessary to computerise the business process or function. Source code program statements and instructions are a more detailed version again. They expand on the steps necessary to get a machine that, after all can only count up to one, to perform the business process or function. The object code derived from this source code is a more detailed version still. It expands on the steps necessary to get a particular kind of computer to perform the business process or function. And so on.

My point is that the continuum source code-object code-machine code, upon which the software protection debate is focused, is just the tail end. The creativity, the intellectual property, is higher up.

In his book, Software Perspectives, Peter Freeman proposes a systemic conceptualisation of software that fits well with the functional focus I am suggesting. To quote Freeman: "Software provides the active portion of the system, the part that makes the system seem alive. While it is obvious that the underlying hardware is an essential ingredient of any computer system,... the nature of the general purpose computer is precisely that it can do anything (and thus nothing) that a particular set of software instructs it to do. One reason that it is unproductive to think of software as consisting only of programs, is that this usually connotes only the procedural part of the programs, leaving out the all important data... It is useful to focus on the fact, that the software that makes up the applications functions is composed of both data and functions. All the information produced during development is important. Successful software development is directly tied to how well one deals with all the information present,... much of which is something other than directly executable programs or their embedded data. Hence," Freeman writes, "I am led to the conceptual generalisation that software is all the information produced during development" (Freeman 1987, 7-8).

While they are not computer programs, I think that systems designs, systems and program specifications, and so on, are computer software. The creativity, the intellectual property, is in the designs and specifications. It is these that turn the general purpose information processing device, the multifunctional stored program controlled device, the computer, into a functional, useful machine. It is these that define the function, and it is these we should be protecting. The protection of programs and data should be seen as no more than the means to this end. Not as an end in itself.

The problem that remains is the one of the levels of abstraction. If we venture up the continuum of increasingly generalised (re)descriptions of functions in pursuit of the intellectual property, where do we stop? Perhaps Freeman's wording "involved in the development process" is enough, but I doubt it. Clearly the idea-expression dichotomy is more nearly a continuum. The question is, how far can we push copyright protection towards the idea end of it before it breaks down?

So, to summarise, I think there is an important distinction between computers and mono-functional stored program controlled devices, and it might be worth considering different software protection for these categories. I think it is the designs and specifications that are the real intellectual property, the subject matter to be protected. The protection of program code is just a means to this end, and I suspect we will only get it right if we keep this in mind.

References

Copyright Law Review Committee 1993 Draft Report on Computer Software Protection, CLRC, Canberra, June 1993.

Freeman, P. 1987 Software Perspectives; The System is the Message, Addison-Wesley, California.

---

^[*] Centre for International Research on Communication and Information Technologies.