James Cook University Law Review
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
James Cook University Law Review |
|
DR A. G. PLACE∗
Abstract
The rapid evolution of software within the last 30 years has seen it move from a method of calculating mathematical expression to a technology in its own right that has permeated almost every technology and economic endeavour. For this reason courts have justified the patenting of software. Under current patent law, the system is particularly sensitive to control and interpretation, and must maintain a delicate balance between inventors’ rights and public access. As the assessment of novelty, inventive step and utility require identifying the field of technology, establishing the common general knowledge and constructing a hypothetical skilled addressee, there is an inseparable nexus between the patenting of software-related inventions and the nature of software and its designers. This paper provides a legal and technical perspective on the issues of software patents. The basis of the paper is that while software has evolved to become patentable, it is still rapidly evolving, which is putting tension on current interpretations and ‘traditional principles’.
The software industry is a rapidly growing industry that has positioned itself within almost every other field of economic endeavour. For this reason, software patents cause a lot of confusion in the technical and legal professions alike. The complexity and misunderstanding associated with software patents can be attributed to software’s rapid evolution from a scientific and mathematical calculator to a pervasive product that has permeated almost every aspect of modern technology (and life).
The patent system is an evolving process which is directed by public policy, legislation and judicial interpretations. This evolution is observed in the European Commission’s proposal for a directive to ensure patent protection for software throughout the EU[1] and within the intellectual property chapter of the recent Australia-United States Free Trade Agreement[2]. The public policy that underpins the patent system is at risk if an adequate balance between inventors’ rights and public access is not maintained.
Particularly in the technical field, software is often considered to be unsuitable for copyright and patent protection. Although it is one area of technology that is afforded both protections, neither is capable of providing adequate protection.
This paper examines the issues of software patents within Australia in light of current legislation, judicial decisions and international trends. The concept of software patents is discussed in relation to historical conventional patents. A critical issue when discussing ‘software patents’ is to recognise that this is not a new class of patent and no new legislation was brought into effect to enable the granting of patent protection to software. As with other technologies, the patentability of software-related inventions must be assessed on ‘traditional principles’.
To adequately understand the history of judicial interpretations, it is advantageous to understand clearly what software is, and how it has evolved in parallel with judicial interpretations. To this end a description of the evolution of software and an outline of some of its important properties is provided.
The relatively recent acceptance of software-related inventions within the scope of patentable subject matter has raised issues relating to patent validity which have not been adequately clarified.
To understand ‘software patents’ first it is necessary to understand software, as legal interpretations and protections provided to software have not suddenly changed, but instead tracked the technological developments of hardware and software.
The invention and development of the integrated circuit in the 1960s set computer hardware and software on a rapid series of technological developments. Initially computers were hard wired to perform particular mathematical functions based on switch settings. These switches were replaced with punched cards that were read in a sequence to provide instructions, but the functions remained limited and the tasks remained mostly mathematical in nature. By 1973 the microprocessor[3], floppy disk and computer networks had been developed, which provided a means to produce a more general purpose computer with greater functionality. To cope with this increased functionality and complexity, Operating Systems (OS) were developed to allow software to control hardware in standard ways and provide greater accessibility to users. By 1977 consumer based computers[4] were on the market, and were quickly followed in 1981 by the International Business Machines Personal Computer (IBM PC) and the Microsoft Disk Operating System (MS DOS). The Apple Macintosh (1983) and Microsoft Windows (1985) further increased the accessibility of computer hardware and software.
In keeping up with the computer platform a significant shift in the dynamics between hardware and software occurred. Computer hardware is focused now on providing a faster and more generic platform, allowing the software to dictate function. Programs now can be written for this more generic machine[5]. Software has developed alongside hardware and can be responsible for determining the system function. It can now be considered equivalent or indistinguishable. Software can be written with unlimited freedom of expression in any programming language to achieve the same outcome, the code does not monopolise its function. The engineering concepts of hardware, firmware and software have continually blurred over time. For dealing with intangible ‘things’ (data or information) the choice to use dedicated hardware, software or a combination of both typically comes down to an engineering decision.
The rapid development of software and its migration, from merely controlling the mathematical functions of hardware, to providing general purpose processes that are integrated into almost every aspect of modern society, has fuelled the question of how software fits into ‘traditional’ systems of protection. Software has notable and unique characteristics that need to be considered when dealing with this question.
Typically, software development is decentralised and relies on a highly layered system from hardware through to user interface with a high proportion of software developed ‘in house’ and never provided to the public. The completed product is often produced by applying the code optimising compiler to produce a non-human readable ‘material form’ which is difficult to associate back to the original work. As the trend for computer hardware is to become generic, the compiled software is mobile[6] and intangible[7], making it difficult to deal with in isolation[8]. This environment of layered development, distributed development and mobile product has given rise to the ‘network effect’ whereby the first to market usually becomes the de facto standard and resulting in significant market power,[9] which must be kept in check.[10]
A significant issue in controlling and protecting software is providing a definition that is flexible enough to recognise these characteristics and that includes the idea that software is more than just the code. The World Intellectual Property Organisation (WIPO) provided an early legal definition of a computer program when dealing with copyright laws,[11]similar to the definition provided by the Copyright Act 1968 (Cth)[12] and the United States Copyright Act[13], but focusing on the code as a series of instructions that perform a particular function. Although copyright is automatically afforded to an expression of code in material form,[14] it does not protect any ideas or inventions used in producing the code. The Patents Act 1990 (Cth) does not attempt to define software as it provides protection to any ‘manner of manufacture’[15] and must include all technologies[16]. The Patents Act’s lack of classification of technology, ie software, and the past exclusion of software from patentability, when combined with the expansive scope of software, have lead to misunderstandings of what is excluded from patentability.
From its beginning, software was considered by the courts and patent offices as unpatentable,[17] and although this cannot be said to have restricted its advancement, there were continuing attempts to patent claimed inventions that consisted of, or included, software. When examining the patentability of software, the material form presented and the real function of the software must be separated.[18]
The precedents leading to the software exclusion from patentability were based on ‘traditional principles’,[19] and focused on the fact that software was not ‘a manner of manufacture’ as no tangible thing was produced. However, it is argued here that these traditional principles applied to the nature and understanding of software at the time and not to software in general.
The following sections examine the historical status of software within patent law, with respect to the artificial time reference of 1977. This reference corresponds to both the first successful instances of patents involving software and the technological release of consumer oriented computers.
Early decisions set the tone for patent law’s initial rejection and ultimate acceptance of software patents. As ‘products’ became more complicated and less tangible, boundaries between patentable subject matter and purely intellectual ideas blurred. Patent law, when dealing with developing technologies, was made to examine how claims were directed to laws of nature and the interpretation of industrial application.
In 1844 the telecommunications industry was launched with the words ‘What hath God wrought?’ during a demonstration of the telegraph[20]. However it was held that a claim ‘to the use of electro-magnetism, however developed’ could not be supported by the specification and that future inventors would be restricted in using the new discovery when applying new techniques.[21] As later applied, this decision allowed non-patentable material to be included in a claim if it was considered adequately connected with the invention.[22]
Australian patent law also expanded what was considered a ‘manner of manufacture’ by accepting a broad interpretation of the Morton J ‘rule’[23] which resulted in the interpretation, provided by the National Research Development Corp (NRDC) decision that:[24]
1. a product is ‘vendible’ if ‘its value to the country is in the field of economic endeavour’,[25] and;
2. a ‘product’ needs only to be something in which ‘an artificially created state of affairs’ may be observed.[26]
The issue of patenting of software-related inventions initially was considered to involve claims that were not statutory subject matter, or encroached on processes that were traditionally reserved as non-patentable.
The Banks Report[27] examined the British Patent system and recommended the exclusion of computer programs[28]. However, during the same period the UK Courts decided that a program as a product was patentable[29] and that a claim directed to a method involving a computer modified through software should be accepted[30].
United States courts at this time were rejecting claims for software related inventions.[31] In Benson[32], the United States Supreme Court considered whether solving mathematical problems to converting one form of numerical representation to another was an abstract intellectual concept. In applying the Morse decision, the Court found that claims that were ‘not limited to any particular art or technology, to any particular apparatus or machinery, or to any particular end use…purported to cover any use of the claimed method in a general-purpose digital computer of any type.’ The Supreme Court held that such a patent ‘would wholly pre-empt the mathematical formula and in practical effect would be a patent on the algorithm itself’.[33] However the Court also stated that uncertainty now existed with attempts to indirectly patent programs by drafting claims as a process, rather than the program itself, and that this should not be permitted.[34] The Supreme Court followed this decision by holding that an application should not be successful if the only difference between a conventional method and that described in an application resided in the mathematical algorithm or formula, which itself was not patentable.[35]
Despite the NRDC decision, the Australian Patent Office’s attitude to software-related inventions continued to adhere to traditional interpretations, leading to the conclusion that software was not appropriate material for letters patent.[36] The Patent Office outlined its objections as follows:[37]
a) Computer programs as a product in material form could not be patented as the material form was not novel;
b) A method of programming a computer to solve a problem was not patentable since it was considered a mathematical solution;
c) Programming in general was considered an advantageous manner of operation and not patentable; and
A computer programmed in a particular way is patentable only if it differed from all known computers.
These reasons reflect the attitude of computers in the day that the hardware was the central component and the software was used to perform relatively simple computations and decisions, as enabled by the hardware. It was even considered public policy that a hardware proprietor should be able to use his machine freely ‘as he thinks fit’.[38]
The Patents Act 1977 (UK) diverged from Australian patent law by expressly excluding the patentability of ‘software as such’.[39] However, the English Court of Appeal, when interpreting the phrase ‘software as such’, found that patentability should not be excluded by the mere fact that it involved a computer program, concluding that the claimed technical result must be considered as a whole and exclusions applied if the technical result was prohibited,[40] thus continuing the patentability of software-related inventions. Patenting in relation to intellectual information also required that if the claim was not categorised solely by an intellectual connotation of matter, the claim as a whole must be assessed.[41]
Despite United States courts allowing claims for inventions involving computer programs through a strict interpretation of Benson, they still maintained that the mere methods of solving mathematical equations in software were not patentable. It was not until 1981 that the United States Supreme Court held that a computer program related invention was patentable.[42] The United States Supreme Court held that the patent for industrial equipment, including software plus machine, was not just an abstract algorithm, and thus was valid subject matter for a patent.[43]
The United States Court of Custom and Patent Appeals (CCPA) developed a test in light of Supreme Court decisions.[44] The test essentially allowed method or process claims for inventions which were implemented by software, provided the claims did not attempt to monopolise or ‘pre-empt’ traditional unpatentable subject matter, such as mathematical algorithms or laws of nature. The CCPA held that if ‘a mathematical algorithm is implemented in a specific manner to define structural relationships between the physical elements of the claim (in apparatus claims) or to refine or limit claim steps (in process claims), the claim being otherwise statutory, the claim passes muster’.[45] The CCPA however maintained that:
[i]f, however, the mathematical algorithm is…presented and solved by the claimed invention,… and is not applied in any manner to physical elements or process steps, no amount of post-solution activity will render the claim statutory; nor is it saved by a preamble merely reciting the field of use of the mathematical algorithm.[46]
This test shows that at this time there was still a presumption that software was used to execute mathematical steps, and that software algorithms could be described in mathematical terms.
Despite these restrictive decisions regarding software as mere mathematical steps, the Court of Appeals for the Federal Circuit ruled that derived numbers, e.g. share prices, could have a real and tangible effect and should be considered to be a valid subject for a patent.[47]
The United States Supreme Court then established the Freeman-Walter-Abele test[48] for claims that included inherently unpatentable subject matter. The test required examining the relationship between the subject matter and the claim as a whole as well as assessing whether the subject matter was essential in its application to the process steps. Clearly distinguishing the Patents Act 1977 (UK), the Australian Patent Office imported the Freeman-Walter-Abele test established by the United States Supreme Court.[49]
The question of whether computer software-related inventions constituted a ‘manner of manufacture’, and therefore patentable subject matter as required by the Patents Act 1990 (Cth)[50], first came before an Australian Federal Court in the IBM case.[51] The Federal Court held that the use of a mathematical formula within a computer used to perform a function was a patentable invention, provided it had a commercially useful effect and met the threshold for novelty and non-obviousness.[52] This prompted the Australian Patent Office (APO) to issue new guidelines for determining the patentability of software-related inventions; the new test simply posed the question: ‘Does the invention claimed involve the production of some commercially useful effect?’[53]
The Full Federal Court affirmed the IBM decision and the 1992 APO Guidelines[54] by finding that a program created to store, retrieve and display Chinese language characters for assembly in text to be used in word processing, by offering utility in achieving retrieval and graphical representations of desired characters in the economic endeavour of word processing, was a ‘manner of manufacture’ and therefore a patentable invention. The Full Federal Court was careful to assert that the Patents Act 1990 (Cth) had not followed the British legislation of 1977.
The APO further revised the Manual of Practice and Procedure[55] to reflect the findings of the IBM and CCOM cases. The APO considers software to be patentable ‘if it includes 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’, but excludes software that is ‘merely a procedure for solving a given type of mathematical problem is not patentable’.[56]
The previous section showed that broadly software can be regarded as patentable subject matter by being considered a ‘manner of manufacture’. Patent validity must therefore reside in the separate heads[57] of novelty, inventiveness and utility[58]. The High Court applied an additional threshold of inventiveness brought through statutory interpretation of the preamble ‘a patentable invention is an invention that…’, continuing the established general principles within patent law.[59] The tests for novelty, inventiveness and utility further require the construction of a ‘prior art base’ and a ‘person skilled in the art’.[60] The following requirements are based on more traditional principles and must be applied on a case by case basis. This section attempts to highlight some of the issues involved in assessing these requirements in the context of software-related inventions.
As judges cannot be experts in all fields of science and technology, a fictitious person who is skilled in the art must be interposed.[61]
The first step in constructing this hypothetical skilled addressee is to identify the field of technology relevant to the patent, as this narrows the required skills and common general knowledge that the hypothetical addressee will posses as of the priority date.[62] The construction of a skilled addressee therefore further involves the establishing of the appropriate field, the appropriate skill level and the common general knowledge.
If the relevant field is cast inappropriately, a person may appear to be inventive, but when examined within the correct field, may not be considered inventive. This is a particular difficulty in software-related inventions where software developers routinely bridge many technologies as a matter of course and the field of the invention may not be the same as the common general knowledge possessed by the software engineer.
Constructing a person skilled in the art is also problematic due to the variety of skill levels distributed across all technologies that engage in software development. Does the person skilled in the art accumulate skills from each field of technology from which prior art information is collected? Is the person an experienced programmer, a person in the area of technology claimed, or a combination of all? It is generally assumed a person skilled in the art is not ‘exceptionally skilled’ nor inventive, but prepared to apply common knowledge.[63]
It has been held that any claim of invention should be compared against the background of that field[64]. A skilled team can also be constructed of individual addressees of varying levels of skill,[65] or could consist of highly skilled addressees who each possess a PhD qualification.[66] Additionally, the skilled addressee must also be constructed based on the grounds of the challenge. Different skills may be needed to advance the art and test obviousness, as opposed to just implementing the art to test sufficiency.[67]
The hypothetical team should be credited with the best available equipment and sufficient time to carry out the work.[68] Funding and equipment should not be a constraint to a hypothetical addressee when assessing inventiveness or insufficiency.
Under the Patents Act 1990 (Cth) the hypothetical addressee is able to consider the available common general knowledge and multiple pieces of prior art information provided it is reasonably expected that they would have been ascertained, understood, and regarded as relevant.[69] Common general knowledge possessed by the hypothetical addressee may include prior art that would form part of the common general knowledge in the particular field for a particular hypothetical addressee, or the combination of prior art if the hypothetical addressee would have regarded the disclosures as a single source of information[70].
Constructing the prior art base for software is exceedingly difficult because software spans almost every field of technology. Therefore, does the scope of the prior art base extend only to the areas of application for the claim or to all areas in which software is used? A further issue in determining the contents of a prior art base involves collecting and assessing material that has been made publicly available. The decentralised nature of the software industry hampers this task as software developments are not readily published[71] and typically are supplied to the public in machine readable form. Although this could be considered a public disclosure, it remains difficult to dissect the software and identify any underlying invention for the purpose of assessing novelty and inventive step. The extent of software licenses may also need consideration when addressing public disclosure for non-exposed inventions[72].
Expert evidence can be used to help provide an example of the hypothetical addressee[73], although there are consequences when the expert’s qualifications do not precisely match those of the hypothetical addressee. Although cases exist where no person can be found to possess the required skills of the hypothetical addressee, particularly within the jurisdiction, the court is willing to concede that a person might exist.[74]
Providing expert evidence is made more difficult with the enquiry requiring the expert to determine whether the steps taken would be tried as a matter of course. Since retrospective analysis of the patent is discouraged, the expert must provide this analysis without the aid of seeing the patent.[75] This has lead to the Federal Court requesting the patentee to provide a statement of the involved ‘inventive step’,[76] to provide a basis from which to assess obviousness.
The hypothetical addressee, once constructed in conjunction with the appropriate common general knowledge and prior art, is interposed to aid the court in establishing novelty, inventive step and utility. These are separate enquiries that need to be conducted upon a challenge to patent validity.
Novelty further requires that the claim is not pre-empted by the prior art,[77] requiring analogous use[78]. Although software is usually built from pre-existing modules with some controlling code, it is inappropriate to dissect the claims into old and new elements, and ignore the presence of the old elements in the analysis[79]. The basic test for anticipation or want of novelty is the same as that for infringement;[80] however given the lack of paper publications, a paper anticipation[81] would require great care in comparing essential integers and mechanical equivalents. Although software development commonly is achieved by reusing modules, it is not permissible to mosaic individual pieces from the prior art when assessing novelty.[82]
Inventive step, assessed on the basis of obviousness, would have a threshold set by both the construction of a ‘person skilled in the art’ and ‘prior art base’ as stated earlier. An inventive step can be found in processes that are foreign to the normal use of computers and, for that reason, are inventive.[83] Inventiveness when considered ‘nothing but a claim for a new use of an old substance’[84] is not sufficient quality of inventiveness.[85] The doctrine of ‘analogous use’ is difficult to apply when migrating current practices and methods to digital form.[86] However, if an integer is new and otherwise patentable, its inclusion in a collocation satisfies section 18(1)(a) of the Patents Act 1990 (Cth), provided a patentee limits that grant in any way when claiming the collocation[87].
The process of developing software is generally as a solution to a given problem provided to, or conceived by, the programmer. Once given the problem, it is difficult to understand whether there are inventive steps required in providing the solutions. It can be conceded that there will be such cases where there has been a difficulty in achieving a result for a period of time but it is generally a matter of applying software engineering principles to that problem.
The High Court has clarified the test for obviousness under the Patents Act 1952 (Cth),[88] finding that it differs from that of the United Kingdom and is more closely aligned to that of the United States. The United Kingdom test for obviousness relies on the ‘obvious to try’[89] or ‘problem solution’ [90] approaches, where the ‘obvious to try’ approach has been accepted as a valid test for obviousness under the Patents Act 1977 (UK)[91]. United States courts rejected the ‘obvious to try’ approach[92] on the basis that ‘research is not undertaken with complete blindness but rather with some semblance of a chance for success’. United States courts rely on testing whether a person skilled in the art, by considering the prior art as a whole, would consider the claimed subject matter to be obvious.[93]
The High Court also rejected the ‘obvious to try’ approach,[94] instead favouring the United States approach.[95] However, despite rejecting the ‘problem solution’ approach, it was found that there remains a need to establish the problem to be solved. The Court preferred a test formulated on the question: would the hypothetical addressee at the relevant date, in all the circumstances, which include a knowledge of all the relevant prior art and of the facts of the nature and success of the prior art, directly be led as a matter of course to try the claimed invention, in the expectation that it might well produce a useful or better alternative to the prior art?[96] However, this is only useful for research and not flashes of genius[97]. It has been expressed that this approach raises the bar for testing ‘obviousness’, especially since it requires no starting point for the enquiry.[98]
Inutility should be distinguished from insufficiency despite understandable overlap.[99] Insufficiency exists when the result promised in the specification cannot be achieved because of deficiencies in the specification, whereas inutility exists when the result promised in the specification cannot be achieved by following the specification.
Utility requires that the invention should enable the addressee to achieve the result claimed through the use of that invention,[100] allowing minor obvious modifications to be applied. As the addressee should have skills that are lower, or equivalent, to those required to test inventiveness, lowering the threshold for inventive step should increase the disclosure requirement and vice versa. As code, or even a detailed software design, is not generally provided within a patent specification, significant skill may be required of the ‘person skilled in the art’ to effect what is claimed, which in turn should be considered when evaluating inventive step.
For utility, the test is whether the complete specification read as a whole contains sufficient instruction to enable the hypothetical addressee to produce each claim without invention of his/her own.[101]
Technology, like the law, is continually evolving and adapting current solutions to emerging situations. Especially in rapidly changing times, there may not be an exact common law precedent for a new societal or technological problem. So judges have a creative role to solve such uncertainties, however much some might suggest otherwise.[102] A balance between intellectual property laws and competition laws, including compulsory licensing, and misuse of market power though the leverage of intellectual property, must also be addressed.
Despite a global trend to harmonise intellectual property laws, patent law in Australia continues to be distinguished from United Kingdom approaches. The European Commission has proposed a directive to harmonize EU member states’ laws to ensure legal protections for software throughout the EU. In the proposal, the Commission noted that the status of computer-implemented inventions, including software, was ambiguous because member states interpret patent law differently from one another and from the Munich-based European Patent Office (EPO)[103].
It has been recognised that there is trend to align with the United States on approaches to patent law, a trend expected to continue under the Australia-United States Free Trade Agreement. This is despite published concerns over the direction Australia is taking, including those by the former Chair of the Intellectual Property and Competition Committee[104] and United States Federal Trade Commission (FTC)[105].
The former Chair of the Intellectual Property and Competition Committee, predicted a worrying future for patent law in Australia in which harmonisation with United States patent law through the FTA increased patent protection, disproportionately favouring producer interests.[106] Ironically, this is occurring despite a recent report by the FTC highlighting problems in the U.S. patent system.[107]
The FTC report highlighted two specific problems in the United States, namely the granting of excessively broad patents, and the granting of too many trivial patents, both having anti-competitive effects.
The FTC recognised that much of this thicket of overlapping patent rights results from the very nature of the technology where computer hardware and software contain an incredibly large number of incremental innovations. This has given rise to more and more patents being sought on incremental inventions, so that firms have enough bargaining chips to obtain access to others’ overlapping patents. The FTC recommended that the validity of a patent should be determined based on a ‘preponderance of the evidence’[108] and a toughening of legal standards when evaluating whether a patent is ‘obvious’[109].
Locally there are also concerns with many aspects of patent law, from the tests used to assess patent validity, to the patent system’s struggle to strike a balance between the inventors’ and the public’s rights.
The process of testing patent validity has introduced so many layers of obscurity that there is now a real danger of entering upon a degree of unreality that goes well beyond reasonable hypothesis[110] so that even conscientious decision-makers could lose the way and forget that the statute ultimately is the only true statement of the governing law.[111] The law has become encumbered with verbal expressions, sometimes remote from the statute, which are often picked up and applied to a case in place of the statute itself. Despite these expressions of concern when formulating general tests for obviousness[112], a test must be provided in each case. This is made more difficult by the requirement that to assess inventiveness it must be established that the skilled addressee possesses none.
It is clear that there is significant scope for error when it is difficult to determine a field of technology for a software development, a person or team skilled in the art, the level of skill possessed by the developers and the varying amounts of common general knowledge retained by developers. This is compounded by the recursive proposition raised when testing whether there has been an inventive step, which requires the construction of a hypothetical addressee who is not inventive. Since even a scintilla of inventiveness is sufficient to support patentability, any narrowing of the field of technology containing or lowering the skill or knowledge level of the hypothetical addressee could result in a finding of invention. In regard to software, typically built on incremental improvements, what is seen as inventive could be merely novel.
Even when a hypothetical skilled person can be constructed, there are still real hurdles in establishing obviousness. This stems from the higher threshold for testing obviousness set by the requirement that the person be ‘directly led’ rather than the approach be ‘obvious to try’, and is compounded by the issues relating to expert witnesses when providing evidence as a person skilled in the art.
As software design often consists of satisfying a specification, it is arguable that the test for obviousness within the development of software would be suitable to the ‘problem solution approach’[113], if it was accepted by Australian courts, as it would align with the typical development processes of the industry.
Additionally, a public interest served by patents is promoting the disclosure of innovations that might otherwise stay hidden as trade secrets. Despite this interest, most patent applications do not require the disclosure of the software code itself.[114]
Many of these issues are clearly relevant to most fields of technology, but are compounded where software-related inventions are involved, due to the ubiquitous nature of software and the fact that software developers are typically highly mobile in careers[115], each having worked in many technological fields with other cross-disciplined developers.
Since software has hardware equivalents there should be no need to handle software differently than an equivalent piece of hardware or restrict the claims to the software implementation.[116] Although it is suggested that software patents should not be allowed because they restrict trade - this is the very purpose of granting a patent. The real issue is whether the threshold for grant is sufficiently high, an issue ultimately left to the courts.
The United States Federal Trade Commission,[117] recognising that the thicket of overlapping patent rights is a consequence of continuing incremental innovation in computer hardware and software technology, recommends the toughening of legal standards when evaluating patent validity[118]. However Australia’s continuing alignment with United States approaches for testing patent validity is further increasing the threshold for obviousness in favour of the patentee.
Under the current system, where testing patent validity has introduced so many layers of obscurity, there is a real danger of creating the position so removed from reality that is becomes unreasonable to construe any reasonable hypothesis.
Based on accepted United States patents there is currently a very low barrier posed by the ‘abstract idea’.[119] Further harmonisation between Australian and United States patent legislation and systems will continue[120] despite warnings that this will foreshadow increases in patent protection.[121]
This paper concludes that despite software being patentable, there remains a lack of understanding as to the nature of software. Limited judicial and legislative direction combined with inadequate methods of assessing the patent validity of software claims does not provide a satisfactory environment for a rapidly growing industry. An effort must be made to improve the quality of patents granted.
∗ BE hons (computer systems) (JCU), PhD (JCU), M Eng Sci (signal processing) (UNSW), Grad Dip IT Law (UTS); Senior Research Engineer for Patentable Research and Development, www.patentable.com.au.
1 Proposal for a Directive of the European Parliament and of the Council on the patentability of computer-implemented inventions [2002] OJ C 151E/129. The European Parliament rejected this proposal in July 2005.
[2] Australia-United States Free Trade Agreement, 18 May 2004, [2005] ATS 1(entered into force 1 January 2005).
[3] United States Patent No 3,821,715 (The Intel 4004).
[4] Initial consumer computers included the Scelbi, IBM 5100, Apple I & II, Commodore PET and TRS-80.
[5] 1978 - 1979 saw the development first spread sheet (VisiCalc) and word processor (WordStar).
[6] Mobility also makes it possible for small software developers to distribute software directly to consumers and is emphasized in network-oriented software (such as Java).
[7] Software only needs a system of execution and not a physical container (‘material form’) to make it useful.
[8] Unlike physical goods, Australian Customs Officers cannot seize software downloaded over the internet.
[9] The near-monopoly positions of IBM in the 1970s and Microsoft in the 1990s are extreme examples of how control of a fundamental standard (in this case, the operating system) leads to control of application software as well.
[10] Eg, Trade Practices Act 1974 (Cth); Fair Trading Act 1989 (Qld); Fair Trading Act 1987 (NSW).
[11] World Intellectual Property Organisation, Model Provisions on the Protection of Computer Software, WIPO Publication No 814, January 1978, 6-19, 12. A ‘computer program means a set of instructions capable … of causing a machine … to achieve a particular function, task or result’.
[12] Copyright Act 1968 (Cth) s 10. A ‘computer program means a set of statements or instructions to be used … in a computer … to bring about a certain result.’
[13] United States Copyright Act, 17 USC § 101 (2000).
[14] Copyright Act 1969 (Cth) s 31(1) (a) specifies the exclusive copyright that subsists in original computer programs, through the inclusion of computer programs in the s 10 definition of ‘literary work’.
[15] Patents Act 1990 (Cth) s 18. Requires novelty, inventiveness (or innovation) and utility.
[16] Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), opened for signature 15 April 1994, [2005] ATS 38, art 27 para 1, (entered into force 1 January 1995) ‘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.’
[17] Butterworths, Patents, Trade Marks & Related Rights: Looseleaf Service (James Lahore ed, 2005 ed) vol 1, 12,510–12,520 [‘Lahore’], discussing British Petroleum Co Ltd’s Application (1968) 38 AOJP 1020, involving patent applications made under the Patents Act 1952 (Cth).
[18] Pamela Samuelson et al, ‘A Manifesto Concerning the Legal Protection of Computer Programs’ (1994) 94 Columbia Law Review 2308, 2317 stating ‘[n]o one would want to buy a program that did not behave…’.
[19] Lahore, above n 17, refers to ‘traditional principles’. In British Petroleum Co Ltd’s Application (1968) 38 AOJP 1020 the applications were rejected as being a solution to mathematical problem and not patentable subject matter.
[20] United States Patent No 1,647 (Samuel Morse). In this patent claim 8 referred to the use of electro-magnetism, however developed, for making or printing intelligible characters, letters, or signs, at any distances, being a new application of that power.
[21] O'Reilly v Morse, 56 US 62 (1853).
[22] Dolbear v American Bell Telephone Co (The Telephone Cases), [1888] USSC 108; 126 US 1, 534 (1888) states ‘the effect of that decision was, therefore, that the use of magnetism as a motive power, without regard to the particular process with which it was connected in the patent, could not be claimed, but that its use in that connection could.’
[23] Re GEC's Application (1942) 60 RPC 1, 4.
[24] National Research Development Corp v Commissioner of Patents [1959] HCA 67; (1959) 1A IPR 63 (‘NRDC’).
[25] Ibid 74, applying Re Elton & Leda Chemicals Ltd’s Application (1957) RPC 267.
[26] Ibid 75, applying Rantzen’s Application (1947) 64 RPC 63 in which an electrical oscillation was considered a product.
[27] United Kingdom, The British Patent System: Report of the Committee to Examine the Patent System and Patent Law, Cmnd 4407 (1970) [Banks Report].
[28] Lahore, above n 17, 12,525. The Banks Committee Report at 487 recommended the exclusions of computer programs in whatever form the invention was presented. This exclusion was to cover methods of programming computers, a computer when programmed in a particular way and cases where the novelty or alleged novelty of the invention lay only in the program.
[29] Gever's Application (1970) RPC 91. The claim related to punched cards used to contain the program being patentable.
[30] Burroughs Corp’s Application (1973) FSR 493.
[31] See, eg, Gottschalk v Benson, [1972] USSC 223; 409 US 63 (1972); Parker v Flook, [1978] USSC 129; 437 US 584 (1978).
[32] Gottschalk v Benson, [1972] USSC 223; 409 US 63 (1972). The claim was for a program of a general computer to perform the mathematical formula associated with converting binary-coded decimal (BCD) numerals into pure binary numerals.
[33] Ibid 72. In support of this it should be noted that, in my opinion, claim 8 and claim 13 in referring to the (i+3)th and (i+1)th binary digit, simply refer to the fact 23+21=10.
[34] Ibid 71. The court acknowledged that direct attempts to patent programs have been rejected on the grounds of non-statutory subject matter.
[35] Parker v Flook, [1978] USSC 129; 437 US 584 (1978). Parker claimed a method of updating an alarm limit, and was denied patentability. Discussion at 595, citing In re Richman, 563 F 2d 1026, 1030 (CCPA, 1977) and O’Reilly v Morse, 56 US 62 (1853) ‘assuming the formula to be within prior art, as it must be’.
[36] British Petroleum Co Ltd’s Application (1968) 38 AOJP 1020, 1021 (‘British Petroleum’). Involved patent applications made under the Patents Act 1952 (Cth). 30,390/63 was directed at programming means which are used to control a computer to operate in a particular way and 30,391/63 the claims were for a new computer, being the computer when programmed. The attitude was restated in Telefon A/B LM Ericssons Application (1974) 44 AOJP 846.
[37] British Petroleum (1968) 38 AOJP 1020, 1021.
[38] Ibid.
[39] Patents Act 1977 (UK) s 1 (2)(c).
[40] Merrill Lynch Inc’s Application, applying Vicom Systems Inc’s Application (1987) Official Journal EPO 119.
[41] Pitman’s Application (1969) RPC 646, 650.
[42] Diamond v Diehr, [1981] USSC 40; 450 US 175 (1981).
[43] Ibid 175. ‘[A] claim drawn to subject matter otherwise statutory does not become non-statutory simply because it uses a mathematical formula, computer program, or digital computer’.
[44] Gottschalk v Benson, [1972] USSC 223; 409 US 63 (1972); Parker v Flook, [1978] USSC 129; 437 US 584 (1978); Diamond v Diehr[1981] USSC 40; , 450 US 175 (1981). The CCPA has been replaced by the Federal Circuit Court of Appeals (Fed Cir).
[45] Re Walter, 618 F 2d 758, 767 (CCPA, 1980).
[46] Ibid (applying Parker v Flook, [1978] USSC 129; 437 US 584 (1978)).
[47] State Street Bank and Trust Co v Signature Financial Group Inc, [1998] USCAFED 107; 149 F 3d 1368 (Fed Cir, 1998). The case involved a claim over a system for doing the bookkeeping for a suite of mutual funds.
[48] Based on Re Freeman, 573 F 2d 1237, 1245 (CCPA, 1978) which proposed the question ‘Does the claim in its entirety wholly pre-empt the algorithm?’; Re Walter, 618 F 2d 758 (CCPA, 1980) which required the equation to be implemented in a specific manner to define structural relationships between the physical elements of the claim (in apparatus claims) or to refine or limit claim steps (in process claims), the claim being otherwise statutory; Re Abele, 684 F 2d 902 (CCPA, 1982).
[49] Australian Patent Office (APO), ‘Guidelines for Considering the Patentability of Computer Program Related Inventions’, Patent Office, Canberra, March 1986.
[50] Patents Act 1990 (Cth), s 18 (1A)(a) requires that the invention ‘is a manner of manufacture within the meaning of section 6 of the Statute of Monopolies’.
[51] International Business Machines Corp v Commissioner of Patents [1991] FCA 625; (1991) 22 IPR 417 (‘IBM’). The case involved a patent application for an improved method of visually representing curve images in computer graphics displays.
[52] Ibid 424. The court held that in offering an advantage which was material (belonged to a useful art) and having commercially useful effect in computer graphics was in the field of economic endeavour and therefore afforded protection in patent law.
[53] IP Australia, Patent Office Manual of Practice & Procedures, (August 1992), 35.65, as discussed in Anne Fitzgerald, ‘Background to Software Patenting in Australia’ (Paper presented at the, Software Engineering Conference, Australia, 29 September to 2 October 1997) 113–114.
[54] CCOM Pty Ltd v Jiejing Pty Ltd (1994) 122 ALR 417 (‘CCOM’). The claims involved a computer programmed to enable the retrieval and display of Chinese language characters for assembly in text to be used in word processing.
[55] IP Australia, Patent Office Manual of Practice & Procedures, (August 1, 2006) 2.9.2.7 (quoting CCOM).
[56] IP Australia, Patents for Computer Related Inventions (2005) 2.
[57] CCOM (1994) 122 ALR 417, 446. This case established the Patents Act 1990 (Cth) s 18 (1) (a)-(d) provided separate heads which emphasised the grounds of patentability.
[58] Patents Act 1990 (Cth) s18 (1) subsection (b) requires (i) novelty and (ii) an inventive step; subsection (c) requires utility.
[59] NV Phillips Gloeilampenfabrieken v Mirabella International Pty Ltd (1995) 32 IPR 449, 452. The threshold of inventiveness and newness involves assessing what is an invention and is separate from that prescribed in Patents Act 1990 (Cth) s 18 (1)(a)-(d), continuing traditional principles from the Patents Act 1903 (Cth) s 4, the Patents Act 1952 (Cth) s 6 and the Patent Design and Trade Marks Act 1883 (UK) s 46.
[60] Patents Act 1990 (Cth) s 7 defines ‘novelty’ and ‘inventive step’.
[61] Aktiebolaget Hassle v Alphapharm Pty Ltd [2002] HCA 59; (2002) 56 IPR 129, 167 148 (Kirby J) (‘Astra’).
[62] ICI Chemicals v Lubrizol Corp Inc (2000) 106 FCR 213, 232.
[63] Valensi v British Radio Corp (1973) RPC 337, 377 provides a description of a diligent worker; General Tire & Rubber Co v Firestone Tyre and Rubber Co Ltd (1972) RPC 457 recognises a team as a person skilled in the art.
[64] The Leonardis v Sartas No 1 Pty Ltd [1996] FCA 449; (1996) 35 IPR 23, 44.
[65] Genentech Inc v Wellcome Foundation Ltd (1989) 15 IPR 423, 548 (Mustill LJ). ‘There is no single standard. Nor do I have any faith in the proposition that formal qualifications are the test, or even a useful test.’
[66] Ibid 509.
[67] Ibid 546.
[68] Ibid 528.
[69] Patents Act 1990 (Cth), s 7 (2), as amended in 2001.
[70] Patents Act 1990 (Cth), s 7 (2), as amended in 2001.
[71] US Congress, Office of Technology Assessment, ‘Computer Software and Intellectual Property-Background Paper’, OTA-BP-CIT-61 (Washington, DC: U.S. Government Printing Office, March 1990) 9.
[72] Most software licenses do not authorise disassembly. Humpherson v Syer (1887) 4 RPC 407, 413, has ‘this information been communicated to any member of the public who was free in law and equity to use it as he pleased?’
[73] Astra (2002) 56 IPR, 167 148 (Kirby J).
[74] Gambro Pty Ltd v Fresenius Medical Care South East Asia Pty Ltd [2004] FCA 323, 356-359.
[75] Minnesota Mining and Manufacturing Company v Tyco Electronics Pty Ltd [2002] FCAFC 315; (2002) 56 IPR 248, 258; Astra [2002] HCA 59; (2002) 56 IPR 129, 135.
[76] Lahore, above n 17, 15,200.
[77] Commissioner of Patents v Microcell Ltd [1959] HCA 71; (1959) 102 CLR 232, 251 ‘claim for the use of a known material in the manufacture of known articles for the purpose of which its known properties make that material suitable ... cannot be subject matter for a patent’; Re BA’s Application (1915) 32 RPC 348, 349 ‘nothing but a claim for a new use of an old substance’.
[78] NRDC [1959] HCA 67; (1959) 1A IPR 63, 65 (per the majority) when examining the phrase ‘nothing but’.
[79] Commissioner of Patents and Trademarks v Diehr, [1981] USSC 40; 209 USPQ 1, 9 (1981).
[80] Meyers Taylor Pty Ltd v Vicarr Industries Ltd, [1977] HCA 19; (1977) 137 CLR 228, 237.
[81] Nicaro Holdings Pty Ltd v Martin Engineering Co [1990] FCA 40; (1990) 91 ALR 513, 528.
[82] Minnesota Mining & Manufacturing Co v Beiersdorf Ltd [1980] HCA 9; (1980) 29 ALR 29, 57.
[83] IBM [1991] FCA 625; (1991) 22 IPR 417, 424.
[84] Re BA’s Application (1915) 32 RPC 348, 349.
[85] NRDC [1959] HCA 67; (1959) 1A IPR 63, 65, interpreting Commissioner of Patents v Microcell Ltd (1959) 1A IPR 52 and Re BA's Application (1915) 32 RPC 348.
[86] British Liquid Air Co Ltd v British Oxygen Co Ltd (1909) 26 RPC 509, 532. Analogous use requires ‘the application of a known device to its ordinary purpose under analogous circumstances’.
[87] B.U.S.M. v Fussel (1908) 25 RPC 631, 649.
[88] Astra [2002] HCA 59; (2002) 56 IPR 129.
[89] Johns Manville Corporations Patent (1967) RPC 479, 493.
[90] Implementing Regulations to Part III of the European Patent Convention Rule 27(1)(c) [1991] OJ EPO 4ff. Rule 27 (1)(c) requires the application to ‘disclose the invention, as claimed, in such terms that the technical problem…and its solution can be understood…’.
[91] Brugger v Medic-Aid (1996) RPC 635, 661, accepting the ‘obvious to try’ test, and further extending the test to a second round of ‘obvious to try’ tests after the initial tests are concluded.
[92] Application of Tomlinson, 360 F 2d 928, 931 (CCPA, 1966).
[93] Re Wood 599 F 2d 1032 (CCPA, 1979).
[94] Astra [2002] HCA 59; (2002) 56 IPR 129, 148.
[95] Ibid 149, citing Application of Tomlinson, 360 F 2d 928 (CCPA, 1966).
[96] Ibid 142-143 citing Graham J in Olin Mathieson Chemical Corp v Biorex Laboratories Ltd (1970) RPC 157, as referred to in the argument in Wellcome Foundation Ltd v VR Laboratories (Aust) Pty Ltd [1981] HCA 12; (1981) 148 CLR 262, 267.
[97] Astra [2002] HCA 59; (2002) 56 IPR 129, 139.
[98] Kim O’Connel and Julian Cooke, ‘Australia: A Patentee’s Paradise?’ (2003) 25 European Intellectual Property Review 481.
[99] Lahore, above n 17, 15,205, citing Tetra Molelectric Ltd’s Application (1977) RPC 290, 297; Nicholls v Kershaw (1910) 27 RPC 237, 250–1; Monnet v Beck (1897) 14 RPC 777, 847.
[100] Welch Perrin Co Pty Ltd v Worrel [1961] HCA 91; (1961) 106 CLR 588, 601-2 (Menzies J). A claim will fail on the grounds of inutility if it contains matter which will prevent it from achieving the desired result even if a skilled person would be able to avoid the inappropriate or inapplicable material.
[101] Thomas Blanko White, Patent For Invention (5th ed, 1983) 4-502.
[102] Justice Michael Kirby, ‘Judicial Dissent’ (Speech delivered at Inter Alia, the Law Students’ Society of James Cook University, Cairns, 26 February 2005).
[103] Aaron C. Chatterjee, ‘Europe struggles over software patents’ (2004) 41 IEEE Spectrum 63.
[104]Henry Ergas, ‘Patent protection an FTA complication’, Australian Financial Review, 24 February 2004. Mr. Ergas was the former chairman of the Intellectual Property and Competition Review Committee that conducted the review of intellectual property legislation under the competition principles agreement.
[105] Federal Trade Commission (FTC), To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy (2003).
[106] Ergas, above n 104.
[107] FTC, above n 105.
[108] Ibid 8-10 (recommendation 2).
[109] Ibid 10-12 (recommendation 3).
[110] ICI Chemicals & Polymers Ltd v Lubrizol Corp Inc [2000] FCA 1349; (2000) 49 IPR 513, 533 67.
[111] Astra [2002] HCA 59; (2002) 56 IPR 129, 167 (Kirby J).
[112] Johns Manville Corporations Patent (1967) RPC 479 (Diplock LJ); Astra [2002] HCA 59; (2002) 56 IPR 129.
[113] IP Australia, above n 55, 2.5.16 (citing HPM Industries Pty Ltd v Gerard Industries [1957] HCA 47; (1957) 98 CLR 424, 437).
[114] Business Software Alliance, headquartered in Washington, D.C. www.bsa.org.
[115] Jim Bright, ‘Change of Career’ (2004) June/July Professional Network a publication of the Association of Professional Engineers, Scientists and Managers, Australia (APESMA) (discussing research that ‘people entering the workforce today will have, on average, five different careers (not five different jobs’)).
[116] Welcome Real Time v Catuity Inc [2001] FCA 445; (2001) 51 IPR 327, 353; supported by State Street Bank & Trust Co v Signature Financial Group, [1998] USCAFED 107; 149 F 3d 1368 (Fed Cir, 1998), where it was considered that ‘business methods should be subject to the same legal requirements for patentability as applied to any other process or method.’
[117] FTC, above n 105.
[118] Ibid 8-12 (recommendations 2 and 3).
[119] United States Patent No 6,078,938 (Method and system for solving linear systems) is a patent assigned to Motorola Inc that combines a programming technique to a known mathematical procedure to produce the known results to that procedure on a computer. United States Patent No 6,807,536 (Method and system for computing singular value decomposition of matrices and low rank approximations of matrices) is a patent assigned to Microsoft Inc that implements known mathematical procedures with known results on a computer.
[120] Australia-United States Free Trade Agreement, 18 May 2004, [2005] ATS 1, art 17.9, para 14 (entered into force 1 January 2005).
[121] Ergas, above n 104.
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/JCULawRw/2005/2.html