AustLII Home | Databases | WorldLII | Search | Feedback

Journal of Law, Information and Science

Journal of Law, Information and Science (JLIS)
You are here:  AustLII >> Databases >> Journal of Law, Information and Science >> 2003 >> [2003] JlLawInfoSci 3

Database Search | Name Search | Recent Articles | Noteup | LawCite | Help

Hughes, Anton --- "A Comment on Software, Patents, Innovation and Openness" [2003] JlLawInfoSci 3; (2003) 14 Journal of Law, Information and Science 50

A Comment on Software, Patents, Innovation and Openness

ANTON HUGHES[*]

Abstract

The history of intellectual property protection in computer software reveals a unique experiment whereby all possible paradigms of protection have been tried. This article looks at whether the current trend towards unfettered patenting of software is likely to succeed where other paradigms have failed. An analysis of the rationales of the patent system and the nature of innovation in the software industry reveals that the balance has been poorly struck, with open source software projects being particularly vulnerable. Given that the problems which exist are due to the unique nature of the software industry, the author looks at ways in which industry-specific reforms can be approached.

1. Introduction

Software and intellectual property have had a stormy relationship. The patent system is the most recent intellectual property paradigm to be applied to provide innovator incentives to software developers. This is a task at which both trade secrets and copyright have failed. Yet recent events in Europe – a popular movement determined to put an end to EU moves to pass a directive requiring software patents – suggest that cracks are appearing in the armour of the once seemingly unstoppable darling of the maximalist camp. This article explores the issue of why software patents are such a thorny issue, by comparing the rationale of the patent system to the way in which the software industry innovates. One subset of the software industry is of particular interest – the Open Source Software (“OSS”) community, which provides a mechanism for global, collaborative innovation based on principles of freedom and sharing.

Section 1 looks at the history of intellectual property in the software industry, charting a course from complete freedom through each of the major protection paradigms, arriving at the current trend towards an increased availability of software patents. Section 2 considers the rationales which underpin patent protection, pointing out that there is more to patents than providing incentives for innovation – namely a furtherance of the public interest.

Section 3 looks at various features of the software industry which distinguish it from other industries in which patents are available. The nature of the software industry makes it poorly suited to coping with the demands of the patent system. This theme is developed through Section 4 where an important subset of the software industry – the free and open source software development community – is shown to be particularly vulnerable to the whims of patent holders in ways that conflict with the rationale behind awarding patents. In Section 5, various options for reform are considered, from the abolition of software patents to private initiatives. The success of the patent system in other industries means that a software-industry-specific solution is the most practical and likely to succeed, so various technology-specific solutions are discussed.

2. The growth of intellectual property protection for software

The history of intellectual property protection for computer software reveals a unique experiment. Following an initial period of openness, successive intellectual property paradigms have been applied to protect software – first trade secrets, then copyright, now patents. Trade secrets and copyright have both failed to adequately protect all valuable aspects of software, begging the question of whether patents will be any more successful.

2.1 Openness

Early in the development of the computer science discipline, there was little commercial involvement and therefore little need to protect new discoveries. Software development was regarded as a pure science, and ‘[f]or scientific results to be justified, they must be replicable.’[1] To be replicated, the source code of software innovations had to be shared, either by direct means or through publication in freely available journals.[2] For example, the Unix operating system, initially developed at AT&T’s Bell Labs by Thompson and Ritchie in the late 1960s and early 1970s, was normally distributed ‘with a complete set of source code and development tools.’[3] During a sabbatical at the University of California in Berkeley, Thompson installed Version 6 of Unix on a mainframe there. This single installation then formed the basis of the freely distributed ‘Berkeley Software Distribution’ or BSD. The BSD project was the major driver in the further development of Unix for a number of years.[4] Along the way the BSD project gave rise to, amongst other things, the TCP/IP stack which forms the backbone of the Internet.[5]

This concept of computer science as a pure science, somewhat akin to mathematics, is consistent with the initial view that software was unpatentable. Software, viewed as a computer-based implementation of a mathematical algorithm, was thought not to be patentable subject matter. The leading case on patentable subject matter in Australia is NRDC v Commissioner of Patents.[6] According to NRDC, ‘a process, to fall within the limits of patentability … must be one that offers some advantage which is material, in the sense that the process belongs to a useful art as distinct from a fine art .. that its value to the country is in the field of economic endeavour’. As a pure science, software fell in ‘the category of laws of nature, discoveries, schemes and plans, and fine arts as opposed to useful arts’.[7] A similar position was adopted in the US. [8]

2.2 Trade secrets

As commercial organisations began to see value in software development, they wanted to be able to protect their investments from imitation. The usual method by which software products were protected was a combination of secrecy and trade secret law.[9] At the time it was extremely difficult and time consuming to decompile object code into source code,[10] and thus first-movers gained a period of exclusivity in which to recover the costs of their investment.

As the state of the art improved, decompilers and other reverse engineering methods got better and the lead-time got shorter. Although reverse engineering remains a time consuming task, the lead time has shortened to such a degree that trade secrets have ceased to be an effective way of ensuring an appropriate reward for innovative research and development.[11] This trend towards a breakdown of the trade secret paradigm is not unique to software. Across many, if not all industries, technological developments have eroded the natural lead time which reverse engineering once awarded to innovators, with most modern products bearing their know-how on their face.[12]

2.3 Copyright

With trade secrets no longer offering adequate incentive to innovators, software came to be recognised as a literary work under the copyright regime. There is a sense in which it was reasonable to extend copyright protection to software, as ‘at first glance, a program appears to be a textual work. Source code is clearly some form of text, even if in a strange language not easily read by the casual observer’.[13] However, as noted by the court in the Apple case,[14] use of the copyright paradigm to protect object code distorts the conceptual consistency of a literary work,[15] as object code is not ‘intended to afford either information and instruction, or pleasure, in the form of literary instruction’[16] to any person. Although it serves a practical purpose in directing the operation of a computer, this has been insufficient to give rise to copyrights in previous cases.[17] Be that as it may, the post-Apple amendments to the Copyright Act 1968 in 1984 have specifically included ‘a computer program or compilation of computer programs’ within the definition of a literary work in s10(1) thereby ending the debate.

This may have been the end of the story, but for a more serious problem. The copyright paradigm only protects the literal text of software, and ‘the primary source of value in a program is its behaviour, not its text’.[18] Although text and behaviour are directly linked, they are not identical. It is possible for any skilled programmer to ‘copy the behaviour of a program exactly, without appropriating any of its text’.[19] In this situation, the ‘imitation’ will not infringe the copyright in either source or object code, but will behave in a manner indistinguishable to an end user from the original. Despite these limitations, copyright continues to be an important method of protecting innovators from the direct replication of software products.[20]

2.4 Patents

Patents, unlike copyright, do protect behaviour. So here and in the US the original reluctance to allow software patents was overcome. To date there have only been two Australian cases in relation to software patents, so to properly understand the development of the current position, it is necessary to refer to US law.

In the US, the 1970s case of Gottschalk v Benson[21] had confirmed that ‘mathematical algorithms (not just formulae) were … non-patentable subject matter’.[22] As software was considered to be a mere ‘concatenation of unpatentable algorithms’,[23] it too was held to be unpatentable.

The first moves towards patentability began in 1981 in the case of Diamond v Diehr,[24] with the United States Supreme Court upholding a patent on a computer-based invention that contained software to continuously monitor the temperature inside a synthetic rubber mould.[25] The basis of the court’s decision was that the appellant did not seek to patent the software itself, only ‘the particular application that [the applicant] had invented in the context of an ‘otherwise statutory’ industrial process’.[26] This decision was confirmed in a subsequent line of cases which also gave the test its name.[27] The Freeman-Walter-Abele test sought to limit the patentability of pure algorithms through a two step process. The first step was to ‘determine whether an invention contains a mathematical algorithm’.[28] If so, then the algorithm should be removed from the invention, and the remaining part analysed for patentability. Whilst representing a genuine attempt to place limits on the patentability of software algorithms, in practice this ‘doctrine of the magic words’ proved to be a minor obstacle to clever patent attorneys wanting to get software inventions patented.[29]

In Australia, a 1984 review of the Australian Patent Office (‘APO’) guidelines saw the Freeman-Walter-Abele test become the basis of the Australian approach to patentability of software.[30] However, its applicability was rejected by the Federal Court in the 1991 case of IBM Corporation v Commissioner for Patents.[31] An application by IBM to patent ‘a method for producing a visual representation of a curve image’ was rejected by the Delegate of the Commissioner of Patents on the basis that it merely recited an algorithm. On appeal to the Federal Court, Burchett J overturned the Delegate’s decision, on the basis that the invention produced ‘a commercially useful effect in computer graphics’.[32] According to His Honour, ‘[i]t is by the production of some useful effect that patent law has distinguished, so far as it has distinguished, between the discovery of a principle of science and the making of an invention’.[33] His Honour supported his reasoning by way of an analogy between a new use of a known algorithm, and the new use of a known chemical compound in NRDC.[34]

The Full Federal Court affirmed the correctness of Burchett J’s approach in the later case of CCOM Pty Ltd v Jiejing Pty Ltd.[35] This case involved a counter-claim for revocation made in infringement proceedings in relation to ‘a computer processing apparatus for assembling text in Chinese characters’. In upholding the patentability of this software invention, the Court reaffirmed the need to focus on the NRDC requirements of patentability.[36] The APO manual now takes the following statement from CCOM v Jiejing as the test of patentability for computer related inventions:

[whether there is] a mode or manner of achieving an end result which is an artificially created state of affairs of utility in the field of economic endeavour.[37]

In the view of the APO, the following will almost always be patentable, as they fit the definition of a ‘mode or manner’ described above:

(a) source code for patentable computer software;

(b) executable code for patentable computer software, which is in a machine readable form; and

(c) a computer, when programmed to achieve any result which has utility in the field of economic endeavour.[38]

The overall result has been that although in theory algorithms remain unpatentable per se, in practice software implementations of algorithms are regularly patented.[39] And by all accounts, the number of these patents is growing.[40] Perhaps the best known example of a patented algorithm is the recently-expired Lempel-Ziv-Welch (‘LZW’) compression algorithm patented by Unisys,[41] which came to be used in GIF and PDF files and in the standard for Postscript documents.[42] A less well known algorithm is the Viterbi algorithm which ‘helps calculate the probability that a sequence of observations represents an intended ideal construction (e.g. that a mispronounced word is actually that word). It was published in 1967 and has a vast field of applications’.[43] This one algorithm is estimated to have given rise to 4,470 patents in different fields of technology from data transmission to business process optimisation.[44] The reach of this algorithm into the Australian jurisdiction can be confirmed by a search of published patent data for the word Viterbi in patent application titles.[45]

The Australian position is roughly consistent with the current US requirement of a ‘useful, concrete and tangible result’ as espoused in State Street Bank & Trust v. Signature Financial Group,[46] although Christie and Syme suggest that the Australian requirement is more liberal, with Australia having “one of the least stringent tests in the developed world”.[47]

Many take the current protection regime for software to be as it should. For example, the Australian Patent Office believe the High Court to have endorsed the patentability of software products in Data Access Corp v Powerflex[48] when they stated that ‘[i]n form, the definition of a computer program seems to have more in common with the subject matter of a patent than a copyright’.[49] Whilst this correctly recognises the inability of the copyright paradigm to protect the behavioural elements of a software product,[50] it seems more likely that both the US and Australia have replaced the earlier problems of under-protection with those of over-protection. Two factors in particular suggest that US law is now too liberal in awarding software patents:

1. impressionistic evidence indicates that the pace of software innovation in the United States was very high before the enhanced intellectual-property rights were established and has not increased since;[51]

and

2. ‘most programmers (in contrast to the executives of the software companies) believe that the new rules significantly hamper their work’.[52]

In order to determine the nature of the mismatch between the patent system and the needs of the software industry, it is necessary to investigate the patent paradigm in more detail to determine how it meets the needs of the software industry – a task to which the next three sections of this article are directed.

3. Rationales underpinning patent protection

Senator Robert Ray in the Second Reading Speech for the Patents Act 1990, put forward the rationale of the patent system as follows:

The essence of the patent system is to encourage entrepreneurs to develop and commercialise new technology.[53]

Implicit in Senator Ray’s statement is an assumption that the encouragement of entrepreneurs to develop new technologies is an unassailably valuable goal. The primary justification is an economic one. Patents create an incentive to innovate by solving the problem of market failures arising from free-rider imitation.[54] The period of exclusivity of exploitation of the invention allows investors to recover the quantum of their investment in research and development. As Bessen puts it, ‘unless would-be competitors are constrained from imitating an invention, the inventor may not reap enough profit to cover [the cost of research and development]’.[55]

The debates in this context rage about whether the patent system, as it currently stands, acts as an incentive or a disincentive to innovate. On one hand, maximalists (or high protectionists) argue that stronger rights create stronger incentives to innovate, and favour expansive interpretations of patentable subject matter. On the other hand, minimalists are deeply sceptical of the monopolistic nature of patent rights and would prefer to see the scope of patents confined to ‘the minimum demonstrably necessary’.[56] The different standpoints are well illustrated in relation to whether Australian law should recognise an exemption from infringement for experimental use (research). In the maximalist view, reducing the ability of patent holders to control the full range of uses of their invention amounts to an unjustifiable tax on rights holders which reduces their incentive to innovate.[57] According to minimalists, an experimental use exception is consistent with the patent bargain in which the innovator is rewarded with a monopoly in exchange for transferring the know-how to the public domain.[58]

3.1. A case of pure economics?

One feature which both the maximalist and minimalist positions share is their commitment to a narrow economic analysis. Even in the discussion of highly sensitive issues such as gene patenting, ‘much of the most “critical” intellectual property scholarship tends to be relentlessly utilitarian’.[59] Whilst it might be expected that maximalists have the most to gain by using economics to exclude non-economic political, moral and social interest groups,[60] it seems that the minimalists are the most ardent supporters of pure economics.[61] This reluctance to embrace a wider range of concerns is attributable to a number of factors, including:

• a belief by minimalists that their economic argument is a strong one,[62] combined with a concern that maximalist arguments based on natural rights and the importance of property are more easily understood by and more powerfully delivered to the ‘unwashed masses’;[63]

and

• intellectual property scholars are more comfortable with the apparent certainty of narrow economics-driven enquiries.

Although the economic debate is useful in that it allows some comparison of competing claims, it often disguises hidden assumptions about the nature of innovation and fails to account for complex political, moral, social and cultural factors which cannot easily be reduced to equation form. For example, it is much simpler to equate increases in social welfare as a result of gene patents with an increase in gross domestic product, than it is to determine the fairness of the distribution of wealth which flows from the award of particular gene patents.

Patents were not always the province of dry economic theorising, however. At the time of the introduction of the patent system in 1623, patents were a politically charged issue which went to the heart of the political power struggles between the monarch and parliament.[64] The Statute of Monopolies from which the Australian patent system continues to draw key concepts of patentability, was passed in order to restrict the power of the monarch to award monopolies as political favours.[65]

There are signs that patents are becoming a political hotbed once again. For example, a popular movement to prevent the European Union from passing a directive requiring member countries to permit software patents has had recent success in delaying, possibly permanently, this directive through the EU Parliament.[66] A key component of this popular movement was a petition against software patents which has so far collected almost 380,000 signatures.[67]

Boyle suggests at least two lines of inquiry which should supplement the economic analysis of innovation:

1. An articulation of the implicit choice between systems of innovation on environmental, distributional, deontological or other grounds.[68]<

/p>

Examining the ‘structure of incentives that intellectual property policy sets up in the policy making process itself.’[69]

3.2. Alternative Methods of Encouraging Innovation

Another failing of intellectual property scholarship has been an implicit assumption that patents are the only (or at least the only effective) way of encouraging innovation. However, there are at least five main strategies that can be employed by governments that wish to solve the problem of the free rider effect discouraging innovators, of which patents rights represent only one. [70] Governments can:

• engage in technological innovation themselves;

• subsidize innovative activities by private actors (e.g. grants given to artists);

• give post-hoc prizes or rewards to persons and organizations that provide the public socially beneficial innovations (e.g. the Nobel prize);

• help innovators to conceal from the public information essential to implement their innovations, thus increasing their ability to charge persons who wish to take advantage of those breakthroughs; or

• confer intellectual-property rights upon innovators.

This being the case, when any extension to the scope of the patent system is proposed, ‘the potential availability of the other four approaches should always be considered’.[71] Allowing particular patents is only justifiable when it is the optimal method of encouraging innovation. In fact, it would also be advisable to consider the possibility of innovation in the absence of any incentive. As Loughlan states, when a patent is awarded for an invention which would have been developed anyway, ‘then society endures the social costs of patenting with no corresponding benefit and there is accordingly a net social loss’.[72]

3.3. The public interest

Scholars pursuing an understanding of the economics of innovation only are in danger of neglecting half of the patent equation. In addition to creating incentives for entrepreneurs, the patent system is directed to furthering the public interest. This additional function is balanced against the interests of inventors by ‘allowing inventors exclusive rights for a limited period to exploit their inventions, while at the same time promoting competition and innovation by ensuring that such inventions are fully disclosed to the public’.[73] This balance is similarly articulated by Article 7 of the Treaty on Trade-Related Aspects of Intellectual Property (‘TRIPS’):

The protection and enforcement of intellectual property rights should contribute to the promotion of technological innovation and to the transfer and dissemination of technology, to the mutual advantage of producers and users of technological knowledge and in a manner conducive to social and economic welfare, and to a balance of rights and obligations.[74]

By requiring inventors to disclose the details of their inventions to the public, two purposes are served. Firstly, once the patent period expires, proper disclosure enables competitors to use the technology and thus ensure that the technology enters the public domain. Secondly, it allows the state of the art in that industry to move forward by encouraging others to build upon the ideas contained therein, either through ‘inventing around’ the patent or developing a better version.[75]

The public interest also comes into play in the discussion of the extent to which public policy considerations should influence the structure of the patent system. Largely, this concerns the scope of patentable subject matter, although it also involves questions as to what type of innovation should be promoted within society. Those who argue for a recognition of public policy considerations find themselves pushed aside by the narrow confines of the economics-centred debate discussed above. For example, the Australian Law Reform Commission (‘ALRC’) in their review of the gene patenting debate, whilst recognising significant social and ethical issues existed in relation to gene patenting, decided that it would be better to address them outside the patent system.[76] This approach is also supported by the courts, with Anaesthetic Supplies Pty Limited v Rescare Limited[77] being authority for the proposition that it is for Parliament, not the courts or the Patent Office to decide whether social or ethical issues are relevant to the question of patentability. In Bristol-Myers,[78] Finkelstein J explained that the question of patentability ‘cannot depend upon the resolution of moral or ethical issues. Judges should not be called upon to resolve moral questions and, speaking generally, legal principles are not to be ascertained by reference to standards of ethics or morality’.[79]

One can understand the reticence of judges to engage in moral debates, but public policy considerations are central to the purpose of the patent system.[80] Encouraging new technologies, and particular forms of innovation have social and ethical outcomes. Failure to consider these outcomes amounts to an implicit acceptance of these them without proper debate. Further, despite pronouncements to the contrary, the courts in both Anaesthetic Supplies and Bristol-Myers showed an ability ‘to balance a variety of competing considerations, such as the likelihood that the grant of a patent would impede the willingness on the part of members of medical colleges to share information about new discoveries and new methods of treatment, in a competent manner’.[81]

4. The nature of innovation in the software industry

Having established the goals of the patent system, it is possible to measure its performance in relation to the software industry by looking at the needs of that industry and how they interrelate to the patent system. This section will show how the patent system is failing the software industry by failing to take into account the distinctive nature of innovation in this field.

4.1. Sequential innovation

Perhaps due in part to the sharing ethos developed in the early stages of the development of computer science, and in part due to the importance of familiarity to users,[82] ‘[p]rogrammers commonly adopt software design elements - ideas about how to do particular things in software - by looking around for examples or remembering what worked in other programs’.[83] As such, most new developments of the technology tend to be cumulative grain-sized innovations rather than big leaps forward. As such, it is extremely unlikely that any software invention should be considered to be new or non-obvious in the relevant sense.

This point is well illustrated by the criticisms made of the Amazon 1-click business method patent from within the industry. The 1-Click system fundamentally relies on ‘cookie’ technology, ‘a small amount of data placed on client computers to add state and session management capabilities to the World Wide Web, which was introduced in 1994 … [and] had been deployed on thousands of sites well in advance of [Amazon’s] 1997 patent application’.[84] Paul Barton-Davis, one of the founding Amazon programmers, accurately captures the cumulative nature of innovation in the field, and the negative impact which improperly awarded patents is likely to have on innovation:

Tools like the GNU C and C++ compilers, Perl, the Unix mmap(2) call, the Berkeley DBM library: these are all software tools that contain far more innovative and significant ideas than Amazon.com's 1-Click (TM) system, and in turn depend upon ideas and innovations that came before them. If developers of these tools, or the researchers and software engineers who worked on the systems that led to their development, had sought and been awarded patents on the many marvellous computational methods that their tools embody, Amazon.com's early existence would have been a costlier and less efficient one.[85]

Yet courts and patent offices consistently under-emphasise the importance of existing technology to software development. This may be in part due to a lack of documented prior art,[86] but could also be explained by the adoption of analogy with cases where a new use is found for an existing product. A leading example in Australia would be the NRDC case, where an invention was found to subsist in the new use of a known chemical compound.[87] The effect of this under-emphasis is that large rewards are available as the result of minimal research. The lure of large profits thus encourages opportunists to obtain and enforce broad patents at the expense of innovation in the field.[88] It seems that Loughlan is right to suggest that the question of who is the free-rider depends on context,[89] and here the finger should be pointed at Amazon rather than Barnes and Noble.[90]

4.2 Complementarity

Whilst direct replication of software in the copyright sense is clearly deserving of protection, imitation in the functional sense can also play an important part in software development, in that different attempts to implement identical functionality will often differ in the approach taken. This new implementation may achieve improved performance, greater intuitiveness or successful commercialisation. In this way, the imitation complements the original design.[91] Because patents reserve all exploitation of an invention to the patent holder, an alternative implementation of functional equivalence is likely to fall within the scope of the patent. Thus an alternative is unlikely to be attempted.[92] In this market then, both competition and innovation are lessened, and the public interest in new technologies is poorly served.

4.3 Poor documentation

In the software industry, much of the prior art lies in areas outside the purview of Patent Offices. In many cases, the only complete documentation of a software invention will be the source code itself. This is not to say that no documentation exists for these projects, as user guides, application programming interfaces[93] and other developer documentation are often available. However, the software industry should be distinguished from other fields such as engineering, where implementation details are generally published in academic or industry journals.[94] It is true that free and open source software projects make source code available. However at present the dominant model for development is still the proprietary model, where source code is a jealously guarded secret. Even assuming access to the source code is available, it is only understandable to programmers well-versed in the language it is written in, and certainly not in a form accessible to patent examiners with only a short time to determine the extent of the prior art base. The problem is even greater in relation to the growing market for Internet technologies. The ephemeral nature of the Internet can often mean that key components of the prior art base may have disappeared off the face of the web by the time a patent application comes to be considered, or infringement proceedings are afoot.[95]

The lack of a suitably complete prior art base diminishes the quality of software patents by lessening the threshold for inventive step and novelty, and encourages an increased volume of software patents. Another consequence can be that it is difficult to mount an effective challenge to invalid patents. The end result is an increased number of patents of dubious value reducing both innovation and competition.

4.4 Network effects

Network effects are common in software development, where having a trained base of users for a particular product creates barriers to entry for competitor products. A network effect refers to ‘the social advantages that arise when all of the users of a particular type of technology adhere to the same standards and thus can share their work and move easily between machines and businesses’.[96] A simple example is the Microsoft Word file, which is a de facto standard for storing and exchanging word processor documents. Competitors in the word processing market such as OpenOffice.org and WordPerfect must include support for reading and writing Word files because computer users expect anyone to be able to access files in this format. This gives Microsoft a great deal of power in the word processing software market, a power which they use to leverage related markets such as spreadsheets and presentation software.[97]

Network effects may also arise as a result of the adoption by the industry of a particular standard. For example, the TCP/IP protocol is the standard for the transmission of digital information over computer networks, particularly the Internet. This however, is an example of a standard which is not controlled by any one organisation, something which has been the key to its widespread adoption.[98]

When a standard falls within the scope of a patent, this gives the patent holder a right to enforce a licence fee from not only all competitors needing their products to interoperate with the standard, but also users of competing products.[99] This is because the patent cannot be ‘invented around’ without deviating from the standard. Thus network effects amplify the impact of patents tremendously by removing the bargaining power of competitors.

Network effects also raise the related issue of when to allow reverse engineering in order to allow interoperability and encourage competition, as compared to reverse engineering which facilitates free-riding.[100] The importance of reverse engineering to ensure interoperability is specifically excused under the Copyright Act 1968 (Cth).[101] However, the status of reverse engineering is not at all clear in the Patents Act 1990 (Cth). If it is allowed, it would be allowed under an experimental use exception. Assuming such a defence exists, a commercial purpose would most likely render it unavailable in the majority of cases.

4.5 Ease of infringement

For software products, there is no one-to-one correlation between patents and products, a vastly different scenario to that which existed for the physical machines which the patent system was originally designed to protect. A patentable software ‘invention’ may only require a few lines of code to implement. A complete software project will involve thousands, or possibly millions of lines of code.[102] Also, patents are not considered a useful source of technological information within the software industry.[103] Further, developers are given a disincentive to become aware of possible patent infringement. In the US, wilful infringement of patents can give rise to triple damages,[104] whilst in Australia innocent infringement can prevent an award of damages or an account of profits.[105] These factors, combined with the structure of the industry discussed below combine to mean that patent searches are not regularly conducted by software developers at the commencement of a project, unlike other industries such as biotechnology where such searches are commonplace.[106] Thus it is fair to say that most software developers are walking into a minefield of infringement suits with their eyes firmly shut.

For those parties wishing to avoid infringement, the only solution is to negotiate licences with each patent holder. The sheer volume of patents involved can make this task so time-consuming and costly that it is not worth continuing.[107]

4.6 Industry structure

The software development community is estimated to include 10 million active programmers worldwide.[108] This is due to both the demand created by the ubiquity of computers in all facets of modern life, and the fact that to become a programmer requires only a low cost computer and a desire to learn.[109] Further, despite the prevalence in the public consciousness of large firms such as IBM and Microsoft, the software development community is generally made up of small-to-medium enterprises (‘SMEs’).[110] SMEs are particularly vulnerable to the problems so far discussed. Their limited budgets means that they can afford neither substantial licensing fees, the costs involved in challenging invalid patents and defending against infringement, nor the maintenance of defensive patent portfolios. Thus when the patent system fails the software industry, these are the players who get hit the hardest.

4.7 Pace of change

The pace of change in the software industry is rapid, largely driven by the similarly fast-paced improvements in computer hardware. The pace of change within the hardware industry is referable to Moore’s law – a prediction made by Intel cofounder Gordon E. Moore in 1965, that computing power will double every 18 months.[111] Although its accuracy can be challenged and its application to software less clear,[112] it has become somewhat of a mantra within the information technology field which captures the ‘innovate or die’ outlook driving the industry forward.

The major concern for an industry which develops at such a pace is the length of the patent term. 20 years is an eternity in information technology. The useful life of a software invention is more like 3-5 years.[113] Thus once 5 years is reached, the patent has arguably achieved its goal of rewarding innovation. When patents persist after this term, their continued existence can sometimes threaten new technologies which through historical accident come within the dinosaur patent’s claims.[114] Here the patent is no longer protecting an invention, but providing a potential windfall to its user, with a negative impact on innovation.

5. Open Source Software development

The Open Source phenomenon, although a creature of copyright, provides a unique counterpoint to the expansionist trends of intellectual property in recent times. Open Source Software (‘OSS’) is licensed on very liberal terms which allow more than just access to the source code of the software. Generally speaking, the following features distinguish OSS licences:

• free redistribution by any party without payment of a royalty or distribution fee must be allowed;

• the program must be distributed with source code, or the there must be a well-publicised means of obtaining the source code for no more than the reasonable cost of reproduction;

• the licence must allow modifications and derived works, and must allow them to be distributed under the same terms as the licence of the original software.[115]

Free software is a subset of open source software which uses the same development methodology, but requires a greater degree of freedom to be granted:

• freedom to use the program for any purpose;

• freedom to examine the source code;

• freedom to distribute the program to others;

• freedom to improve the program.[116]

OSS represents an important subset of the software industry, because it embraces an alternative model of innovation and global collaboration which furthers the public interest in the development of new technologies. Having established in the previous section how patents threaten the software industry, it will be shown here how the threat to OSS is particularly acute.

5.1 Why OSS is important to the software industry

In an industry which rarely provides access to core documentation, the open source development model provides an important source of prior art. Being prefaced on a culture of openness, OSS development takes place in full public view. Version control systems[117] and bug trackers[118] provide an annotated history of the development of the code. The workings of the code are supplemented by developer-to-developer, developer-to-user and user-to-user conversations on mailing list archives[119] and discussion boards.[120] Further, because there is no need to maintain secrecy, implementation details are disclosed in project documentation or developer weblogs.[121]

The OSS development model produces superior software, at least in some circumstances. Bessen concludes on economic grounds that ‘proprietary [software] provision fails to effectively meet the needs of many customers in markets where customers have highly disparate needs and products are complex’.[122] Quantitative measures of market share, reliability, performance, scalability, security and total cost of ownership also suggest that such software is, at the very least, worthy of serious consideration.[123]

OSS can create competition in markets where proprietary companies cannot. For example, the Mozilla browser project began when Netscape open-sourced its Navigator product 1998, in light of rapidly declining market share as a result of an unwinnable ‘browser war’ with Microsoft’s Internet Explorer.[124] Netscape has long passed into obscurity, but a derivative version of the browser, Mozilla Firefox has recently reduced Internet Explorer’s market share to its lowest point in three years.[125] By bringing competition back into the browser market, Firefox is reintroducing innovation to what had become a stagnant market.[126]

The Internet is a case-in-point of the importance of commons-based innovation in the software industry. At the logical level, the TCP/IP protocol which ships data from server to user and peer to peer is beyond the control of any one organisation or person. This ‘packet neutrality’[127] distinguishes TCP/IP from all network designs which have gone before, and is the key to its successful adoption by a wide range of application developers.[128]

The point is not that the commons model should displace proprietary-based models. The two can, and do, happily coexist.[129] The point is rather that commons-based innovation provides an alternate route to the same goal as that of the patent system which will sometimes lead to the furtherance of the public interest in a more efficient manner.[130] To allow its demise because of an unquestioning acceptance of the dominant political view of the patent-holding elite would be a mistake of massive proportions.

5.2 Why OSS is vulnerable

To an extent, the problems faced by OSS developers are the same facing all software developers – as discussed above the nature of the industry itself makes patents too easy to get, too easy to infringe and too broad in scope. However, there are a number of factors which make the software patent problem more likely to create havoc for OSS developers. Firstly, the nature of the parties involved in OSS development makes them particularly vulnerable. OSS development is in most cases undertaken by a widely distributed group of volunteers and small companies. OSS developers are generally either small companies who make their money out of implementation and administration of the software on customer projects, or in many cases volunteers.[131]

Open source projects harness the collaborative power of the Internet which allows co-operation on projects on a global scale exceeding even that of the software industry monoliths.[132] The geographical distribution of project participants together with the ability of any party to ‘fork’ the code,[133] a central tenet of the open source paradigm, illustrates the distributed nature of OSS development in which no central body has ultimate control. Sometimes a project will have one or more leaders,[134] or a non-profit organisation which holds the intellectual property and hosts development services,[135] but the distributed nature of these projects are more analogous to Grokster than Napster.[136] This may be the key to the success of these projects, but the downside is that legal responsibility for patent infringement is likely to be directed at individual contributors and even users of the software. Whilst some may suggest that the sheer difficulty of pursuing a distributed target makes enforcement less likely,[137] many open source advocates have long felt that patents represent the biggest threat to this form of software development.[138] In a recent study undertaken by Open Source Risk Management, it was found that the Linux kernel potentially infringes 283 US software patents which have yet to be tested by the courts.[139] Stallman has estimated that the Linux kernel represents only 0.25% of a complete GNU/Linux system,[140] suggesting a possible total of between 30,000 and 300,000 possible infringements overall.[141] Although ‘a third of the patents are owned by Linux backers, including Hewlett-Packard, IBM, Novell, and Oracle, which are unlikely to assert claims’, such a precarious position is cause for concern.[142]

Microsoft’s actions in particular suggest patent enforcement against open source projects, particularly Linux, can be expected in the future. The first signs came from a leaked Microsoft document in 1998 which stated that ‘the effect of patents and copyright in combating Linux remains to be investigated’.[143] Secondly, a leaked memo from a senior Hewlett-Packard executive in 2002 asserts that ‘Microsoft will soon be launching a patent-based legal offensive against Linux and other free software projects’.[144] Finally, a provision in the recent Sun-Microsoft settlement over Java specifically allows Microsoft to pursue infringement claims against OpenOffice.org users and developers.[145]

Software patents are regularly used by large firms in a way not intended by the patent system. Although licensing represents a substantial revenue stream for parties holding a large patent portfolio,[146] IBM in 1990 noted that their primary purpose for collecting patents was the ability to mobilise their patent portfolio to ensure cross-licensing.[147] Similarly, in a 1991 Microsoft internal memo, Bill Gates noted that Microsoft’s strategy for addressing the problems of software patenting should be ‘patent exchanges with large companies and patenting as much as we can’.[148] Through cross-licensing, large firms like IBM can ‘measure the average’ of the patent system, avoiding the harm whilst enjoying the benefit.[149] Unfortunately, small players in the industry are unable to participate in this sort of risk minimisation due to the cost of maintaining such a portfolio. The problem is even more pronounced for distributed, multi-party OSS projects who lack a central body to fund and harvest a large patent portfolio, or to negotiate licensing agreements to cover project participants en masse.

A particular danger exists in regard to the inclusion of patented technologies in industry standards. For example, the MPEG-2 standard for displaying video on a computer using software contains 39 US patents. To implement the standard, an OSS project would have to pay licence fees for each of these patents or risk infringement proceedings. This constitutes an impenetrable barrier to entry for any open source project wishing to compete in this arena, and thus both innovation and competition in this market are reduced.

In summary, patents pose a real threat to free and open source software development. The loss of this valuable form of global collaboration and innovation, the likes of which has never been seen before, is irreconcilable with the aims of a patent system which claims to promote innovation and further the public interest. However, identifying the nature of the problem is only the beginning. The real challenge is finding the best way to adapt the patent system to remove the impediments it has placed in the way of innovation in the software industry, whilst not leaving innovators unprotected from imitation.

6. Reform

The previous sections have revealed a mismatch between the needs of the software industry and the raison d’être of the patent system. However, the patent system can and does work well for other industries, such as the pharmaceutical industry.[150] Thus any proposal for wholesale change of the patent system is likely to meet with stiff resistance from such industries, and with good reason. It is therefore necessary to consider the available framework in which an industry-specific solution to software patent problems can be addressed.

6.1 In support of a technology-specific solution

The ALRC’s recent report into the patentability of genetic materials[151] considered the feasibility of technology-specific solution. A number of factors were held to weigh against such an approach. Firstly, it was asserted that this approach would increase the complexity of patent law.[152] Yet the patent system is not technology-neutral at all – industry-specific nuances are currently addressed through the application of the seemingly neutral patentability requirements. For example, in biotechnology cases, the courts have focused on the uncertain nature of biotechnology, almost to the extent that it is ‘as much a black art as a science’.[153] As a result, US courts take a liberal view of non-obviousness, but uphold the disclosure requirement very strictly.[154] In the software cases however, these same courts take the view that software developers are so skilled as to make programming a ‘mere clerical function’,[155] which results in a strict application of the non-obviousness standard, whilst allowing a mere abstract functional description to satisfy the disclosure requirements.[156] Thus although the ALRC’s recommended approach may suffice to keep the legislation simple, the case law is already a complex terrain in which industry-specific jurisprudence is hampered by limited judicial expertise,[157] hindsight bias,[158] precedent[159] and reasoning by analogy.[160] The veil of technology-neutrality only serves to prevent a policy-based analysis of the needs of particular industries.

Secondly, the ALRC said that a technology-specific approach would lead to ‘inconsistency between the way in which genetic materials and technologies are treated under Australian patent law and the patent laws of other jurisdictions’.[161] This is an interesting contention given that Australia as a sovereign nation should be primarily concerned to implement laws which further the best interests of its citizens, and patent law represents a significant economic and political instrument with which to achieve this goal.[162] However, the flexibility to tinker with the Australian patent system is, in practical terms, governed by the World Trade Organisation (‘WTO’) Treaty on Trade-Related Aspects of Intellectual Property (‘TRIPS’) and further circumscribed by the Australia-US Free Trade Agreement. TRIPS sets minimum standards of intellectual property protection which must be implemented by all WTO member countries. Failure to meet these standards could potentially result in the imposition of trade sanctions. Article 27 of TRIPS requires that ‘patents shall be available for any inventions, whether products or processes, in all fields of technology, provided that they are new, involve an inventive step and are capable of industrial application’. Article 27 further mandates that ‘patents shall be available and patent rights enjoyable without discrimination as to … the field of technology’. Subject to limited exceptions,[163] this appears to require a technology neutral approach to patents.[164] But this is not the only possible interpretation of Article 27. Firstly, TRIPS requires only that patents be available for ‘inventions’, but does not define the term. This suggests some leeway in excluding certain subjects from TRIPS as non-inventions.[165] Further, the meaning of the phrase ‘without discrimination’ should be carefully considered. Discrimination is be defined as follows:

discrimination

1 unfair treatment of a person, racial group, minority; action based on prejudice.

2 subtle appreciation of matters of taste.

3 the ability to see fine distinctions and differences.[166]

It is submitted that the first definition is clearly preferable in this context, and thus ‘without discrimination’ could be written ‘without prejudice or unfair treatment’. Understood in this way, it seems reasonable to argue that technology-specific alterations are possible, so long as the protection thereby afforded is equivalent to the minimum standard of patent protection set out in TRIPS.

The final impediment the ALRC saw for a technology-specific approach was that it would set an ‘undesirable precedent for the way in which the patent system should accommodate new technologies in the future’.[167] In fact, the precedent has already been set. A sui generis protection regime already exists for circuit layouts as set out in the Circuit Layouts Act 1989 (Cth). The Act is based largely on the US Semiconductor Chip Protection Act 1984 and is a specific requirement of TRIPS (Article 35). The protections afforded under this Act differ from patent protection in that they recognise the importance of reverse engineering, protecting it in much the same was as the US Copyright Act protects ‘fair use’.[168]

6.2 Abolishing software patents

A number of people, including free and open source software figures such as Richard Stallman,[169] Linus Torvalds, [170] Tim O’Reilly[171] and Bruce Perens[172] argue that the only solution to the problems which software patents cause software developers is to abolish software patents altogether.[173]

The abolition of patents for software could be achieved in Australia through legislative action to specifically exclude software patents from the definition of patentable subject matter. Alternatively, it would be open to the courts to exclude software from the ambit of the ‘manner of manufacture’ test on the grounds that software patenting would be ‘generally inconvenient’.[174] As discussed above it could be argued that something which does not meet Australia’s ‘manner of manufacture’ test is not in fact an ‘invention’ and is not covered by TRIPS.[175] On this interpretation, the abolition of software patents is at least theoretically possible.

Whatever the theoretical possibilities, the practical reality is that the battle to prevent software patents in Australia and the US has in all likelihood been lost. In the US, over 80,000 software patents have already been issued. In Australia, the APO has supported software patents since 1985.[176] So it may be that the question of software patentability is one for the history books in these countries.[177] However, in Europe the debate about software patentability is likely to continue for some time yet. In recent months, a popular movement was been successful in delaying, although not preventing,[178] the passage of a European Directive on software patenting through the European Parliament despite strong pressure from the US.[179] Whatever the eventual result in Europe, the current debate draws much needed attention to the problems which software patents create and ignites a spark of hope for battle-weary minimalists that the tide may one day turn their way.

6.3 A sui generis regime for software

Ten years ago, a group of leading computer science and intellectual property scholars, rather than investigating ‘how existing laws can or should apply to computer programs’,[180] turned their efforts to a ‘normative analysis of the kind of legal protection that would be socially desirable for software and how it might best be accomplished’.[181] Their conclusion was that a sui generis regime was the only way to correctly strike the balance between creating incentives to innovate and upholding the public interest.[182]

Given the high degree of substitutability between hardware and software solutions, it is unsurprising then that the chosen regime is conceptually similar to the Circuit Layouts regime described above.[183] More importantly perhaps, the Circuit Layouts Act represents a unique attempt at rebalancing the innovation-disclosure equation. This is because the original US Act upon which our Act is based[184] made various concessions to the minimalist concerns that ‘there should be no protection at all since competition in the industry provided the impetus which made America the world’s leader in chip innovation.’[185] The circuit layouts regime seeks a resurrection of the trade secret paradigm through the creation of a compensatory liability regime in which an artificial lead time of two years is given to the registrant of a grain-sized innovation, in exchange for disclosure of the details of the innovation. All those wishing to use the innovation during the lead time must pay a royalty to the innovator, or else wait out the liability period.[186]

The historical role of trade secrets in protecting software products, the sequential nature and rapid pace of innovation in the field suggest that a similar regime is a good fit. Such a regime imports copyright-like protection against cloning and allowing independent creation, but unlike copyright it can protect the value inherent in a program’s behaviour. The regime suggested by Samuelson et al further refines the current dual copyright-patent protection approach by granting ‘a period of automatic anti-cloning protection and an opportunity to register innovative software compilations or subcompilations in order to get a longer period of blockage or a period of compensation under a standard licensing arrangement’.[187] The main difference from the circuit layouts regime is that the whole product would not need to be registered – a recognition of the many ideas which combine to make up a single software product.

One attraction of such a system is that it can be implemented with or without the continued availability of patents for software inventions. If implemented in substitute for the current software patenting regime, it could support an interpretation of the patentable subject matter requirement which recognises that the sequential nature of software innovations means they would always fall short of the inventive step requirement, and hence are not inventions. Further, Reichman has shown how under such a system no-one is worse off than they would be if software patents were available.[188] Thus it may be argued that the minimum levels of protection required under TRIPS are being met, albeit in a slightly different fashion.

If software patents are to be kept, then the provision of an alternative regime for software innovations would alleviate the need for courts and the Patent Office to continually lower the inventive step threshold in order to protect sub-patentable innovations. The regime also stands to increase the prior art base by giving an incentive to disclose in the form of a longer period of protection, and would reduce the number of patent applications, which would also increase the likelihood that only truly non-obvious, novel and useful software inventions were awarded patents.

6.4 Mitigating the damage

In the current Australian political climate, where our government waits to boldly go where the US has gone before, radical responses to the software patent issue such as those discussed above are unlikely to be implemented. Thus consideration must be given to other ways of limiting the negative impact of software patents.

GOVERNMENT RESPONSES

In the field of software development, the question of adequate disclosure is inexorably linked to breadth of claims. Patents are not supposed to protect ideas, only inventions. By enforcing the disclosure requirement much more strictly, the scope of claims would be considerably narrowed. This would have the additional benefit of limiting the useful period of the patent without altering the patent term required under TRIPS, and hence preventing opportunistic enforcement of doubtful patents such as the BT hyperlink patent[189] and the Eolas patent.[190] Further, such a move would have a positive impact on the usefulness of patent literature to those in the industry and allow proper scrutinisation of awarded patents to improve their quality. The most difficult question in this context is to determine to what level disclosure should be made. At one end of the spectrum is the current approach which requires only a high level functional description of the software. At the other is the copyright approach which protects only the exact implementation.[191] A mid-range possibility would be detailed description in pseudo-code in human readable form.[192] The biggest problem facing any solution focused on the grant of new patents however, is that it only applies to future patent applications. Thus, whilst such a solution has merit, the software industry would have to remain in a holding pattern for the next 20 years until the current crop of patents expire – a virtual eternity, as discussed above.

Compulsory licensing is one means of addressing the power of patent holders created by network effects. It operates by mitigating the power of patent holders to exclude competitors from the market. Under a compulsory licensing scheme, competitors gain licences to use a patented technology without the authorisation of the patent holder.[193] However, a ‘reasonable’ licence fee must still be paid, and a court order obtained, which represents a considerable barrier to entry for new competitors, especially small-to-medium enterprises and open source projects. Such an approach also does little to improve the quality or scope of patents which are currently being awarded. Another alternative is that when software product including a patented component becomes a standard, a free access right for the purposes of reverse engineering should be imposed.[194]

Various solutions have been suggested to address the imitation and reverse engineering problem. O’Rourke favours a ‘fair use’ defence, along the lines of the US doctrine of fair use available in copyright law.[195] Another possibility is a narrow construction of the claims of the invention to exclude independently implemented functional equivalents.[196]

Other solutions suggest an affirmative defence of alleged infringer. Fisher suggests a patent abuse defence designed ‘to prevent the developer of a software program that enjoys market power in one context … from leveraging that power into control of another field’.[197] This is clearly designed to minimise the advantages which network effects supply to patent holders. Some have suggested that a variation in which an outright defence against infringement should be available to the developers of open source and free software.[198] Stallman acknowledges what is perhaps the biggest problem with such a solution – namely that simply arguing for open source protection alone is not likely to be enough to encourage reform. Support from proprietary software developers was a key component of the successful movement against software patents in Europe.[199]

PRIVATE INITIATIVES

One pragmatic response to the problems facing open source developers and users was the formulation of the Open Source Risk Management corporation (‘OSRM’). This organisation was formed in the wake of the somewhat doubtful claims by SCO that portions of the Linux kernel infringed SCO’s rights as holder of the copyrights to UNIX.[200] The company offers ‘comprehensive, low-cost vendor-neutral [risk mitigation consulting, best practice protocols and indemnification service] to end-users, developers, and vendors’.[201] Litigation insurance will prove useful for those who pay the premiums. For those who do not, the OSRM study into Linux patent infringement suggests that patent holders may have a hard time enforcing their patents, as the distributed nature of open source projects means that an injunction would be practically meaningless.[202] This may be the case, but not something which developers should be too ready to risk their livelihoods on.

IBM has just extended its umbrella of patent protection to all open source projects by gifting licences to use any of 500 specified patents.[203] In order to gain access to the patents, a project has to be working on or using software distributed under a licence approved by the Open Source Initiative. Together with a promise to enforce their patents against any organisation pursuing infringement action against an OSS project, this unprecedented move brings a level of indemnity to OSS developers not dissimilar to the defensive portfolio strategy discussed above. IBM and others have previously promised to use their patents to protect open source projects they are involved in,[204] however this goes well beyond a mere promise, to be a legally binding commitment which will hopefully provide encouragement for those hesitating to embrace the OSS development model. It can only be hoped that other patent holders follow suit, thereby establishing a ‘patent commons’.

Richard Stallman had previously suggested a similar strategy for companies concerned to collect patents for defensive portfolios.[205] He suggests that these smaller companies declare publicly that their patents will only be used for defensive purposes, never for aggression. This addresses the situation where a small innovative company fails and their assets are then auctioned off to a ‘patent parasite’ looking for easy money. Such representations would form a basis for a counter-claim based on estoppel.

Another final strategy directs the global collaboration model of OSS development at defeating bad patents. The Public Patent Foundation[206] is a not-for-profit organisation which volunteers to search for prior art which will invalidate bad patents. The Foundation has recently been successful in having Microsoft’s FAT file system patent re-examined and invalidated by the US Patent Office.[207] Compared to the scale of the patent problem, this is a mere drop in the ocean, but is certainly better than nothing, and could act as an important mechanism for defeating particularly contentious patents through collective action in the future.

7. Conclusion

The stormy relationship of software and intellectual property seems likely to continue for some time to come. The patent system is supposed to create an incentive to innovate by solving the problem of market failures arising from free-rider imitation. Yet a narrow focus on the economics of innovation, and a failure to consider alternative forms of innovation is preventing a policy-based analysis of the system’s effectiveness by excluding social and ethical questions central to the public interest the patent system is supposed to promote.

It has been shown how the patent system is failing the software industry by ignoring the unique nature of innovation in this field. A professed commitment to technology neutrality and failure to consider the importance of alternative forms of innovation, such as open source software development, has straight-jacketed the software industry into various forms of private initiative aimed at harm minimisation. Private initiatives are doing their best to prevent this possibility, but there is a sense in which they amount to “swatting mosquitoes to cure malaria”.[208]

In the absence of a government commitment to consider vital reforms, opportunists are using software patents in ways irreconcilable with the aims of the patent system. Not only is this likely to impede software innovation across the board, but it could bring about the end of a valuable alternative form of global collaboration and innovation that is Open Source Software. It can only be hoped that governments wake up to the danger before it is too late.

Bibliography

Advisory Council on Intellectual Property, ‘Patents and Experimental Use’ (2004) Options Paper.

Allison JR, Lemley MA, Moore KA and Trunkey RD, ‘Valuable Patents’ (2003) George Mason Law & Economics Research Paper No. 03-31; UC Berkeley Public Law Research Paper No. 133:

http://ssrn.com/abstract=426020

(2 November 2004).

‘Amazon One-Click Shopping’:

http://www-cse.stanford.edu/classes/cs201/projects-99-00/software-patents/amazon.html

(11 November 2004).

Australian Law Reform Commission, ‘Genes and Ingenuity: Gene Patenting and Human Health’ (2004) ALRC 99:

http://www.austlii.edu.au/au/other/alrc/publications/reports/99/

(2 November 2004).

Australian Patent Office, Manual of Practice and Procedure:

http://www.ipaustralia.gov.au/resources/manuals_patents2.shtml

(2 November 2004).

Babcock C and Greenemeier L, ‘Open Source Stress’ Information Week, 9 August, 2004:

http://www.informationweek.com/story/showArticle.jhtml?articleID=26806464

(3 January 2005).

Bangeman E, ‘Will Microsoft sue OpenOffice users?’ ArsTechnica, 16 September, 2004:

http://arstechnica.com/news.ars/post/20040916-4192.html

(14 January 2005).

Barr J, ‘Live and let license’ Linuxworld.com, 23 May, 2001

http://www.itworld.com/AppDev/350/LWD010523vcontrol4/

(14 January 2005).

Barr J, ‘HP memo forecasts MS patent attacks on free software’ Newsforge, 19 July, 2004:

http://www.newsforge.com/article.pl?sid=04/07/19/2315200

(14 January 2005).

Barton-Davis P, ‘I Oppose Amazon.com’s 1-Click Patent’:

http://www.equalarea.com/paul/amazon-1click.html

(12 January 2005).

Bessen J, ‘Open Source Software: Free Provision of Complex Public Goods’:

http://www.researchoninnovation.org/opensrc.pdf

(11 November 2004).

Bessen J and Maskin E, ‘Sequential Innovation, Patents and Imitation’ (2000) MIT Working Paper:

http://www.researchoninnovation.org/patent.pdf

(11 November 2004).

Bezos J, ‘An Open Letter from Jeff Bezos on the Subject of Patents’ O’Reilly Network, 3 September, 2000:

http://www.oreilly.com/news/amazon_patents.html

(18 January 2005).

Bicknell C, ‘Cashing in on Bogus Patents’ Wired News, 19 October, 2000:

http://www.wired.com/news/business/0,1367,39528-2,00.html

(12 January 2005).

Boyle J, ‘Enclosing the Genome: What the Squabbles over Genetic Patents Could Teach Us’ To be published in F. Scott Kieff (ed), Perspectives on the Human Genome Project:

http://www.law.duke.edu/ip/pdf/enclosing.pdf

(1 September 2003).

Burk DL and Lemley MA, ‘Is Patent Law Technology-Specific?’ (2002) 17 Berkeley Tech. Law Journal 1155:

http://ssrn.com/abstract=349761

(2 November 2004).

Burk DL, ‘Cyberlaw and the Norms of Science’ (1999) Intellectual Property and Technology Forum:

http://infoeagle.bc.edu/bc_org/avp/law/st_org/iptf/commentary/content/burk.html

(18 March 2005).

Caplan P, ‘Patents and Open Standards’ (2003) 14(4) Information Standards Quarterly:

http://www.niso.org/press/whitepapers/Patents_Caplan.pdf

(18 January 2005).

Carroll J, ‘Software patents need shelter from the storm’ ZDNet UK, 17 September, 2003:

http://comment.zdnet.co.uk/other/0,39020682,39116437,00.htm

(5 January 2005).

CCOM Pty Ltd and Another v. Jiejing Pty Ltd and Others [1994] FCA 396; (1994) 122 ALR 417:

http://www.austlii.edu.au/au/cases/cth/federal_ct/unrep6887.html

(2 November 2004).

Chartrand S, ‘Patents: Disproving Idea Ownership’ New York Times, 23 October, 2000:

http://www.nytimes.com/2000/10/23/technology/23PATE.html

(22 November 2004).

Christie A and Syme S, ‘Patents for Algorithms in Australia’ (1998) 20 Sydney Law Review 23.

Cockfield A, ‘Towards a Law and Technology Theory’ (2004) 30 Manitoba Law Journal 383:

http://ssrn.com/abstract=615088

(11 November 2004).

Cohen JE and Lemley MA, ‘Patent Scope and Innovation in the Software Industry’ (2001) 89 Californian Law Review 1:

http://www.law.georgetown.edu/faculty/jec/softwarepatentscope.pdf

(2 November 2004).

Collins Australian Dictionary and Thesaurus, 3rd Edition, HarperCollins, 2004.

Commonwealth of Australia, ‘Patents Bill 1990: Second Reading’ Senate, 29 May, 1990

http://parlinfoweb.aph.gov.au/PIWeb/view_document.aspx?id=562046&table=HANSARDS

(2 November 2004).

Copyright Law Review Committee, ‘Chapter 8: Circuit Layouts in Copyright and Contract’ 23 August 2004:

http://sgeag001web.ag.gov.au/agd/www/Clrhome.nsf/0/937143BFE876FAD2CA256C4F000F065F?OpenDocument

(19 January 2005).

Dibbell J, ‘We Pledge Allegiance to the Penguin’ Wired Magazine.:

http://www.wired.com/wired/archive/12.11/linux_pr.html

(11 November 2004).

DiBona C, Ockman S and Stone M, ‘Chapter 1: Introduction’ in Di Bona C & Ockman S (eds) Open Sources: Voices from the Open Source Revolution, O’Reilly & Associates, 1999.

Drjya MJ, ‘Looking to the Changing Nature of Software for Clues to Its Protection’

http://www.dryjapat.com/look.htm

(3 January 2005).

EuroLinux Alliance:

http://petition.eurolinux.org/signatures.html

(11 January 2005).

Fisher WW, ‘The Growth of Intellectual Property: A History of the Ownership of Ideas in the United States’ in Eigentum im internationalen Vergleich, Vandenhoeck & Ruprecht, 1999, 265-91:

http://cyber.law.harvard.edu/people/tfisher/iphistory.pdf

(2 November 2004).

Fisher WW, ‘Intellectual Property and Innovation: Theoretical, Empirical, and Historical Perspectives’ (2001) 37 Industrial Property, Innovation, and the Knowledge-based Economy, Beleidsstudies Technologie Economie:

http://cyber.law.harvard.edu/people/tfisher/Innovation.pdf

(2 November 2004).

Forsyth M, ‘Biotechnology, Patents and Public Policy: A Proposal for Reform in Australia’ (2000) 11 AIPJ 202.

Foundation for a Free Information Infrastructure, ‘Viterbi: 1 Algorithm, 4470 Patents’:

http://swpat.ffii.org/patents/effects/viterbi/index.en.html

(5 January 2005).

Free Software Foundation, ‘GNU General Public License’ 16 February 1998:

http://www.gnu.org/copyleft/gpl.html

(3 January 2005).

Garfinkel SL, Stallman RM and Kapor M, ‘Why Patents are Bad for Software’ (1991) Issues in Science and Technology:

http://lpf.ai.mit.edu/Links/prep.ai.mit.edu/issues.article

(11 November 2004).

Gates W, ‘Challenges and Strategy’:

http://www.bralyn.net/etext/literature/bill.gates/challenges-strategy.txt

(10 March 2005).

Gleick J, ‘Patently Absurd’ New York Times Magazine, 12 March, 2000:

http://www.nytimes.com/library/magazine/home/20000312mag-patents.html

(11 November 2004).

Hamm S, ‘Mozilla is Gaining on Godzilla’ BusinessWeek Online, 12 January, 2005:

http://www.businessweek.com/technology/content/jan2005/tc20050112_0827_tc119.htm

(14 January 2005).

Henkel J and Tins M, ‘Munich/MIT Survey: Development of Embedded Linux’ Institute for Innovation Research, Technology Management and Entrepreneurship, University of Munich, 10 May 2004:

http://opensource.mit.edu/papers/henkeltins.pdf

(20 January 2005).

Hicks M, ‘Eolas Patent Invalid’ eWeek, 5 March, 2004:

http://www.eweek.com/article2/0,4149,1543839,00.asp

(18 January 2005).

International Business Machines Corporation v Commissioner of Patents [1991] FCA 625; (1991) 105 ALR 388:

http://www.austlii.edu.au/au/cases/cth/federal%5fct/unrep5261.html

(2 November 2004).

IP Australia, ‘Grants of Standard Patents by Technology Group 1992-2003’:

http://www.ipaustralia.gov.au/pdfs/statistics/Standard%20Grants%20by%20Tech%2092-03.xls

(12 January 2005).

Irlam G and Williams R, ‘Software Patents: An Industry at Risk’ Submission to the Patent and Trademark Office on Patent Protection for Software-Related Inventions

http://lpf.ai.mit.edu/Patents/industry-at-risk.html

(5 January 2005).

Jaques R, ‘RIAA Launches P2P file sharing legal blitz’ vnunet.com 19 November 2004:

http://www.vnunet.com/news/1159534

(20 January 2005).

Kahin B, ‘Why Europe should be wary of software patents’ (2003) Managing Intellectual Property:

http://www.si.umich.edu/~kahin/mip.html

(13 January 2005).

Kenwood CA, ‘A Business Case Study of Open Source Software’ October 2001

http://www.mitre.org/work/tech_papers/tech_papers_01/kenwood_software/index.html

(11 November 2004).

Kerneltrap.org, ‘Interview: Richard Stallman’:

http://kerneltrap.org/node/4484

(4 January 2005).

League for Programming Freedom, ‘Against Software Patents’ 28 February 1991:

http://lpf.ai.mit.edu/Patents/against-software-patents.html

(11 November 2004).

League for Programming Freedom, ‘Unisys/Compuserve GIF Controversy’:

http://lpf.ai.mit.edu/Patents/Gif/Gif.html

(11 November 2004).

Lessig L, ‘Introduction’ in Free Software, Free Society, Selected Essays of Richard M. Stallman, GNU Press, Boston, 2002.

Lessig L, ‘The Architecture of Innovation’ (2002) 51 Duke Law Journal 1783.

Lohr S, ‘I.B.M. to Give Free Access to 500 Patents’ New York Times, 11 January, 2005:

http://www.nytimes.com/2005/01/11/technology/11soft.html

(12 January 2005).

Loughlan P, ‘Patents: Breaking into the Loop’ [1998] SydLawRw 24; (1998) 20 Sydney Law Review 553.

LPS Group, ‘Market Trends’:

http://www.informationholdings.com/lps/market.html

(18 January 2005).

Marson I, ‘Software patent directive adopted’ ZDNet UK, 8 March 2003:

http://news.zdnet.co.uk/0,39020330,39190497,00.htm

(8 March 2005).

McKeough J, Stewart A and Griffith P, Intellectual Property in Australia, 3rd ed, LexisNexis Butterworths, Sydney, 2004.

McKusick MK, ‘Twenty Years of Berkeley Unix: From AT&T-Owned to Freely Redistributable’ in Di Bona C & Ockman S (eds) Open Sources: Voices from the Open Source Revolution, O’Reilly & Associates, 1999.

Merges RP, ‘As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform’ 6 September 1999 Social Science Research Network:

http://ssrn.com/abstract=180748

(11 November 2004).

Miller R, ‘US to EU on Software Patents: ‘We sold out, you should too’’ Newsforge, 9 June, 2004:

http://trends.newsforge.com/trends/04/06/09/1447245.shtml

(5 January 2005).

‘Monopolies of the Mind’ Economist, 11 November, 2004:

http://economist.com/opinion/displayStory.cfm?story_id=3376181

(11 November 2004).

Morgan JV, ‘Chaining Open Source Software: The Case Against Software Patents’:

http://lpf.ai.mit.edu/Patents/chaining-oss.html

(11 November 2004).

Nicol D and Nielsen J, Patents and Medical Biotechnology: An Empirical Analysis of Issues Facing the Australian Industry (2003) Centre for Law and Genetics Occasional Paper No. 6.

Novell Corporation, ‘Novell Statement on Patents and Open Source Software’:

http://www.novell.com/company/policies/patent/

(5 January 2005).

Nuffield Council on Bioethics, ‘The ethics of patenting DNA’ (2002) Nuffield Council on Bioethics:

http://www.nuffieldbioethics.org

(2 November 2004).

Olexy PD, ‘Avoiding Willful Infringement And Increased Damages: An Overview of Selected Decisions’:

http://www.sughrue.com/clientfiles/aviod_wi(4).htm

(13 January 2005).

Open Source Initiative, ‘Halloween Document II (Version 1.4)’:

http://www.opensource.org/halloween/halloween2.php

(14 January 2005).

Open Source Initiative, ‘The Open Source Definition’:

http://www.opensource.org/docs/definition.php

(14 January 2005).

O’Reilly T, ‘An Open Letter to Jeff Bezos’ O’Reilly Network, 28 February, 2000:

http://www.oreilly.com/cgi-bin/amazon_patent. comments.pl

(12 January 2005).

O’Reilly T, ‘Ask Tim: Open Source, Patents & O’Reilly’ O’Reilly Network, 28 February, 2000:

http://www.oreilly.com/pub/a/oreilly/ask_tim/2000/amazon_patent.html

(22 November 2004).

O’Reilly T, ‘My Conversation with Jeff Bezos’ O’Reilly Network, 2 March, 2000:

http://www.oreilly.com/pub/a/oreilly/ask_tim/2000/bezos_0300.html

(11 November 2004).

O’Reilly T, ‘Jeff Bezos/Tim O'Reilly on Patent Reform’ O’Reilly Network:

http://tim.oreilly.com/patents/PTO_long.html

(11 November 2004).

O’Reilly T, ‘The Internet Patent Land Grab’ O’Reilly Network:

http://tim.oreilly.com/patents/cacm3.html

(11 November 2004).

O’Reilly T and Stallman R, ‘Stallman on the Harm of Patents: An Exchange Between Richard Stallman and Tim O'Reilly’ O’Reilly Network, 13 March, 2000:

http://www.oreillynet.com/pub/a/policy/2000/03/09/stallman.html

(11 November 2004).

O’Rourke MA, ‘Toward a Doctrine of Fair Use in Patent Law’ (2000) 100 Columbia Law Review 1177.

Patents Act 1990 (Commonwealth).

Perelman M, ‘The Political Economy of Intellectual Property’ (2003) 54(8) Monthly Review:

http://www.monthlyreview.org/0103perelman.htm

(6 November 2004).

Perens B, ‘Software Patents vs. Free Software’:

http://perens.com/Articles/Patents.html

(15 January 2005).

Pfaffenberger B, ‘The Coming Software Patent Crisis: Can Linux Survive?’ Linux Journal, 9 August, 1999:

http://www.linuxjournal.com/node/5079

(11 November 2004).

Public Patent Foundation, ‘At PubPat’s Request, Patent Office Rejects Microsoft’s FAT Patent’ 30 September, 2004:

http://pubpat.org/Microsoft_517_Rejected.htm

(19 January 2004).

Ravicher D, ‘Position Paper: Mitigating Linux Patent Risk’ Open Source Risk Management 2 August 2004:

http://www.osriskmanagement.com/pdf_articles/linuxpatentpaper.pdf

(3 January 2005).

Reichman JH, ‘Of Green Tulips and Legal Kudzu: Repackaging Rights in Subpatentable Innovation’ (2000) 53 Vanderbilt Law Review 1753.

Reynolds R and Stoianoff N, Intellectual Property: Text and Essential Cases, Federation Press, Sydney, 2003.

Roberts R, ‘Why Linux is Wealthier than Microsoft’ BusinessWeek Online, 19 November, 2003:

http://yahoo.businessweek.com/technology/content/nov2003/tc20031119_9737.htm

(18 January 2005).

Samuelson P, ‘Toward a New Politics of Intellectual Property’ Paper presented at World Wide Web 2002:

http://www.sims.berkeley.edu/~pam/papers/CACMNewPolitics3.pdf

(2 November 2004).

Samuelson P, Davis R, Kapor M and Reichman J, ‘A Manifesto Concerning the Legal Protection of Computer Programs’ (1994) 94 Columbia Law Review 2308.

Sauer G, ‘BT’s ‘Hyperlinking’ Patent Litigation Fails’ (2003) 51 NSW Society for Computers and the Law Journal:

http://www.nswscl.org.au/journal/51/Glen_Sauer.html

(18 January 2005).

Sol S, ‘From the Beginning’ Web Developer’s Virtual Library, 19 July, 1998:

http://wdvl.com/Internet/UNIX/Intro/history.html

(22 March 2005).

Stallman R, ‘The Anatomy of a Trivial Patent’:

http://lpf.ai.mit.edu/Patents/anatomy-trivial-patent.txt

(11 November 2004).

Stallman R, ‘The Danger of Software Patents’ 2004 Cyberspace Law and Policy Seminar (audio recording), Sydney, 14 October 2004:

http://images.indymedia.org/imc/sydney/stallman%20lo.ogg

(11 November 2004).

Stallman R, ‘The Danger of Software Patents’. Speech at Cambridge University 25 March 2002:

http://lpf.ai.mit.edu/Patents/danger-of-software-patents.txt

(11 November 2004).

‘There is no Moore’s Law for Software’:

http://www.geniebusters.org/15_Mooreslaw.htm

(14 January 2005).

Torvalds L & Cox A, ‘Open Letter on Software Patents from Linux Developers’ 21 September, 2003:

http://www.effi.org/patentit/patents_torvalds_cox.html

(15 January 2005).

Tuomi I, ‘The Lives and Death of Moore’s Law’ (2002) 7(11) FirstMonday:

http://www.firstmonday.dk/issues/issue7_11/tuomi/

(14 January 2005).

Unisys Corporation, ‘LZW Patent Information’:

http://www.unisys.com/about__unisys/lzw

(15 January 2005).

United States Department of Justice, ‘United States v. Microsoft: Current Case’:

http://www.usdoj.gov/atr/cases/ms_index.htm

(14 January 2005)

Wearden G, ‘Serious setback for European software patents’ ZDNet UK, 17 November, 2004:

http://uk.builder.com/manage/business/0,39026582,39224199,00.htm

(5 January 2005).

Weiser PJ, ‘The Internet, Innovation and Intellectual Property Policy’ (2003) 103 Columbia Law Review 534.

Wheeler D, ‘Why Open Source Software/Free Software? Look at the Numbers!’ 7 November 2004:

http://www.dwheeler.com/oss_fs_why.html

(14 January 2005).

Wikipedia, ‘Application Programming Interface’:

http://en.wikipedia.org/wiki/Application_Programming_Interface

(12 January 2005).

Wikipedia, ‘Code Forking’:

http://en.wikipedia.org/wiki/Code_forking

(18 January 2005).

Wikipedia, ‘Forums’:

http://en.wikipedia.org/wiki/Forums

(12 January 2005).

Wikipedia, ‘Moore’s Law’:

http://en.wikipedia.org/wiki/Moore%27s_law

(14 January 2005).

Wikipedia, ‘Version Control’:

http://en.wikipedia.org/wiki/Version_control

(14 January 2005).

Williams P, ‘Patent threat to open source is limited’ vnunet.com, 8 September, 2004:

http://www.pcw.co.uk/news/1157906

(5 January 2005).


[*] BE (Computer Systems), LLB. The author is currently studying towards an LLM at the University of Tasmania.

[1] DiBona C, Ockman S and Stone M, ‘Chapter 1: Introduction’ in Di Bona C & Ockman S (eds) Open Sources: Voices from the Open Source Revolution, O’Reilly & Associates, 1999.

[2] Merton and others identified the norm of communalism as being of central importance to the scientific community, as early as the 1950s. See for example, Merton R, ‘The Normative Structure of Science’ in The Sociology of Science (University of Chicago Press, 1973). cited in Burk DL, ‘Cyberlaw and the Norms of Science’ (1999) Intellectual Property and Technology Forum:

http://infoeagle.bc.edu/bc_org/avp/law/st_org/iptf/commentary/content/burk.html

(18 March 2005)

[3] Sol S, ‘From the Beginning’, Web Developer’s Virtual Library, 19 July, 1998:

http://wdvl.com/Internet/UNIX/Intro/history.html

(22 March 2005)

[4] The BSD distribution continues to this day in various forms, such as the FreeBSD, NetBSD and OpenBSD projects. FreeBSD forms the basis of Apple’s Macintosh OSX operating system.

[5] For a more detailed history of the BSD project, see McKusick MK, ‘Twenty Years of Berkeley Unix: From AT&T-Owned to Freely Redistributable’ in Di Bona C & Ockman S (eds) Open Sources: Voices from the Open Source Revolution, (O’Reilly & Associates, 1999).

[6] National Research and Development Council v Commissioner of Patents [1959] HCA 67; (1959) 102 CLR 252.

[7] Reynolds R and Stoianoff N, Intellectual Property: Text and Essential Cases, (Federation Press, Sydney, 2003), at 268, fn 46.

[8] See Cohen JE and Lemley MA, ‘Patent Scope and Innovation in the Software Industry’ (2001) 89 California Law Review 1:

http://www.law.georgetown.edu/faculty/jec/softwarepatentscope.pdf (2 November 2004) at 8.

[9] Fisher WW, ‘Intellectual Property and Innovation: Theoretical, Empirical, and Historical Perspectives’ (2001) 37 Industrial Property, Innovation, and the Knowledge-based Economy, (Beleidsstudies Technologie Economie):

http://cyber.law.harvard.edu/people/tfisher/Innovation.pdf

(2 November 2004) at 23.

[10] A necessary step to understand the way the software works.

[11] Above n 9, at 23.

[12] See Reichman JH, ‘Of Green Tulips and Legal Kudzu: Repackaging Rights in Subpatentable Innovation’ (2000) 53 Vanderbilt Law Review 1753.

[13] Samuelson P, Davis R, Kapor M, and Reichman J, ‘A Manifesto Concerning the Legal Protection of Computer Programs’ (1994) 94 Columbia Law Review 2308, at 2316.

[14] Apple Computer Inc v Computer Edge Pty Ltd [1947] ArgusLawRp 3; (1984) 53 ALR 225.

[15] Ibid, at 184 per Gibbs CJ.

[16] Hollinrake v Truswell [1894] UKLawRpCh 158; [1894] 3 Ch 420 at 424-425.

[17] Ibid.

[18] Samuelson et al, above n 13 at 2315.

[19] Ibid, at 2318.

[20] There is a vast body of literature on the relationship of copyright and software which it is beyond the scope of this paper to consider. For an excellent account discussion of the strengths and weaknesses of copyright protection for software, see Samuelson et al, above n13.

[21] [1972] USSC 223; 409 U.S. 63 (1972).

[22] Cohen & Lemley, above n 8 at 8.

[23] Ibid.

[24] [1981] USSC 40; 450 U.S. 175 (1981).

[25] Fisher WW, ‘The Growth of Intellectual Property: A History of the Ownership of Ideas in the United States’ in Eigentum im internationalen Vergleich, (Vandenhoeck & Ruprecht, 1999), 265-91:

http://cyber.law.harvard.edu/people/tfisher/iphistory.pdf

(2 November 2004) at 270.

[26] Cohen & Lemley, above n 8 at 8.

[27] In re Freeman 573 F.2d 1237 (C.C.P.A. 1978); In re Walter 618 F.2d 758 (C.C.P.A. 1980); In re Abele 684 F.2d 902 (C.C.P.A. 1982).

[28] Drjya MJ, ‘Looking to the Changing Nature of Software for Clues to Its Protection’:

http://www.dryjapat.com/look.htm

(3 January 2005) at [40].

[29] Cohen & Lemley, above n 8 at 9.

[30] Australian Patent Office, Guidelines for Considering the Patentability of Computer Program Related Inventions, 1986. Cited in Reynolds & Stoianoff, above n 7 at 268.

[31] [1991] FCA 625; (1990) 105 ALR 388.

[32] Ibid at 395.

[33] Ibid at [14].

[34] Above n 6 at [16].

[35] CCOM Pty Ltd and Another v. Jiejing Pty Ltd and Others [1994] FCA 396; (1994) 122 ALR 417:

http://www.austlii.edu.au/au/cases/cth/federal_ct/unrep6887.html (2 Nov 2004).

[36] Ibid, at [100], [106], [128].

[37] Australian Patent Office, Manual of Practice and Procedure:

http://www.ipaustralia.gov.au/resources/manuals_patents2.shtml

(2 November 2004) at 8.2.7.2.

[38] Ibid, at 8.2.7.3.

[39] Ibid, at 8.2.7.4.-8.2.7.6.

>[40] APO figures show that the number of patents awarded in the information technology field have increased from around 200 per year in the years 1994-1997 to 597 in 2003. See IP Australia, ‘Grants of Standard Patents by Technology Group 1992-2003’:

http://www.ipaustralia.gov.au/pdfs/statistics/Standard%20Grants%20by%20Tech%2092-03.xls

(12 January 2005).

[41] See Unisys Corporation, ‘LZW Patent Information’: http://www.unisys.com/about__unisys/lzw

(15 January 2005) for more information on this patent. Although the original patent has expired, Unisys have patents pending for alleged improvements to the algorithm. For more on the controversy surrounding this patent, see League for Programming Freedom, ‘Unisys/Compuserve GIF Controversy’:

http://lpf.ai.mit.edu/Patents/Gif/Gif.html

(11 November 2004)

[42] Stallman R, ‘The Danger of Software Patents’ Speech at Cambridge University 25 March 2002:

http://lpf.ai.mit.edu/Patents/danger-of-software-patents.txt

(11 November 2004).

[43] Foundation for a Free Information Infrastructure, ‘Viterbi: 1 Algorithm, 4470 Patents’:

http://swpat.ffii.org/patents/effects/viterbi/index.en.html

(12 January 2005)

[44] Ibid.

[45] A search was made on 22 March 2005, using the search page provided by IP Australia at:

http://apa.hpa.com.au:8080/ipapa/qsearch?

A total of 54 results were returned, the most recent being application 2003255971. This application, filed by Koninklijke Philips Electronics N.V. is entitled ‘Viterbi Bit Detection Method and Advice’ and has a priority date of 26 September 2002.

[46] [1998] USCAFED 107; 149 F.3d 1368 (Fed. Cir. 1998), cert. Denied, 525 U.S. 1093 (1999) at 1373. See the discussion of the case in see Cohen & Lemley, above n8 at 10.

[47] Christie A and Syme S, ‘Patents for Algorithms in Australia’ (1998) 20 Sydney Law Review 23 at 527.

[48] (1999) AIPC 91-154.

[49] See APO, Manual, above n37 at 8.2.7.3.

[50] See the discussion in section 1.3 above.

[51] Fisher, above n 9 at 24-25.

[52] Ibid, at 25.

[53] Commonwealth of Australia, ‘Patents Bill 1990: Second Reading’ Senate, 29 May, 1990’:

http://parlinfoweb.aph.gov.au/PIWeb/view_document.aspx?id=562046&table=HANSARDS

(2 November 2004).

[54] Loughlan P, ‘Patents: Breaking into the Loop’ [1998] SydLawRw 24; (1998) 20 Sydney Law Review 553 at 567.

[55] Bessen J and Maskin E, ‘Sequential Innovation, Patents and Imitation’ (2000) MIT Working Paper:

http://www.researchoninnovation.org/patent.pdf

(11 November 2004) at 3.

[56] Boyle J, ‘Enclosing the Genome: What the Squabbles over Genetic Patents Could Teach Us’ To be published in F. Scott Kieff (ed), Perspectives on the Human Genome Project:

http://www.law.duke.edu/ip/pdf/enclosing.pdf

(1 September 2003) at 9.

[57] Advisory Council on Intellectual Property, ‘Patents and Experimental Use’ (2004) Options Paper. at 25 per submissions of Epitan Ltd and GTG Technologies.

[58] Ibid, at 24 per Eisenberg.

[59] Boyle, above n 56 at 13.

[60] See Loughlan, above n 54 at 555-559.

[61] Boyle, above n 50 at 11.

[62] Ibid at 9.

[63] One can’t help but notice the irony of such an argument, namely, that the commoner won’t understand the commons. On the political reach of the maximalist camp see Perelman M, ‘The Political Economy of Intellectual Property’ (2003) 54(8) Monthly Review

http://www.monthlyreview.org/0103perelman.htm

(6 November 2004).

[64] Loughlan, above n 54 at 553.

[65] Ibid.

[66] See Wearden G, ‘Serious setback for European software patents’ ZDNet UK, 17 November, 2004

http://uk.builder.com/manage/business/0,39026582,39224199,00.htm (5 January 2005)

[67] EuroLinux Alliance:

http://petition.eurolinux.org/signatures.html

(11 January 2005).

[68] Boyle, above n 56 at 13-15.

[69] Ibid, at 15.

[70] Fisher, above n 9 at 2-3.

[71] Ibid, at 4.

[72] Loughlan, above n 54 at 569.

[73] Nuffield Council on Bioethics, ‘The ethics of patenting DNA’ (2002) Nuffield Council on Bioethics:

http://www.nuffieldbioethics.org

(2 November 2004) at 12.

[74] Emphasis added.

[75] See Burk DL and Lemley MA, ‘Is Patent Law Technology-Specific?’ (2002) 17 Berkeley Tech. Law Journal 1155:

http://ssrn.com/abstract=349761

(2 November 2004) at 8-9.

[76] Australian Law Reform Commission, ‘Genes and Ingenuity: Gene Patenting and Human Health’ (2004) ALRC 99:

http://www.austlii.edu.au/au/other/alrc/publications/reports/99/ (2 November 2004)

[77] [1994] FCA 1065; (1994) AIPC 91-076, 28 IPR 383.

[78] Bristol-Myers Squibb Co. v F H Faulding & Co. Ltd. [2000] FCA 316; (2000) 170 ALR 439.

[79] Ibid, at 472.

[80] See Forsyth M, ‘Biotechnology, Patents and Public Policy: A Proposal for Reform in Australia’ (2000) 11 AIPJ 202.

[81] Ibid, at 212.

[82] See below.

[83] Samuelson et al, above n13 at 2330-2331.

[84] O’Reilly T, ‘An Open Letter to Jeff Bezos’ O’Reilly Network, 28 February, 2000:

http://www.oreilly.com/cgi-bin/amazon_patent.comments.pl

(12 January 2005). This letter encapsulates many of the objections to the patent and became a petition which stopped accepting signatures after the number of signatories reached 10,000.

[85] Barton-Davis P, ‘I Oppose Amazon.com’s 1-Click Patent’:

http://www.equalarea.com/paul/amazon-1click.html

(12 January 2005).

[86] Discussed below.

[87] See NRDC, above n 6 at 261-262.

[88] See Fisher, above n 9 at 18.

[89] Loughlan, above n 54 at 571.

[90] Amazon obtained a preliminary injunction against Barnes & Noble which forced the latter to add another click to their ‘Express Lane’ ordering system. See Amazon.com Inc v BarnesAndNoble.com Inc 73 F Supp 2d 1228 (US District 1999).

[91] Bessen, above n 55 at 3-4.

[92] Ibid, at 4, 5.

[93] These are defined as follows: ‘An application programming interface (API) is a set of definitions of the ways in which one piece of computer software communicates with another. It is a method of achieving abstraction, usually (but not necessarily) between lower-level and higher-level software’. See Wikipedia, ‘Application Programming Interface’

http://en.wikipedia.org/wiki/Application_Programming_Interface (12 January 2005).

[94] See Cohen J, ‘Reverse Engineering and the Rise of Electronic Vigilantism: Intellectual Property Implications of ‘Lock-Out’ Technologies’ (1995) 68 California Law Review 1091 at 1178. Cited in Burk & Lemley, above n75 at 17.

[95] See Bicknell C, ‘Cashing in on Bogus Patents’ Wired News, 19 October, 2000:

http://www.wired.com/news/business/0,1367,39528-2,00.html

(12 January 2005).

[96] Fisher, above n 9 at 25.

[97] Microsoft’s strategy of only making Word available within the Microsoft Office bundle has ensured that the other bundled products, Excel and Powerpoint, are de facto standards in their markets.

[98] For a discussion of the reasons why, see Lessig L, ‘The Architecture of Innovation’ (2002) 51 Duke Law Journal 1783 at 1789.

[99] Technically, use of the product would fall within the scope of the patent holder’s right to exploit the invention. It is assumed however that a patent holder is unlikely to sue its own customers.

[100] For a detailed analysis of the status of reverse engineering in the US, see Weiser PJ, ‘The Internet, Innovation and Intellectual Property Policy’ (2003) 103 Columbia Law Review 534 at 552-568. In support of a copyright-like fair use defence for patent law, see O’Rourke MA, ‘Toward a Doctrine of Fair Use in Patent Law’ (2000) 100 Columbia Law Review 1177.

[101] S47D.

[102] See the discussion of the Linux project below, which was found to potentially infringe 283 patents.

[103] This point was originally put forward by Mandeville et al in relation to patents generally, although it can no longer be considered a statement of general authority. (For empirical evidence that this is no longer the case in the biotech industry see Nicol D and Nielsen J, Patents and Medical Biotechnology: An Empirical Analysis of Issues Facing the Australian Industry (2003) Centre for Law and Genetics Occasional Paper No. 6). Stallman and Kahin have recently echoed Mandeville’s claim in relation to the software industry. See Mandeville, T D, Lamberton, D M, Bishop, E J, Economic Effects of the Australian Patent System (1982) Australian Government Publishing Service, Canberra at 213. Cited in Loughlan, above n54. See also Stallman R, ‘The Danger of Software Patents’ 2004 Cyberspace Law and Policy Seminar (audio recording), Sydney, 14 October 2004:

http://images.indymedia.org/imc/sydney/stallman%20lo.ogg

(11 November 2004); Kahin B, ‘Why Europe should be wary of software patents’ (2003) Managing Intellectual Property:

http://www.si.umich.edu/~kahin/mip.html

(13 January 2005).

[104] 35 USC § 284. For an overview of the decisions on wilful infringement, see Olexy PD, ‘Avoiding Willful Infringement And Increased Damages: An Overview of Selected Decisions’:

http://www.sughrue.com/clientfiles/aviod_wi(4).htm

(13 January 2005).

[105] Section 123(1) Patents Act 1990 (Cth).

[106] In relation to patent searching practices in the biotechnology industry see Nicol D and Nielsen J, ‘Results Chapter 5: Anticommons Issues’ in Patents and Medical Biotechnology, above n103.

[107] This is the well-documented ‘anticommons’ effect. See O’Rourke, above n100, at 1179.

[108] Carroll J, ‘Software patents need shelter from the storm’ ZDNet UK, 17 September, 2003:

http://comment.zdnet.co.uk/other/0,39020682,39116437,00.htm

(5 January 2005).

[109] Ibid.

[110] See Samuelson et al, above n 13 at 2377.

[111] For an overview see Wikipedia, ‘Moore’s Law’

http://en.wikipedia.org/wiki/Moore%27s_law

(14 January 2005). For a more in-depth analysis, see Tuomi I, ‘The Lives and Death of Moore’s Law’ (2002) 7(11) FirstMonday:

http://www.firstmonday.dk/issues/issue7_11/tuomi/

(14th January 2005)

[112] See ‘There is no Moore’s Law for Software’:

http://www.geniebusters.org/15_Mooreslaw.htm

(14th January 2005).

[113] See for example Bezos J, ‘An Open Letter from Jeff Bezos on the Subject of Patents’ O’Reilly Network, 3 September, 2000:

http://www.oreilly.com/news/amazon_patents.html

(18 January 2005).

[114] The British Telecom Sargent patent, discussed below, is a classic example of this sort of opportunistic enforcement of a dinosaur patent.

[115] Open Source Initiative, ‘The Open Source Definition’

http://www.opensource.org/docs/definition.php

(14 January 2005).

[116] Barr J, ‘Live and let license’ Linuxworld.com, 23 May, 2001:

http://www.itworld.com/AppDev/350/LWD010523vcontrol4/

(14 January 2005).

[117] Version (or revision) control systems record the history changes to files such as source code and project documentation. See Wikipedia, ‘Version Control’

http://en.wikipedia.org/wiki/Version_control

(14 January 2005) for more information and examples of common version control software.

[118] Bug trackers are used to manage the reporting and resolution of errors in computer software. See for example:

http://www.bugzilla.org

[119] Mailing lists are a system which allow for widespread distribution of email to many Internet users. Typical OSS projects archive mailing list discussions in web-searchable format. See for example:

http://mailman.sourceforge.net

[120] A web application that allows for online discussion. See Wikipedia, ‘Forums’:

http://en.wikipedia.org/wiki/Forums

(12 January 2005) for more information.

[121] A weblog (or blog) is an online journal or diary. Software developers on OSS projects often document their work through this medium. Blogs of individuals are sometimes aggregated together to form a ‘blog planet’. For an example relating to the Mozilla project, see:

http://planet.mozilla.org

[122] Bessen J, ‘Open Source Software: Free Provision of Complex Public Goods’

http://www.researchoninnovation.org/opensrc.pdf

(11 November 2004) at 21.

[123] David Wheeler has combined many such sources together into one document which also includes some discussion of non-quantitative measures. See Wheeler D, ‘Why Open Source Software/Free Software? Look at the Numbers!’ 7 November 2004:

http://www.dwheeler.com/oss_fs_why.html

(14 January 2005).

[124] This browser war formed the basis of the high-profile antitrust suit against Microsoft by the US Department of Justice. See United States Department of Justice, ‘United States v. Microsoft: Current Case’:

http://www.usdoj.gov/atr/cases/ms_index.htm

(14 January 2005) for information relating to the case.

[125] Hamm S, ‘Mozilla is Gaining on Godzilla’ BusinessWeek Online, 12 January, 2005:

http://www.businessweek.com/technology/content/jan2005/tc20050112_0827_tc119.htm

(14 January 2005).

[126] Part of the appeal of Firefox has been the new features it offers to users such as tabbed browsing and popup ad blocking. See:

http://www.mozilla.org/products/firefox

[127] Lessig, above n 98 at 1789.

[128] Ibid, at 1788-1790.

[129] Such is the nature of the Internet where open protocols run on proprietary hardware, see Lessig, above n 98 at 1790.

[130] See Bessen J, above n 122 for an economics-heavy proof of this proposition.

[131] The author bases this proposition on his personal experience with the Zope project community (http://zope.org/). Empirical evidence gathered in relation to Embedded Linux also supports this view, with 139 of 259 survey participants working for organisations with less than 50 people, and 81 participants working for companies with 200 or more staff. See Henkel J and Tins M, ‘Munich/MIT Survey: Development of Embedded Linux’ Institute for Innovation Research, Technology Management and Entrepreneurship, University of Munich, 10 May 2004

http://opensource.mit.edu/papers/henkeltins.pdf

(20 January 2005) at 7.

[132] Roberts R, ‘Why Linux is Wealthier than Microsoft’ BusinessWeek Online, 19 November, 2003

http://yahoo.businessweek.com/technology/content/nov2003/tc20031119_9737.htm

(18 January 2005).

[133] Code forking is a consequence of the liberal modification and redistribution rights given in OSS licences. If any one party is unhappy with the direction a project is taking, they can take a copy of the source code and use it to launch their own project. See Wikipedia, ‘Code Forking’

http://en.wikipedia.org/wiki/Code_forking

(18 January 2005).

[134] For example, Linus Torvalds is the project leader for the Linux kernel project.

[135] For example the Apache Software Foundation (see <http://apache. org> ) or the Mozilla Foundation (see <http://mozilla.org/> ) host the project website, version control and mailing lists which allow project collaboration in addition to holding IP rights in the project code.

[136] On the legal implications of distributed versus centralised control see Metro-Goldwyn-Mayer Studios, Inc v Grokster, Ltd US District Court Central District of California (25 April 2003).

[137] Williams P, ‘Patent threat to open source is limited’ vnunet.com 8 September, 2004:

http://www.pcw.co.uk/news/1157906

(5 January 2005). On the other hand, the Recording Industry Association of America in the US are certainly having a good shot at pursuing similarly distributed filesharers with copyright infringement suits. See Jaques R, ‘RIAA Launches P2P file sharing legal blitz’ vnunet.com 19 November 2004

http://www.vnunet.com/news/1159534

(20 January 2005).

[138] See for example Babcock C and Greenemeier L, ‘Open Source Stress’ Information Week, 9 August, 2004

http://www.informationweek.com/story/showArticle.jhtml?articleID=26806464

(3 January 2005). Also see Section 5.2 below.

[139] Ravicher D, ‘Position Paper: Mitigating Linux Patent Risk’ Open Source Risk Management 2 August, 2004

http://www.osriskmanagement.com/pdf_articles/linuxpatentpaper.pdf

(3 January 2005) at 1.

[140] This is what people mistakenly refer to as the Linux operating system, which Stallman insists is more correctly called GNU/Linux. See Stallman, above n 97.

[141] Ibid.

[142] For example, the preamble to the GNU General Public Licence has for a decade recognised that “any free program is threatened constantly by software patents”. See Free Software Foundation, ‘GNU General Public License’ 16 February 1998:

http://www.gnu.org/copyleft/gpl.html

(3 January 2005)

[143] See Open Source Initiative, ‘Halloween Document II (Version 1.4)’:

http://www.opensource.org/halloween/halloween2.php

(14 January 2005)

[144] Barr J, ‘HP memo forecasts MS patent attacks on free software’ Newsforge, 19 July, 2004

http://www.newsforge.com/article.pl?sid=04/07/19/2315200

(14 January 2005).

[145] Bangeman E, ‘Will Microsoft sue OpenOffice users?’ ArsTechnica, 16 September, 2004

http://arstechnica.com/news.ars/post/20040916-4192.html

(14 January 2005). OpenOffice.org is a free alternative to Microsoft Office.

[146] In 1999, IBM collected $1.3 billion from licensing, with 40% of that believed to have come from patent licensing. See LPS Group, ‘Market Trends’:

http://www.informationholdings.com/lps/market.html

(18 January 2005).

[147] Cited in Irlam G and Williams R, ‘Software Patents: An Industry at Risk’ Submission to the Patent and Trademark Office on Patent Protection for Software-Related Inventions:

http://lpf.ai.mit.edu/Patents/industry-at-risk.html

(5 January 2005).

[148] Gates W, ‘Challenges and Strategy’:

http://www.bralyn.net/etext/literature/bill.gates/challenges-strategy.txt

(10 March 2005).

[149] Stallman, above n 42.

[150] Fisher undertakes a comparison of the impact of the patent system on four different industries, and concludes that patents work well in the pharmaceutical industry. See Fisher, above n 9.

[151] ALRC, above n 76.

[152] Ibid, at 6.14.

[153] Burk & Lemley, above n 75 at 48.

[154] Ibid, at 23-35. In Australia, the same trend has emerged for non-obviousness, but not for disclosure. This may be attributable to either the much smaller number of Australian cases on the matter, or the generally more liberal standards adopted in the Australian jurisdiction. On the obviousness standard in Australia, see Nicol & Nielsen, above n103 at 365-369.

[155] Burk & Lemley, above n 75 at 48.

[156] Ibid, 7-22.

[157] Ibid, at 52-54.

[158] Ibid, at 54-56.

[159] Ibid, at 56-58.

[160] Ibid, at 58-59.

[161] Ibid, at 6.15.

[162] Loughlan, above n 54 at 577.

[163] Where the prevention of commercial exploitation is necessary to protect ‘ordre public’ or morality, or to avoid serious prejudice to the environment. See Articles 27(2) and (3); Christie & Syme, above n 47 at 544-545.

[164] This interpretation was taken by many submissions to ALRC gene patent Inquiry see above n 76 at Ch 6, fn14.

[165] Christie and Syme, above n 47 at 544-545.

[166] Collins Australian Dictionary and Thesaurus, 3rd Edition, HarperCollins, 2004.

[167] ALRC, above n 76 at 6.18.

[168] McKeough J, Stewart A and Griffith P, Intellectual Property in Australia, 3rd ed, LexisNexis Butterworths, Sydney, 2004 at 269.

[169] Stallman R, above n 42.

[170] Torvalds L & Cox A, ‘Open Letter on Software Patents from Linux Developers’ 21 September, 2003:

http://www.effi.org/patentit/patents_torvalds_cox.html

(15 January 2005).

[171] O’Reilly T, ‘The Internet Patent Land Grab’ O’Reilly Network:

http://tim.oreilly.com/patents/cacm3.html

(11 November 2004).

[172] Perens B, ‘Software Patents vs. Free Software’

http://perens.com/Articles/Patents.html

(15 January 2005).

[173] In fact, the League for Programming Freedom have been arguing for their abolition since at least 1991. See League for Programming Freedom, ‘Against Software Patents’ 28 February 1991:

http://lpf.ai.mit.edu/Patents/against-software-patents.html

(11 November 2004).

[174] This is one of the grounds for exclusion under the Statute of Monopolies 1623 on which our manner of manufacture test relies. See Christie & Syme, above n 47 at 543-544.

[175] Ibid, at 545.

[176] The Freeman-Walter-Abele test was introduced after a 1985 review of the Patent Office Guidelines.

[177] Cohen & Lemley, above n 8 at 4.

[178] At the time of writing, the European Council has passed the Directive, despite requests from a number of member nations not to do so. See Marson I, ‘Software patent directive adopted’ ZDNet UK, 8 March 2003:

http://news.zdnet.co.uk/0,39020330,39190497,00.htm

(8 March 2005)

[179] Miller R, ‘US to EU on Software Patents: ‘We sold out, you should too’’ Newsforge, 9 June, 2004

http://trends.newsforge.com/trends/04/06/09/1447245.shtml

(5 January 2005).

[180] Ibid.

[181] Samuelson et al, above n 13 at 2310.

[182] Ibid.

[183] Ibid, 2415-2419.

[184] Semiconductor Chip Protection Act 1984 US. This legislation formed the basis of the WIPO draft of the 1989 Washington Treaty for which the Australian legislation was passed. Circuit Layout protection is now required by Articles 35-38 of TRIPS. See Copyright Law Review Committee, ‘Chapter 8: Circuit Layouts’ in Copyright and Contract 23 August 2004:

http://sgeag001web.ag.gov.au/agd/www/Clrhome.nsf/0/937143BFE876FAD2CA256C4F000F065F?OpenDocument

(19 January 2005) at 8.03-8.04.

[185] McKeough, above n 168 at 268.

[186] Reichman has elsewhere observed that many of the major problems facing the patent system are due to a failure to protect these small grain-sized innovations. See Reichman J, above n 8.

[187] Samuelson, above n 13 at 2417-2418.

[188] Reichman, above n 12. In fact, some are likely to be better off.

[189] BT’s Sargent Patent (US patent 4,873,662) was filed in 1977 but not awarded until 1989. BT claimed the patent covered the use of hyperlinking and in 2000 began enforcement proceedings against Prodigy. Summary judgement was awarded to Prodigy in 2002. The case brought harsh criticism from the US software industry. British Telecommunications PLC v Prodigy Communications Corporation (2002) 217 F Supp 2d 399; US Dist LEXIS 15521 (22 August 2002). For a good overview of the case see Sauer G, ‘BT’s ‘Hyperlinking’ Patent Litigation Fails’ (2003) 51 NSW Society for Computers and the Law Journal:

http://www.nswscl.org.au/journal/51/Glen_Sauer.html

(18 January 2005).

[190] The Eolas patent (US patent 5,838,906) covers a method for embedding and invoking interactive applications in Web browsers. Eolas was awarded US$521 million in an infringement suit against Microsoft over their ActiveX technology. The public outcry in the wake of this judgement led to a re-examination in which the patent was found invalid on the basis of submitted prior art. The Eolas-Microsoft litigation is ongoing. See Hicks M, ‘Eolas Patent Invalid’ eWeek, 5 March, 2004:

http://www.eweek.com/article2/0,4149,1543839,00.asp

(18 January 2005).

[191] See Burke TP, ‘Software Patent Protection: Debugging the Current System’ (1994) 69 Notre Dame Law Review 1115, 1158-1160: ‘In fostering the trade-off between the interests of inventors and the public, the source code is the best way to explain the algorithm.’ Cited in Burke & Lemley, above n 75, footnote 45 at 13-14.

[192] Toedt DC, ‘Patents for Inventions Utilizing Computer Software: Some Personal Pointers’ (1992) 9(10) Computer Law 12. Cited in Burk & Lemley, above n 69, footnote 45 at 13.

[193] A party wishing to exploit the patented invention may apply under s.133(1) Patents Act 1990 (Cth) to the court for an order requiring the patentee to grant a licence.

[194] Weiser, above n 100.

[195] See O’Rourke, above n 100.

[196] Cohen & Lemley, above n 8.

[197] Ibid, at 28.

[198] Unknown speaker in Stallman, above n 103.

[199] Ibid.

[200] For an overview of the nature of the claims and the current state of litigation, see:

http://sco.iwethey.org/

[201] Taken from the company’s website:

http://osriskmanagement.com/

(5 January 2005)

[202] Williams, above n 137.

[203] Lohr S, ‘I.B.M. to Give Free Access to 500 Patents’ New York Times, 11 January, 2005

http://www.nytimes.com/2005/01/11/technology/11soft.html

(12 January 2005).

[204] See for example, Novell Corporation, ‘Novell Statement on Patents and Open Source Software’:

http://www.novell.com/company/policies/patent/

(5 January 2005).

[205] Stallman, above n 103.

[206] See <http://pubpat.org/> (18 January 2005).

[207] Public Patent Foundation, ‘At PubPat’s Request, Patent Office Rejects Microsoft’s FAT Patent’ 30 September, 2004:

http://pubpat.org/Microsoft_517_Rejected.htm

(19 January 2004).

[208] Stallman, above n103.


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/journals/JlLawInfoSci/2003/3.html