Journal of Law, Information and Science
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Journal of Law, Information and Science |
|
by ANDREW CHRISTIE[*]
The Australian Copyright Law Review Committee (‘CLRC') recently published its long-awaited Draft Report on Computer Software Protection. In this Report the CLRC sets out its initial recommendations for reform of intellectual property protection for software and data. One of the key issues addressed by the Report is the appropriate form of protection for computer programs. The CLRC concludes that computer programs should continue to be protected by copyright, and in particular should continue to be treated as ‘literary works' for copyright purposes. This article reviews, and criticises, the CLRC's deliberations and proposals on this topic. It argues that the CLRC has failed to address the crucial issue of what aspect of a computer program is the most deserving of intellectual property protection. A consideration of that issue suggests that it is the design of a computer program which should be protected. Copyright is not the right regime for providing appropriate protection for that subject matter.
It is more than 15 years since the World Intellectual Property Organisation (‘WIPO’) formulated its Model Provisions on the Protection of Computer Software.[1] It is a decade since the so-called ‘temporary amendments’ were made to the Australian Copyright Act 1968 (‘Copyright Act’) to deal expressly with computer programs.[2] And it is less than a year since the Autodesk saga reached its sorry conclusion in the High Court of Australia.[3]
What do these events tell us? A sporting (Australian Rules football) metaphor may assist. In the ongoing game of intellectual property protection for computer programs, it was over 15 years ago that Australia (and indeed the rest of the industrially-developed world) failed to pick up the ball (being the ball marked ‘sui generis’). In the following decade, Australia, following the lead of the US,[4] picked a ball, but it was the wrong ball (being the ball marked ‘copyright’). As the subsequent years have demonstrated, Australia (and the rest of the world) have been running not only with the wrong ball, but in the wrong direction.
Now, with the release of the Copyright Law Review Committee’s Draft Report on Computer Software Protection (‘Draft Report’),[5] Australia has the chance to re-write the rules of the game for the next few decades. The concern of this writer is that we have not learnt from the mistakes of our past, and that we are about to continue running in the wrong direction.
To know whether one is heading in the right direction, it is helpful to identify from where one is coming and to where one has arrived. What follows is a very brief review of the history of intellectual property protection for computer programs.
In the early days of computers and programming, the people who wrote and exploited computer programs saw them perhaps somewhat differently from the way lawyers, patent attorneys and industry people see them today. Back then, computer programs were seen as adjuncts to machines. Just as in the Nineteenth Century one could take a punched card and use that to control the way a Jacquard loom weaved cloth, so in the 1950s and 1960s one could use a punched card to control the way an electronic information processing machine processed information. Not unreasonably, therefore, computer programs were seen as adjuncts to machinery.
When it came to the issue of intellectual property protection, that view naturally suggested the regime of patent protection. Accordingly, during the late 1960s and early 1970s applications were made to the Patent Offices, and court cases dealing with those applications were fought, with varying degrees of success.[6] The law on patenting computer programs is still developing. The CLRC does not propose to exclude computer programs from the scope of patent protection.[7]
By the mid-1970s, the commercial value of computer programs was increasingly recognised. Computer programs were no longer bundled with the (usually mainframe) computer being sold, they were a marketable commodity in their own right. Thus industry began to look afresh at what sort of intellectual property protection should be given to them.
In the mid-1970s, the developed countries had the opportunity to pick up the ball marked sui generis. The Latin phrase sui generis means ‘of its own kind’. A sui generis regime is a new regime, being one specially tailored to the relevant subject matter.
In fact, the sui generis approach was adopted not many years later in the US, for the related subject matter of integrated circuit design.[8] But it was not for computer programs. Why? Because copyright had become the favoured form of protection.
The reasons why copyright was favoured over a sui generis regime for protecting computer programs may be simply stated. There appear to be three contributing factors. First, when a client presents a representation of its computer program to its intellectual property law adviser, the representation is invariably in writing. To the intellectual property lawyer, and to the judges in the courts, something in writing looks like a literary work, and that suggests copyright protection.
Secondly, and more significantly, if copyright is used to protect computer programs, it is not necessary to go through the arduous task of drafting and passing new legislation. The US Semiconductor Chip Protection Act of 1984 saw the light of day only after many years of Senate and Congressional hearings, which involved listening to, and being influenced by or ignoring (as the case may be), numerous lobby groups representing the disparate interest parties.[9] If, however, the existing copyright regime is used, it is simply a matter of doing nothing. No new legislation is required.[10]
Thirdly, and most importantly, copyright provides the potential for international protection, pursuant to the Berne Convention.[11] That international treaty, to which all major countries are now a party, requires each country to provide copyright protection to literary and artistic works produced in any other country which is a party to the treaty. If computer programs could be slotted in under the heading of ‘literary work’ for copyright purposes, it was thought that international protection would automatically arise under the Berne Convention.
Given these acknowledged advantages of using copyright, one might well ask why not adopt copyright protection for computer programs? What is the problem? Has it not been the case in the past that copyright has successfully taken up the challenge of protecting the products of new technology?
It is true that in the Nineteenth and Twentieth Centuries, copyright successfully adapted to protect the products of the new inventions of the phonogram, the cinematograph film, the radio, and the television, as these became commercial entities in their own right. It is notable, however, that the products of these subject matters (the sound recording, the film, and the sound and TV broadcast) came to be protected by copyright in a part of the Copyright Act (Part IV) separate from the part (Part III) which gives protection to the traditional copyright subject matter of literary, dramatic, musical and artistic works.
It is submitted that there is a fundamental difference between the Part IV works and computer programs. The Part IV works are essentially new ways of embodying an underlying subject matter in the nature of a Part III work.[12] That is to say, a sound recording is a new embodiment of a musical work and/or a literary work; a film is a new embodiment of a dramatic work and/or a literary work; and a broadcast is a new embodiment of a musical, literary, dramatic and/or artistic work. A computer program, in contrast, is not a new way of embodying a Part III work. This is because the underlying subject matter of a computer program does not have the fundamental characteristic of a Part III work.
The characteristic which is common to and determinative of Part III works is that they are addressed to humans. In particular, they are addressed to humans so as to ‘afford either information and instruction, or pleasure’.[13] A computer program, however, is not addressed to humans. The primary purpose of a computer program is not to give information and instruction, or pleasure, to a human, but to control a machine - whether that machine may be a computer (however that may be defined), an industrial robot, or a washing machine.
It should be clear, therefore, that a computer program inherently is neither a Part III work (since it is not addressed to humans) nor a Part IV work (since it is not an embodiment of a Part III work). It should also be clear that to protect computer programs by copyright is to use a round hole for a square peg. Yet not only has Australia (as well as the US, Japan, the EC countries and many others) chosen to protect computer programs by copyright, it has chosen to protect computer programs as a Part III work - namely, as literary works.[14]
The problem with treating computer programs as literary works is that the copyright protection given to literary works is very wide and very long. For computer programs, this protection is both too wide and too long.
In relation to the width of protection, the Copyright Act provides that one cannot reproduce or make an adaptation of a literary work without the consent of the copyright owner.[15] The problem with this is that one cannot use or analyse a computer program without either reproducing it or making an adaptation of it. To run a computer program, one must load parts of it into random access memory (or some other form of memory store). It seems clear that such storage is an infringing reproduction (or adaptation). If one wishes to analyse the content and structure of a computer program, it is necessary to run it, it is necessary to print out some representation of it, and it may be necessary to ‘decompile’ it (or carry out some other translation of it). Those are all acts of reproduction (or adaptation) too. This is not the case with the traditional (ie. true) literary works, such as books. One can use (ie. read) and analyse a book without reproducing or adapting it, and hence without infringing copyright in it.
The are, however, further problems with protecting computer programs as literary works. Within the field of copyright protection for literary works there has developed the principle of ‘non-literal copying’. Under this principle, the prohibition on copying extends to producing a version of the literary work in a different form of expression. The principle of non-literal copying has been applied to computer programs, so as to find infringement of copyright where the function and appearance of a protected program was reproduced in another program written in a different computer language.[16] It is not the purpose of this article to review the developing law on non-literal infringement of computer programs.[17] It is sufficient to note that the application of the principle of non-literal infringement has generated much controversy in both academic writings and legal decisions.
There is one other key respect in which copyright protection for literary works is too wide for computer programs - namely, non-complete copying. Non-complete copying means copying (whether literal or non-literal) of less than the whole of the protected work. That is to say, it is an infringement of copyright to reproduce a ‘substantial part’ of a protected work.[18] As the decision in Autodesk v Dyason [No. 1][19] shows, a ‘substantial part’ of a computer program may be something extremely small in the context of the overall work, and indeed may be something that is not even protected in its own right as a literary work.[20] It is submitted that the decision in Autodesk v Dyason [No. 1], more than any other recent decision, demonstrates that literary work copyright protection is too wide for computer programs.
The case for the argument that literary work copyright protection is too long for computer programs can be made simply. Under the Copyright Act, copyright for a literary work lasts the life of the author plus fifty years.[21] This means that if a twenty-five year old programmer writes a computer program in 1994 and she lives until the age of sixty-five, the copyright in her program would last until the year 2084. It is doubtful that anyone could justify that duration of protection.
The above discussion should, it is submitted, demonstrate that Australia (and the other developed countries) are probably using the wrong regime (copyright), and certainly using the wrong heading within that regime (literary work), to protect computer programs. But one might be forgiven for feeling optimistic, given that the CLRC has been charged with the brief of re-writing the rules for computer program protection. So, let us consider just what it is that the CLRC is proposing.
The CLRC’s Draft Report commences by stating that the right way to protect computer programs is not through patents, or at least, not through patents exclusively. The reason given is that the patent regime is too costly and too time consuming, due to the need for the drafting and examination of specifications, and the potential for opposition proceedings and appeals against the decision of the Registrar of Patents. In addition, the CLRC believes (quite rightly) that very few computer programs would satisfy the patentability requirements of novelty and inventive step. And in any event, the CLRC says that there is very little support for patent protection for computer programs nowadays.[22]
It is suggested that what the CLRC means by this last justification is that there is very little support for patent protection exclusively. There is, of course, substantial support for patent protection as an adjunct to other protection, as the recent cases in Australia, the US and Europe demonstrate.[23]
Having made it clear that the solution does not lie with patent protection exclusively, the CLRC then goes on to suggest that the ideal form of protection would be a copyright-like sui generis regime.[24] Such a regime would be flexible, and could be adapted to the needs of the subject matter. But despite having identified the correct ball to play with, the CLRC fails to pick it up.
The reason is stated to be as follows:
The industry has become used to copyright protection. It has a familiarity with it. To depart from it entirely would be likely to create a good deal of uncertainty when, despite difficulties here and there, there is a common understanding and approach not only in Australia but in other countries as well.[25]
It is surprising, and indeed disappointing, to read that the main reason a law reform body rejects an option for reform is the preference of a familiar devil over an unfamiliar one. It is submitted, however, that the real reason the CLRC does not want to recommend a sui generis system has to do with the politics of global trade. Reading between the lines, the CLRC seems concerned that Australia would be threatened with trade sanctions from trading partners (for which read ‘the US’) if it was to break ranks on the appropriate form of protection for computer programs. Whatever may be the likelihood of that possibility, it is submitted that it is a political not legal issue, and hence is one with which a law reform body should not be predominantly concerned - or at least, is one which should not be decisive to a law reform body if the fundamental principles of good law-making point to a contrary conclusion.
Having indicated that copyright should remain the system for protecting computer programs, the CLRC proceeds to state:
that the optimum form of protection would be to remove computer programs from Part III of the Act and place it [sic] in Part IV as a separate category of subject matter.
For those, like this writer, who do not regard computer programs as true literary works, this statement is heartening. Yet hopes for Part IV protection are shockingly dashed in the following paragraph, when the CLRC concludes that Australia’s obligations under the Berne Convention preclude this course of action.[26] Put simply, the CLRC takes the view that the Berne Convention requires Australia to protect computer programs (at least those in source code) as literary works.
It is difficult to see why the CLRC takes this view, particularly when that view is not shared by two of the authorities to whom it turned for advice on this issue. The CLRC quotes extensively from Professor Ricketson’s leading works on intellectual property[27] and on the Berne Convention.[28] Professor Ricketson is of the opinion that the Berne Convention does not require member countries to treat computer programs as literary works.[29] The Office of International Law in the Attorney-General’s Department is of the same opinion.[30] Nevertheless, the CLRC is confident to conclude to the contrary. On what basis?
The CLRC states:
Whilst the Committee is grateful for the assistance given it by the Office of International Law, it is persuaded by the decisions of the High Court in the Apple case and the Autodesk case that programs expressed in source code are true literary works.
This writer is certainly not persuaded by those decisions. One does not have to be an international lawyer to realise that the Apple case[31] and the Autodesk case[32] are not cases interpreting the Berne Convention. Rather, they are cases interpreting the Australian Copyright Act. Decisions on Australian legislation cannot be decisive of the position which obtains under an international Treaty. This is especially so when the international Treaty in question was drafted almost a century before computer programs became a gleam in the eye of copyright lawyers.[33]
The Copyright Act currently defines ‘computer program’ to mean:
an expression, in any language, code or notation, of a set of instructions (whether with or without related information) intended, either directly or after either or both of the following:
(a) conversion to another language, code or notation;
(b) reproduction in a different material form;
to cause a device having digital information processing capabilities to perform a particular function.[34]
This rather long-winded definition may be condensed to the following set of key elements:
- An expression
- in any language, code or notation
- of a set of instructions ...
- intended to cause a digital information processing device to perform a particular function.
It is noteworthy that the definition refers to ‘an expression’. Under the traditional view of copyright protection for literary works, copyright protects only the expression of a work, not the work’s idea. The correctness of this view is open to challenge on a number of grounds,[35] especially when applied to computer programs.[36] This is not the forum in which to debate this important and contentious issue. It is sufficient to note that the words ‘an expression’ in the definition presumably are intended to identify exactly what aspect of a computer program is protected by copyright.
In Autodesk v Dyason [No. 1], Mason CJ, Brennan and Deane JJ add some ‘supplementary comments’ to the main judgment of Dawson J (with whom they agreed).[37] Those somewhat mystifying comments appear to be directed to the question of what is the significance of the words ‘an expression’ in the current Australian definition. It is submitted that, for all intents and purposes, Mason CJ, Brennan and Deane JJ conclude that the words ‘an expression’ can be read out of the legislation.
The current Australian definition of ‘computer program’ contains two major limitations, both of which relate to how the computer program must operate. First, to come within the definition, the program must cause a digital information processing device to perform a particular function. This excludes programs for analogue computers. Admittedly analogue computers are not particularly common, but they do exist and may continue to exist in the future - and thus there is no good reason why intellectual property protection should not extend to them. The CLRC acknowledges this, and accordingly proposes to delete this limitation.[38]
The second limitation on the current definition is that the program must consist of a set of instructions. The possible problem is that this limitation may exclude programs written in the newer forms of programming languages, and in particular programs written in declarative languages.[39] The CLRC recognises the potential of this limitation to cause a problem in the future.[40] For these two reasons, the CLRC proposes that a new definition of computer program be introduced.
One option the CLRC considered was to have no definition of computer program. This is the approach adopted in the relatively new UK copyright legislation.[41] The CLRC recognises that this approach allows for greater flexibility, but rejects it on the ground that it ‘may create uncertainty’.[42]
It is submitted that the greater uncertainty lies not in having no definition, but in having a rigid definition which may not cover future developments (since a number of these are bound to be developments that are presently unforeseeable). In this writer’s view, the preferable approach is to have no definition of computer program, and to leave it to the courts to decide over time whether new developments should be regarded as protectable subject matter.
The CLRC’s preferred approach is to adopt the definition of computer program which appears is the US legislation, which provides:
A computer program is a set of statements or instructions to be used directly or indirectly in a computer to bring about a certain result.
If one accepts that the term should in fact be defined, this definition is rather nifty. Its key elements may be summarised as follows:
- A set of statements or instructions
- to be used in a computer to bring about a certain result.
Put simply, the US approach is to protect a set of statements or a set of instructions. This has a number of advantages over the current Australian definition. First, the definition is not limited to a set of statements or instructions embodied in a given expression. This approach is consistent with this writer’s opinion that the words ‘an expression’ in the current Australian definition are superfluous (indeed misleading), and with this writer’s view as to what Mason CJ, Brennan and Deane JJ intended by their supplementary comments in Autodesk v Dyason [No. 1]. The absence of similar words in the US definition is desirable, since it makes clear that which it is submitted is the true position under copyright law - namely, that the subject matter of protection is not the expression of the instructions (or statements) but the instructions (or statements) themselves, whatever their expression.
The second advantage of the US definition is that there is no reference to ‘language, code or notation’. This is desirable, because it avoids the possible argument that these words preclude protection for a computer program in expressed in microcode (since it is debateable whether microcode is any of a language, a code or a notation).
The third advantage is that there is no reference to ‘digital’. Thus, the US definition does not exclude programs for analogue computers.
Generally, the CLRC’s proposal to adopt the US definition of computer program is reasonable, subject to one reservation. Nowhere in the current Copyright Act is ‘computer’ defined. The CLRC does not make any proposal for defining this key term. That does not appear to be a conscious decision, justified by the desire to accommodate possibly unforeseeable technological developments. Rather, it appears to be an oversight.
More critically, however, it is submitted that the CLRC’s proposal fails to address properly the fundamental issue of precisely what is the intellectual property in a computer program which should be protected. Until that issue is addressed and answered, it is not possible to determine what is the appropriate form of protection for computer programs.
The CLRC says that any new definition of computer program must ‘clearly and unambiguously apply to programs expressed in source code, programs expressed in object code (compiled from source code) and to microcode’.[43] The sense of that objective is not disputed. What is disputed is the implicit view that source code, object and microcode are the only embodiments of computer programs which need to be protected.
Figure 1 is a diagrammatic representation of the various levels in the computing hierarchy.
At Levels 2 to 5 in that hierarchy, software[44] is represented in the forms which the CLRC considers must be protected. It is now clear that Australian courts regard programs in both source code (Levels 4 and 5) and object code (Level 3) as within the definition of ‘computer program’ in the Copyright Act.[45] It is submitted that the position with respect to microcode (Level 2) is somewhat more contentious. There is no case on this issue in Australia. There is, however, authority in the US to the effect that programs in microcode are computer programs for the purpose of the US copyright legislation.[46] This writer has doubts as to whether an Australian court would follow the reasoning and decision in that case. For the purposes of this article, however, it will be assumed that microcode constitutes a computer program under the Copyright Act.
Figure 1 - The Computing Hierarchy
Level 1 is hard-wired logic. Software may be represented at this level by the interconnections made between various logic components in a solid state item. An application-specific integrated circuit (ASIC), being a dedicated (one purpose) processor chip, is the classic example.[47] It is arguable that software represented at this level is a ‘computer program’ for the purposes of the Copyright Act. The matter is far from certain, however, as the only case which appears to have considered the issue, the Autodesk case, contains conflicting decisions on this point.
In the Autodesk case, there arose the question whether the ‘AutoCAD lock’ was a computer program. That device essentially consisted of a clock, a shift register and an exclusive OR gate. In the Federal Court at first instance, Northrop J held that the ‘AutoCAD lock’ was a computer program for copyright purposes.[48] The Full Court of the Federal Court on appeal took a contrary view.[49] On further appeal, the High Court did not decide the issue. Dawson J did, however, express doubt that the AutoCAD lock was itself a computer program protected by copyright.[50]
It is submitted that software at Level 1 should be treated as a ‘computer program’ for copyright purposes, if copyright is to be the regime for providing intellectual property protection for computer programs. The justification is that whatever can be done in software can be done in hardware, and vice versa. That is to say, a particular set of logical manipulations may be performed on data either by a general purpose processor programmed by a ‘computer program’ or by an ASIC (or some other form of hard-wired device). In both cases, the intellectual property potentially worthy of protection is the set of logical manipulations to be performed on the data. If the embodiment of that logic in the form of a ‘computer program’ (ie. as represented at Levels 2, 3, 4 or 5) may be protected by copyright, should not the embodiment of that logic in the form of an ASIC also be capable of protection by copyright? It is submitted that there is no reason in logic or policy why it should not.
It is further submitted that at least one major piece of copyright legislation expressly provides protection for hard-wired logic. Recital 7 of the EC Council Directive of 14 May 1991 on the Legal Protection of Computer Programs[51] (‘EC Directive’) provides:
Whereas, for the purposes of this Directive, the term ‘computer program’ shall include programs in any form, including those which are incorporated into hardware; whereas this term also includes preparatory design work leading to the development of a computer program provided that the nature of the preparatory design work is such that a computer program can result from it at a later stage.
The first part of this definition seems to make clear that logic embodied in hardware (ie Level 1) may constitute a computer program for copyright purposes. In what appears to be a case of gross oversight, the CLRC does not consider the EC definition in its discussion of the appropriate definition of computer program. At paragraph 6.14 of the Draft Report the CLRC states that ‘the approach taken by the European Community has been not to define computer program’. That statement is simply incorrect.
The conclusion that software at Level 1 should be capable of copyright protection leads to a more significant question - namely, should software represented at Level 6 be capable of protection by copyright? At this level, the set of logical manipulations to be performed on data are represented in the most abstract or conceptual of forms, in that the representations are not limited by the constraints of a computer language. At Level 6, the intellectual property may be take the forms of functional specifications, flow charts and algorithms.
Level 6 is the level of design. That is to say, the intellectual property in the embodiments of software at Level 6 is the design of the computer program. It is submitted that the design is the most important aspect of a computer program. It is here that the real creativity, and innovation (if any), will lie. It is at this level that the greater part of the resources of the creator will be spent. The coding of software (ie. turning the Level 6 design into a computer program represented at one of Levels 2 to 5) is a relatively routine process.
These facts suggest that the aspect of a computer program most deserving of intellectual property protection is its design. It is therefore submitted that the representation of a computer program at Level 6 in the computing hierarchy should receive intellectual property protection.
Does copyright provide adequate protection for computer program design? The answer is unclear. It must be doubted whether the current definition of ‘computer program’ in the Copyright Act includes a representation of a program at Level 6 in the hierarchy. The current definition focuses on an expression of a set of instructions intended to cause a digital information processing device to perform a particular function. Functional specifications and other representations of logic at Level 6 are probably too abstract to come within that definition. The US definition of ‘computer program’, which the CLRC recommends Australia adopt, likewise probably does not extend to a Level 6 representation of software.
This should be contrasted with the definition used in the EC Directive. Recital 7 expressly states that the term ‘computer program’ includes ‘preparatory design work’ leading to the development of a computer program. That seems to make it clear that a Level 6 representation of software is a computer program for copyright purposes.
The fact that the current (and indeed the proposed) definition of computer program in the Australian copyright legislation probably does not include a representation at Level 6 does not, of course, mean that there is no copyright protection for the design of a computer program. This is because the principle of non-literal copying means there may be infringement of copyright in the source code or object code of a program where another program has adopted the same, or a substantially similar, design. It was noted earlier in this article that the principle of non-literal copying has generated much controversy in relation to computer programs. When dealing with the issue of infringement by reproduction of the computer program’s design, that principle is in conflict with the idea/expression principle. How this conflict ultimately will be resolved is far from clear. What is clear, however, is that the courts are struggling in their attempts to resolve the conflict and produce an acceptable solution.
In this writer’s view, the fact that the courts are struggling to resolve the conflict between the principle of non-literal copying and the idea/expression principle is indicative of a greater problem - namely, that copyright is not the right regime under which to protect computer programs. But what is the right regime? The fact that the most important and valuable aspect of a computer program is its design suggests that the most appropriate form of intellectual property protection might be some type of design right.
This possibility may sound rather exotic. It is not, however, totally without support. At least one other writer has, very recently, suggested something similar.[52] This is neither the time nor the place in which to explore this possibility in detail. This is the time and place, however, to note forcibly that the CLRC has not even mentioned this possibility in its deliberations. Given that the reform ultimately implemented in this of this area is likely to be on the statute books for many decades, one would hope that the CLRC would at least consider the question of the form of protection in the wider context of exactly what is the intellectual property that is deserving of protection. Its failure to do this is very disappointing.
This writer does not claim to have all the answers. Indeed, he does not even claim to have asked all of the right questions. He does, however, claim to be able to make a small case for the argument that we are not yet far enough down the track in our understanding of what are computer programs, of what aspects of them deserve to be protected, and of how we should protect those aspects, for the CLRC (or indeed anyone else) to claim to be able re-write the rules of the game for the coming thirty years.
[*] B.Sc., LL.B. (Hons) (Melb), LL.M. (Lond), Ph.D (Cantab). Senior Lecturer in Law at the University of Melbourne.
[1] The Model Provisions are contained in (1977) 16 Industrial Property 265.
[2] Copyright Amendment Act 1984.
[3] Autodesk v Dyason [No. 2] [1993] HCA 6; (1993) 176 CLR 300.
[4] See in particular the US Final Report of the National Commission on New Technological Uses of Copyright Works (‘CONTU Report’), 31 July 1978.
[5] Copyright Law Review Committee, Draft Report on Computer Software Protection, June 1993 (ISBN 0 642 19361 4).
[6] The early US cases include In re McIlroy 442 F.2d 1397 (1971) and Gottschalk v Benson [1972] USSC 223; 409 US 63 (1972), In the UK, the issue received judicial consideration in Slee & Harris’s Application (1966) RPC 195, and Burroughs Corporation (Perkins’) Application [1974] RPC 147. An early Australian case was NV Philips Gloeilampenfabriken [1966] 36 AOJP 1020.
[7] See Draft Report, paragraph 4.22.
[8] Semiconductor Chip Protection Act 1984, 17 U.S.C. §901.
[9] For a brief history of the passage of that Act, see Stern, Semiconductor Chip Protection (1986, Law & Business/Harcourt Brace Jovanovich, New York), §1.4.
[10] This was demonstrated by the decision of the High Court of Australia in Computer Edge v Apple Computer [1986] HCA 19; (1986) 161 CLR 171, where it was held that a computer program in source code was protected as a literary work under the Copyright Act prior to the 1984 amendments.
[11] International Convention for the Protection of Literary and Artistic Works, signed at Berne on 9 September 1886 (Paris revision 1971).
[12] It is not necessarily the case that the underlying subject matter is a Part III work protected by copyright, since the underlying subject matter may not itself satisfy the various requirements for protection.
[13] Hollinrake v Truswell [1894] UKLawRpCh 158; [1894] 3 Ch 420, at 428. See also Exxon Corporation v Exxon Insurance [1982] Ch 119, at 143-144; and Computer Edge v Apple Computer [1986] HCA 19; (1986) 161 CLR 171, at 182, 192 and 201. These words were used in these cases to define a literary work. It is submitted that the characteristic of being intended to provide information and instruction, or pleasure, to humans is common to, and determinative of, dramatic, musical and artistic works also.
[14] Copyright Act, section 10(1) (definition of ‘literary work’).
[15] Copyright Act, section 31(1)(a).
[16] See, for example, Whelan Associates v Jaslow Dental Laboratory [1986] USCA3 976; 797 F.2d 1222 (1986).
[17] A discussion of some of the later US cases is contained in Fitzgerald, ‘Square Pegs and Round Holes: Recent Developments in Copyright Protection for Computer Software’ (1993) 4 Journal of Law and Information Science 142. See also the paper by Hunter, ‘Limitations on the Proposals for Reform of Non-literal Copyright and Reverse Engineering of Computer Software’, in this volume.
[18] Copyright Act, section 14(1)(a), when read with sections 31(1)(a) and 36(1).
[19] [1992] HCA 2; (1992) 173 CLR 330.
[20] In the Autodesk [No. 1] case, it was held that a 127-bit look-up table was a ‘substantial part’ of the protected computer program, such that the reproduction of that look-up table by storage of its contents in an EPROM was an infringement of copyright in the protected program. Dawson J, with whom the rest of the court agreed, stated that ‘no reliance was placed upon [the look-up table] as a literary work in itself’: [1992] HCA 2; (1992) 173 CLR 330, 347.
[21] Copyright Act, section 33(2).
[22] Draft Report, paragraphs 4.21 and 4.22.
[23] For recent developments in Australia, see IBM v Commissioner of Patents (1991) 22 IPR 407, and the discussion of this in Old, ‘Patenting Computer-related Inventions in Australia’ (1993) 24 International Review of Industrial Property and Copyright 345. The position in the US was the subject of a recent article in this journal: Swinson, ‘Software Patents in the United States’, (1993) 4 Journal of Law and Information Science 116.
[24] Draft Report, paragraphs 4.23 and 4.24.
[25] Draft Report, paragraph 4.25.
[26] Draft Report, paragraph 4.26.
[27] Ricketson, The Law of Intellectual Property (1984, Law Book Company, Sydney).
[28] Ricketson, The Berne Convention for the Protection of Literary and Artistic Works: 1886-1986 (1987, Queen Mary College/Kluwer, London).
[29] Ibid., 900. Many other leading academic commentators are of the same view. See, for example, Cornish, ‘Computer Programs and the Berne Convention’ [1990] 4 European Intellectual Property Review 129, at 131; and Soltysinski, ‘Protection of Computer Programs: Comparative and International Aspects’ (1990) 21 International Review of Industrial Property and Copyright 1, at 25-27.
[30] Draft Report, paragraph 4.14.
[31] Computer Edge v Apple Computer [1986] HCA 19; (1986) 161 CLR 171.
[32] Autodesk v Dyason [No. 1] [1992] HCA 2; (1992) 173 CLR 330.
[33] Even the most recent revision of the Berne Convention, the Paris Revision of 1971, took place before copyright protection for computer programs was contemplated: Cornish, ‘Computer Programs and the Berne Convention’ [1990] 4 European Intellectual Property Review 129, at 131. WIPO currently has under consideration a proposed Protocol to the Berne Convention which would deal expressly with computer programs. The First Session of the Committee of Experts was held in Geneva on 4-8 November 1991. The Report of that Session is published as WIPO document BCP/CE/II/4.
[34] Copyright Act, section 10(1).
[35] See, for example, Christie, ‘Copyright Protection for Ideas: an Appraisal of the Traditional View’ [1984] MonashULawRw 8; (1984) 10 Monash University Law Review 175; Hopkins, ‘Ideas, their Time has Come: an Argument and a Proposal for Copyrighting Ideas’ (1982) 14 Intellectual Property Law Review 385; and Pessa, ‘Evaluating the Idea-Expression Dichotomy: Rhetoric or a Legal Principle?’ (1991) March/April Computer Law & Practice 166.
[36] Some of the recent writings on this topic include Fetterman, ‘The Scope of Copyright Protection for Computer Software: Exploring the Idea/Expression Dichotomy’ (1990) 36 A.S.C.A.P. Copyright Law Symposium 1; Laurie, ‘Use of a “Levels of Abstraction” Analysis for Computer Programs’ (1989) 17 AIPLA Quarterly Journal 232; Polfanders (1989) 5 Journal of the Copyright Society of Australia 1; and Spivack, ‘Does Form Follow Function? - the Idea/Expression Dichotomy in Copyright Protection for Computer Software (1988) 35 University of California Los Angeles Law Review 723.
[37] [1992] HCA 2; (1992) 173 CLR 330, at 335-6.
[38] Draft Report, paragraph 6.13.
[39] The current definition of computer program is based on the notion of computer programs written in traditional (ie. sequential) programming languages. It is not certain, however, that programs in declarative languages are not within the definition. An argument may be made to the effect that the differences between declarative and sequential programming languages are not such as to make a program in a declarative language outside the current definition.
[40] Draft Report, paragraph 6.11.
[41] Copyright, Designs and Patents Act 1988.
[42] Draft Report, paragraph 6.16.
[43] Draft Report, paragraph 6.09.
[44] The term ‘software’ is used here instead of ‘computer program’, so that the latter may be reserved for software that is found to fall within the definition of ‘computer program’ in the copyright legislation. For the purposes of this discussion, ‘software’ is defined as a set of logical manipulations to be performed on data.
[45] See the supplementary comments of Mason CJ, and Brennan and Deane JJ, in Autodesk v Dyason [No. 1] [1992] HCA 2; (1992) 173 CLR 330, at 335-6.
[46] NEC v Intel 10 USPQ2d 1177 (ND Cal 1989); (1989) 14 IPR 1.
[47] For the purposes of this article, an ASIC is defined to be a programmable array logic (PAL) device, a programmable logic array (PLA) device, or any other application-specific device consisting of interconnections representing Boolean or similar logic.
[48] [1989] FCA 295; (1989) 15 IPR 1, at 25.
[49] Dyason v Autodesk [1990] FCA 469; (1990) 24 FCR 147, at 151 and 167-8 (Lockhart and Sheppard JJ, Beaumont J not deciding).
[50] [1992] HCA 2; (1992) 22 IPR 163, at 171.
[51] 91/250/EEC, Official Journal No L 122/42.
[52] Harkin, ‘A Design Right for Computer Programs?’, (1993) 3 Computer Law & Practice 94.
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/JlLawInfoSci/1993/16.html