Computer Edge Pty Ltd v Apple Computer Inc [1986] HCA 19; (1986) 161 CLR 171 (6 May 1986)

HIGH COURT OF AUSTRALIA

COMPUTER EDGE PTY LTD v. APPLE COMPUTER INC. [1986] HCA 19; (1986) 161 CLR 171
F.C. 86/017

Copyright

High Court of Australia
Gibbs C.J.(1), Mason(2), Wilson(2), Brennan(3) and Deane(4) JJ.

CATCHWORDS

Copyright - Literary work - Infringement - Computer program - American computer marketed in Australia - Source program - Whether literary work - Whether original - Object program - Whether literary work - Infringement by adaptation - Translation - Whether object program translation of source program - Infringement by reproduction - Taiwan computer marketed in Australia containing object program copied from American object program - Whether Taiwan object program reproduction of American source program - Copyright Act 1968 (Cth), ss. 10(1) "adaptation", 22, 31(1), 32, 37, 38.

HEARING

1985, March 12-14; 1986, May 6. 6:5:1986
APPEAL from the Federal Court of Australia.

DECISION

GIBBS C.J.: This appeal from a decision of the Full Court of the Federal Court (Fox and Lockhart JJ, Sheppard J. dissenting) (1984 [1984] FCA 137; 1 FCR 549 raises for consideration the questions whether, and to what extent, the Copyright Act 1968 (Cth), as amended, in the form which it took before the enactment of the Copyright Amendment Act 1984, recognized and protected copyright in computer programs.

2. The first respondent, a Californian corporation, makes and sells computers under the well-known name of "Apple" and the second respondent is its Australian distributor. The first appellant, a Victorian company, has imported and sold in Australia micro-computers under the name of "Wombat" and the second appellant is the managing director of the first appellant. The respondents' case is that the importation and sale in Australia of the Wombat micro-computer was an infringement of the first respondent's copyright in two computer programs, known respectively as "Applesoft" and "Autostart". The alleged infringements occurred before the commencement of the Copyright Amendment Act 1984, and that Amendment Act, which is intended to protect computer programs, has no relevance to the present case: see s.7(2). References hereunder to the Copyright Act will be to that Act before its amendment in 1984.

3. Computer science makes much use of jargon and metaphor, and to enable the matters in issue to be understood it seems desirable to attempt a brief explanation of the meaning of some of the expressions used in that science and to describe the manner in which a computer program is developed. A computer program is a set of instructions designed to cause a computer to perform a particular function or to produce a particular result. A program is usually developed in a number of stages. First, the sequence of operations which the computer will be required to perform is commonly written out in ordinary language, with the help, if necessary, of mathematical formulae and of a flow chart and diagram representing the procedure. In the present case if any writing in ordinary language (other than the comments and labels mentioned below) was produced in the production of Applesoft and Autostart, no question now arises concerning it. Next there is prepared what is called a source program. The instructions are now expressed in a computer language - either in a source code (which is not far removed from ordinary language, and is hence called a high level language) or in an assembly code (a low level language, which is further removed from ordinary language than a source code), or successively in both. Sometimes the expression source code seems to be used to include both high level and low level language. In the present case, the source programs were written in an assembly code, comprising four elements, viz. -

(a) labels identifying particular parts of the program;
alphabet and corresponding to a particular
operation expressed in 6502 Assembly Code (the code
used);

(c) mnemonics identifying the register in the
microprocessor and/or the number of instructions in
the program to which the operation referred to in
(b) related; and

(d) comments intended to explain the function of the
particular part of the program for the benefit of
a human reader of the program.

The writing has been destroyed, although it is possible to reconstruct the mnemonics, but not the labels and comments, which were comprised in it.

4. The source code or assembly code cannot be used directly
in the computer, and must be converted into an object code, which is "machine readable", i.e., which can be directly used in the computer. The conversion is effected by a computer, itself properly programed. The program in object code, the object program, in the first instance consists of a sequence of electrical impulses which are often first stored on a magnetic disk or tape, and which may be stored permanently in a ROM ("read only memory"), a silicon chip which contains thousands of connected electrical circuits. The object code is embodied in the ROM in such a way that when the ROM is installed in the computer and electrical power is applied, there is generated the sequence of electrical impulses which cause the computer to take the action which the program is designed to achieve. The pattern of the circuits in the ROM may possibly be discerned with the aid of an electron microscope but it cannot be seen by the naked eye. Obviously, the electrical impulses themselves cannot be perceived. However the sequence of electrical impulses may be described either in binary notation (using the symbols 0 and 1) or in hexadecimal notation (using the numbers 0-9 and the letters A-F), and it is possible to display the description on the visual display unit of the computer, and to print it out on paper. And, as has been said, it is also possible to reconstruct the mnemonics in the source code. It will have been seen from this account that a program exists successively in source code and in object code, but the object code need not be written out in binary or hexadecimal notation in the process of producing and storing the program.

5. The relevant model of the Apple computer contained six
ROMs which stored both the Applesoft object program and the Autostart object program. The Wombat computer had two ROMs and one EPROM ("erasable programmable read only memory") which stored object programs which the evidence shows were the same as those stored in the six Apple ROMs. An EPROM is a chip which differs from a ROM in that the information on it may be erased and the chip may be reprogramed, but this difference is not material. Nothing turns on the fact that the Applesoft program and the Autostart program were stored on six ROMs in the Apple computer and in only two ROMs and one EPROM in the Wombat computer, for there is no doubt that the programs which appear in the Wombat ROMs and EPROM are the same as those which were prepared for the first respondent and which are used in the relevant Apple computer. It is not clear how the appropriation of the first respondent's programs was effected, but it was possible for the patterns of one ROM to be embodied directly in another ROM or in an EPROM without first representing the object code in binary or hexadecimal notation and without converting the object code back to source code. In other words, the making of the Wombat ROMs and EPROM did not necessarily involve the use of any writing or other visual display of the programs in either source code or in binary or hexadecimal notation.

6. In these circumstances the ultimate question for
decision is whether to the knowledge of the appellants the making of the ROMs and the EPROM used in the Wombat computer would, if those articles had been made in Australia by the first appellant, have constituted an infringement of the first respondent's copyright. It will be sufficient, in citing the relevant sections of the Copyright Act, to refer only to such parts of those sections as deal with literary works, since there is obviously no suggestion that any question arises as to dramatic, musical or artistic works. Section 37 of the Copyright Act provides, inter alia, that the copyright in a literary work is infringed by a person who, without the licence of the owner of the copyright, imports an article into Australia for the purpose of selling the article "where, to his knowledge, the making of the article would, if the article had been made in Australia by the importer, have constituted an infringement of the copyright". Section 38(1) similarly provides, inter alia, that the copyright in a literary work is infringed by a person who, in Australia, and without the licence of the owner of the copyright, sells an article "where, to his knowledge, the making of the article constituted an infringement of the copyright or, in the case of an imported article, would if the article had been made in Australia by the importer, have constituted such an infringement". The learned trial judge made no finding on the question of knowledge and the Full Court ordered that if the present respondents wish to pursue claims for damages in relation to infringements, the matter should be referred back to the learned trial judge for findings on that question; this order is the subject of a cross-appeal.

7. The nature of copyright is described in s.31(1) of the
Copyright Act, which, so far as is possibly relevant for present purposes, provides as follows:

"For the purposes of this Act, unless the
contrary intention appears, copyright, in relation
to a work, is the exclusive right -

(a) in the case of a literary ... work, to do

all or any of the following acts:

(i) to reproduce the work in a material
form;

...

(vi) to make an adaptation of the work;

(vii)to do, in relation to a work that is
an adaptation of the first-mentioned
work, any of the acts specified in
relation to the first-mentioned work
in sub-paragraphs (i) to (v),
inclusive;

..."

The copyright will subsist only if the literary work is "original": s.32. Section 36(1) of the Copyright Act provides, inter alia:

"Subject to this Act, the copyright in a
literary ... work is infringed by a person who, not
being the owner of the copyright, and without the
licence of the owner of the copyright, does in
Australia, or authorizes the doing in Australia of,
any act comprised in the copyright."

The effect of these sections is that the making of the Wombat ROMs and EPROM in Australia would have constituted an infringement of the first respondent's copyright if either -

(1) (a) the Applesoft and Autostart source programs
were original literary works; and

(b) the Applesoft and Autostart object programs

were adaptations of the Applesoft and
Autostart source programs; and

(c) the Wombat object programs embodied in the

ROMs and EPROM were reproductions of the
Applesoft and Autostart object programs; or

(2) (a) the Applesoft and Autostart source programs
were original literary works; and

(b) the Wombat object programs embodied in the

ROMs and EPROM were either adaptations or
reproductions of the Applesoft and Autostart
source programs respectively; or

(3) (a) the object programs embodied in the Apple ROMs
were original literary works; and

(b) the object programs embodied in the Wombat

ROMs and EPROM were reproductions of the
Applesoft and Autostart object programs
respectively.

8. The majority of the Federal Court accepted the first of
these propositions as correct. It is clear that if either the source programs or the object programs of Applesoft and Autostart were original literary works in which copyright might subsist within the meaning of the Copyright Act, the first respondent was the owner of the copyright.

9. It is well settled that in the case of a literary work
copyright does not exist in ideas or information but in "the order of words" (Jefferys v. Boosey [1854] EngR 816; (1854) 4 H.L.C. 815, at p.867 (10 E.R. 681, at p.702)), that is, in the particular form of expression in which the ideas or information are conveyed: see Blackie & Sons Ltd. v. Lothian Book Publishing Co. Proprietary Ltd. [1921] HCA 27; (1921) 29 C.L.R. 396, at p.400; Victoria Park Racing and Recreation Grounds Co. Ltd. v. Taylor [1937] HCA 45; (1937) 58 C.L.R. 479, at pp.498, 511. And, in the case of a literary work, the expression must be in print or writing, as the ordinary meaning of the words "literary work" suggests. In University of London Press, Limited v. University Tutorial Press, Limited (1916) 2 Ch. 601, it was said, at p.608, that "the words 'literary work' cover work which is expressed in print or writing, irrespective of the question whether the quality or style is high". That passage was cited with approval in Ladbroke (Football) Ltd. v. William Hill (Football) Ltd. (1964) 1 W.L.R. 273, at pp.285, 291; (1964) 1 All E.R. 465, at pp.475, 479. However, for the purposes of the Copyright Act "writing" is given an extended definition; it means "a mode of representing or reproducing words, figures or symbols in a visible form", and "written" has a corresponding meaning: see s.10(1).

10. The programs in the present case were first expressed as
source programs, i.e. in writing in an assembly code. A literary work need not be expressed in ordinary language. For example, it is well settled that there may be copyright in a list of words used as a telegraph code, or in a catalogue of sets of type or in a system of shorthand: Copinger and Skone James on Copyright, 12th ed. (1980), par.156. The fact that the programs were expressed in an assembly code creates no obstacle to regarding them as literary works. However, it was submitted, in reliance on Hollinrake v. Truswell (1894) 3 Ch. 420, that the source programs were not literary works because they were merely directions to be understood and used with an apparatus - viz. a computer. Hollinrake v. Truswell was a case in which a cardboard sleeve chart, i.e. a representation of a sleeve designed for a lady's arm with certain scales of measurements on it, intended for practical use in dressmaking, was not a literary work. Davey L.J. said, at p.428: "a literary work is intended to afford either information and instruction, or pleasure, in the form of literary enjoyment". That passage was recently approved by the English Court of Appeal in Exxon Corporation v. Exxon Insurance Consultants International Ltd. (1982) Ch. 119, at pp.143, 144; (1982) R.P.C. 69, at pp.88, 89. The source programs satisfy this test; they afford instruction to the operator keying in the machine that will convert the source code to object code. The source program is not itself an apparatus or contrivance like the sleeve in Hollinrake v. Truswell, nor is it part of an apparatus or contrivance like the printed face of the barometer which was held not to be the subject of copyright in Davis v. Comitti (1885) 52 L.T.N.S. 539 or the perforated sheets which were held to be part of the player piano in Boosey v. Whight (1900) 1 Ch. 122, at p.125. In my opinion, clearly the source programs were literary works.

11. Further, the source programs were "original" literary
works within the meaning of s.32 of the Copyright Act. The expression "original" in that section does not mean that the work must be the expression of original or inventive thought. "The originality which is required relates to the expression of the thought. But the Act does not require that the expression must be in an original or novel form, but that the work must not be copied from another work - that it should originate from the author": University of London Press, Limited v. University Tutorial Press, Limited, at pp.608-609 and see Sands & McDougall Proprietary Ltd. v. Robinson [1917] HCA 14; (1917) 23 C.L.R. 49 and Victoria Park Racing and Recreation Ground Co. Ltd. v. Taylor, at p.511. Originality is a matter of degree, depending on the amount of skill, judgment or labour that has been involved in making the work: see Ladbroke (Football) Ltd. v. William Hill (Football) Ltd., at pp.277-278, 282, 285; pp.469, 473, 475 of All E.R. There is no doubt that the production of the source programs in the present case involved sufficient skill, labour and experience on the part of the authors to satisfy the requirement of originality. For these reasons the source programs were original literary works.

12. The question that then arises is whether the Applesoft
and Autostart object programs embodied in the ROMs were literary works. As has been shown, those programs existed in the form of a sequence of electrical impulses, or possibly in the pattern of circuits that when activated generated those electrical impulses. On any view they were not expressed in writing or print. Although the electrical impulses could be represented by words or figures, the impulses themselves did not represent or reproduce any words and figures. They were not visible or otherwise perceptible, and they were not, and were not intended to be, capable by themselves of conveying a meaning which could be understood by human beings. Obviously, the patterns of the circuits in the ROMs also did not represent or reproduce any words or figures, and were incapable of conveying any meaning. It is true that the object programs might have been printed out in binary or hexadecimal form, but the question whether any such written expression of the programs would have been a literary work is not the question that now falls for decision. We are concerned with the object programs embodied in the ROMs and it seems clearly to follow from the cases already cited, which decide that a literary work is a work expressed in print or writing, that they were not literary works.

13. However, it was submitted on behalf of the respondents
that it is not right to say that a literary work must be expressed in writing; indeed, it was submitted, the work need not be visible. The submission was that a literary work may be made in any material form, including that of a ROM. This argument is based on s.22(1) of the Copyright Act which reads as follows:

"A reference in this Act to the time when, or
the period during which, a literary, dramatic or
musical work was made shall be read as a reference
to the time when, or the period during which, as
the case may be, the work was first reduced to
writing or to some other material form."

The sub-section refers to dramatic and musical as well to literary works. It is obvious enough that a dramatic work or a musical work may first appear in some material form other than writing. That is not so obviously true of a literary work, having regard to the wide definition of "writing" in s.10(1) of the Act, which would seem to include such things as the carving of letters on stone and the printing of Braille. One thing that is clear however is that the time to which s.22(1) refers is the time when the "work", which is defined in s.10(1) to mean "a literary, dramatic, musical or artistic work", is first reduced to writing or to some other material form. Assuming that a ROM is a material form for the purposes of the section, the question becomes whether the object code was a "work" that was reduced to the form of the ROM. As has already been seen, a source program is a literary work, and the print-out of an object program may be a literary work, but the question is whether the object program as embodied in the ROM - the sequence of electrical impulses or the pattern of circuits - was a "work", i.e. a literary work, for it was not a dramatic, musical or artistic work. It seems to me a complete distortion of meaning to describe electrical impulses in a silicon chip, which cannot be perceived by the senses and are not intended to convey any message to a human being and which do not represent words, letters, figures or symbols as a literary work; still less can a pattern of circuits be so described.

14. I therefore hold that the object programs were not
themselves literary works. It is unnecessary to consider whether the ROMs should be regarded as mere adjuncts to a mechanical contrivance (the computer) and for that reason alone should be regarded as not the subject of copyright, on the authority of such cases as Hollinrake v. Truswell. It is further unnecessary to consider whether the fact that Applesoft and Autostart are operational programs (essential to the general functioning of the computer) rather than application programs (designed for specific purposes) would be material to the consideration of such a question.

15. The next question that falls for decision is whether the
Applesoft and Autostart object programs were adaptations of the Applesoft and Autostart source programs. By s.10(1) of the Copyright Act "adaptation" is defined to mean:

"(a) in relation to a literary work in a
non-dramatic form - a version of the work
(whether in its original language or in a
different language) in a dramatic form;

(b) in relation to a literary work in a dramatic
form - a version of the work (whether in its
original language or in a different language)
in a non-dramatic form;

(c) in relation to a literary work (whether in a
non-dramatic form or in a dramatic form) -

(i) a translation of the work; or

(ii) a version of the work in which a story or
action is conveyed solely or principally
by means of pictures; and

(d) in relation to a musical work - an arrangement
or transcription of the work".

It was argued on behalf of the respondents that the embodiment of the object programs in the silicon chips was a translation of the source programs and therefore an adaptation within par.(c)(i) of this definition. Reliance was placed on the Shorter Oxford English Dictionary, which defines "translation" not only as "The action or process of turning from one language into another" but also, in transferred and figurative use, as "The expression or rendering of something in another medium or form". When the definition of "adaptation" in s.10(1) is read as a whole, it suggests that the primary, and not the transferred and figurative, meaning of "translation" is intended. If that were not so the provisions of pars.(a), (b) and (c)(ii) of the definition would appear to be surplusage. Clearly the programs in the ROMs were not a translation of the source programs within the primary meaning of the word, because when the programs were embodied in the ROMs they were not turned into another language, even into another computer language - they were turned into electrical impulses. If it be assumed, contrary to my opinion, that translation in the definition is used also in its transferred and figurative sense, I consider that the object programs in the ROMs were not a translation of the source programs within that usage of the expression. When the dictionary definition refers to "the expression or rendering of something in another medium or form" it must refer to the portrayal, utterance, representation, reproduction or depiction of the thing in a different medium or form. The ROMs did not in any way express or render the source programs; rather, the ROMs were the means of putting into action and making effective the instructions written in the source programs. Just as a person does not (except in a metaphorical sense) translate the instructions for the working of a machine when, following those instructions, he sets the machine in motion, so the electrical charges in the ROMs effectuate, but do not translate, the instructions in the source program. Fox J., in his judgment, refers to the 1979 Report of the United States National Commission on New Technological Uses of Copyright Works, where it is said:

"An object code is the version of a program in
which the source code language is converted or
translated into the machine language of the
computer with which it is to be used."

The word "translated" in this passage is used in a metaphorical sense, and the passage does not provide an example of the use of the word in its ordinary and natural meaning.

16. A further reason for supporting the conclusion that the
object programs were not adaptations is that the definition of "adaptation", and the words of s.31(1)(a)(vii), suggest that an adaptation must itself be a "work". As has been seen, the object programs were not works.

17. It follows, of course, that the Wombat object programs
themselves were not adaptations of the Applesoft and Autostart source programs. The final question for decision then becomes whether the Wombat object programs embodied in the ROMs and EPROM were reproductions of the Applesoft and Autostart source programs. A reproduction within the meaning of the Copyright Act includes a reproduction of a substantial part of the work (see s.14 of the Copyright Act). The notion of reproduction involves two elements - that the infringing work sufficiently resembles the copyright work and that it was produced by the use of the copyright work. The latter element is satisfied in the present case where the appellants clearly appropriated for themselves some of the benefits of the respondents' work. The former element requires that there should be "a sufficient degree of objective similarity between the two works": see Francis Day & Hunter Ltd. v. Bron (1963) Ch. 587, at p.614; Lend Lease Homes Pty. Ltd. v. Warrigal Homes Pty. Ltd. (1970) 3 N.S.W.R. 265, at p.273; S.W. Hart & Co. Pty. Ltd. v. Edwards Hot Water Systems [1985] HCA 59; (1985) 59 A.L.J.R. 729, at p.730; [1985] HCA 59; 61 A.L.R. 251, at p.253. It is impossible to say that there is any objective similarity between the ROMs and EPROM on the one hand and the written source programs on the other. Neither the silicon chips nor the electrical impulses that may be generated in them have the slightest resemblance to the written source programs. The Wombat ROMs and EPROM may well be reproductions of the Apple ROMs but they are not reproductions of the Applesoft and Autostart source programs. The ROMs and EPROM give effect to the same idea as that which is expressed in the source programs but do not resemble the source programs. It is established that a work which is manufactured in accordance with written instructions is not a reproduction of those instructions. Thus it was held in Cuisenaire v. Reed [1963] VicRp 96; (1963) V.R. 719 that a set of coloured rods did not infringe copyright in a book which described a method of teaching mathematics with the aid of such rods, and Pape J. said, at p.734, that "a set of written directions is not 'reproduced' by the construction of an article made in accordance with those directions". Similarly, in Boosey v. Whight it was held that perforated rolls used to produce particular melodies did not infringe the copyright in sheets of music on which the score of those melodies was written and in Brigid Foley Limited v. Ellott (1982) R.P.C. 433 it was held that knitted garments produced by following the instructions in a knitting guide would not be reproductions of the knitting guide itself. The ROMs and EPROM embody the idea, and logical structure, of the source programs, but do not reproduce the expression of the idea and of the logical thought which is to be found in the source programs.

18. The questions involved in this case are of considerable
difficulty, and they have given rise to differences of opinion in the writings on the subject. We were referred to some authorities which are contrary to the conclusions that I have reached. In the United Kingdom, the provisional opinion has been expressed by judges in interlocutory proceedings that a machine code program is either a reproduction or an adaptation of an assembly code program: Sega Enterprises Ltd. v. Richards (1983) F.S.R. 73; Thrustcode Ltd. v. W.W. Computing Ltd. (1983) F.S.R. 502. In Canada and in South Africa it has been held that a program in an object code may be a literary work: I.B.M. Corporation v. Ordinateurs Spirales Inc. (1984) 80 C.P.R. (2d) 187; La Societe D'Informatique R.D.G. Inc. v. Dynabec Ltd. (14 August 1984; unreported); Northern Micro Computers v. Rosenstein (1982) F.S.R. 124. There are some differences between the language of the statutes considered in those cases and that of the Copyright Act, but I am not sure that the differences are material. In the United States the question has been considered in a number of cases and it has been held that an object code is entitled to copyright protection (see particularly Apple Computer Inc. v. Franklin Computer Corporation [1983] USCA3 1033; ; (1983) 714 F.2d. 1240 and Apple Computer Inc. v. Formula International Inc. [1984] USCA9 205; (1984) 725 F.2d. 521) but the statute law in the United States is materially different from that of Australia. I have not found anything in any of those authorities that has persuaded me that a sequence of electrical impulses in a silicon chip, not capable itself of communicating anything directly to a human recipient, and designed only to operate a computer, is itself a literary work, or is the translation of a literary work, within the Copyright Act. It may be regretted that the respondents have no remedy in copyright against the appellants who pirated their programs. However although it would be no doubt right to give the Copyright Act a liberal interpretation, it would not be justifiable to depart altogether from its language and principles in an attempt to protect the products of scientific and technological developments which were not contemplated, or only incompletely understood, when the statute was enacted. To keep copyright law abreast of developing technology is no new problem, as the decision in Boosey v. Whight shows. The provisions of s.21(1) and Pt.IV of the Copyright Act were intended to take account of technological developments unknown when the earlier copyright acts came into force. The Copyright Amendment Act 1984 similarly attempts to make the law adequate to cope with modern developments. Important questions of policy arise when it becomes necessary to consider whether copyright protection should be extended to such a thing as a ROM, and the courts can act only within the existing statutory framework.

19. Since the respondents must fail entirely in the
proceedings the question of knowledge raised by the cross-appeal does not arise.

20. I would allow the appeal, dismiss the cross-appeal and
restore the orders made by the learned trial judge.

MASON AND WILSON JJ.: The respondents instituted proceedings in the Federal Court of Australia seeking relief, inter alia, in respect of the alleged infringement by the appellants of the copyright of the first respondent in two computer programs. The programs are those described as Applesoft object program and Autostart object program respectively. In their primary form each of these programs consists of a series of electrical impulses which initiate and control the basic operations of the computer. They are contained in six ROMs fitted into the respondents' Apple II computer. The computer is marketed in Australia by the second respondent.

2. The appellants import into Australia for the purpose of sale a computer known as "Wombat". It is manufactured in Taiwan. The Wombat is similar in appearance to the Apple II computer and its programs are represented as being compatible with those of the Apple II. The similarity extends even further. The Wombat has fitted within it two ROMs and one EPROM which together contain programs which are virtually identical to the Applesoft object program and the Autostart object program. The finding of the learned trial judge that the former programs were in fact copied from the latter is not contested by the appellants. It is this copying that raises the issues that are now to be determined.

3. At first instance the respondents failed to persuade Beaumont J. that their object programs were protected by copyright in accordance with the Copyright Act 1968 (Cth), as amended ("the Act"). His Honour held that neither these object programs nor the written source programs from which they were derived were original literary works capable of attracting copyright protection. Subsequent to the decision the Parliament enacted the Copyright Amendment Act 1984 (Cth) in order to ensure a degree of protection to computer programs in the future but the terms of that enactment are not material to the present proceedings. It does mean, however, that similar issues that may arise in future proceedings will have a very different statutory context for their resolution. In the meantime the respondents appealed successfully to the Full Court of the Federal Court. It was held by a majority (Fox and Lockhart JJ., Sheppard J. dissenting) that the programs contained in the Wombat ROMs and EPROM were reproductions of the programs contained in the Apple II ROMs, which were themselves adaptations of the Applesoft source program and the Autostart source program respectively. The Full Court was unanimous in concluding that these source programs were original literary works, but divided on the adaptation issue. On the question of infringement, the issue in the Full Court was whether the appellants had the knowledge necessary to constitute infringement under ss.37 and 38 of the Act, namely, the knowledge that the making of the ROMs and EPROM used in the Wombat computer would have constituted an infringement of the respondents' copyright if they had been made in Australia by the importer. A belated attempt by the respondents to establish infringement by reference to s.36 of the Act was rejected by the Full Court on the ground that the issues arising under that section had not been fully canvassed at the trial. A further attempt in this Court to rely upon s.36 failed. In the result the Full Court held that the appellants had the requisite knowledge at least from the commencement of the proceedings and that injunctions should be granted on a quia timet basis, with the proviso that if a claim for damages for past infringements was to be pursued the matter should be referred back to the trial judge for findings concerning prior knowledge.

4. Special leave to appeal having been granted, the appellants now appeal to this Court. The respondents cross-appeal on the issue of knowledge and ask that the matter be referred back to Beaumont J. for an assessment of damages or, at the election of the cross-appellants, an order for an account of profits.

5. Before leaving the decision of the Full Court, it is to be observed that each of the reasons of their Honours includes, in clear and simple language, a description of the operation of a computer sufficient to provide the background to a determination of the issues in the present case. The decision has been reported: [1984] FCA 137; (1984) 53 A.L.R. 225. The descriptions are to be found at pp.229-234 (Fox J.), pp.247-253 (Lockhart J.) and pp.271-274 (Sheppard J.). We do not find it necessary to attempt a similar exposition and express our gratitude to their Honours for relieving us of any obligation to do so. Again, we do not find it necessary to repeat the detailed statement of agreed facts which is conveniently appended to the reasons of Fox J., commencing at p.242. We are broadly in agreement with the reasons and conclusions of Fox J. and Lockhart J. and will attempt to avoid undue repetition.

6. The relevant programs of the respondents were developed in this way. In 1977 Mr Wigginton, who was employed by the first respondent, obtained a copy of the Microsoft Basic program from another manufacturer and set about modifying it. The modification involved considerable skill, time and effort. The entire program was written out and with suitable amendment re-expressed in the shorthand form of language called "6502 Assembly Code". This code consists of mnemonics, each comprising three letters of the alphabet. It is a language which can be readily understood by persons skilled in its use. The program so devised by Wigginton was called "Applesoft" and when expressed in 6502 Assembly Code was described as the Applesoft source program. In its original written form the Applesoft source program included both labels and comments to assist the programmer in converting the program into the Applesoft object program. This conversion was achieved by using a computer to reduce the 6502 Assembly Code in which the Applesoft source program was expressed into machine language consisting of a series of electrical impulses which were then able to be stored on a magnetic disc or tape or permanently installed in the Apple II computer on ROMs. The ROMs are embodied in silicon chips. After its conversion into the Applesoft object program, the Applesoft source program was destroyed and whilst it is possible to extract from the computer a reconstruction of the mnemonics that were employed it is not possible to reproduce the labels and instructions. It is also possible to produce a written description of the object program, expressed either in binary form or in a form of language known as hexadecimal notation.

7. Whether it is the series of electrical impulses stored in the ROMs or the written description in binary or hexadecimal notation that is truly the object program is not entirely clear. However, the case has been conducted on the footing that it is the former rather than the latter (see the judgments of Fox and Sheppard JJ. in the Federal Court, at pp.232 and 273 respectively, which seem to reflect pars. 22 and 22A of the agreed statement of facts at p.244, but contrast that of Lockhart J. at pp.250-251). We propose to deal with the case on this footing as no issue is raised with respect to it in this Court.

8. A similar history attends the creation of the Autostart source program. It was devised in 1977 by Mr Wozniak and later modified by Mr Arkley in 1978. Wozniak was an officer of the first respondent and Arkley was one of its employees. The Autostart source program was then reduced with the aid of a computer into machine language and in that form became known as the Autostart object program.

9. In substance, the case for the respondents is that each of the object programs is an original literary work in which copyright subsists and of which copyright the first respondent is the owner. Alternatively, it is said that each of the source programs is an original literary work and that the object programs are translations and therefore adaptations of the respective source programs. In that character also the object programs are the subject of copyright. The respondents assert that the Wombat ROMs and EPROM are a reproduction of the programs embodied in the Apple II ROMs, namely, the Applesoft object program and the Autostart object program and that, having regard to ss.37 and 38 of the Act, there has been infringement. As we have said, the Full Court was unanimous in holding that the two source programs were original literary works within the meaning of s.32 of the Act and the majority held also that the object programs were adaptations respectively of the source programs. Neither Fox J. nor Lockhart J. found it necessary to determine whether the object programs were themselves original literary works but Sheppard J. held that they were not. In addition the majority held that in the case of each program the Wombat ROMs and EPROM were reproductions in a material form of an adaptation of an original literary work and hence within s.31(1)(a)(i) and (vii) of the Act. The Court was not required to consider whether the Wombat ROMs and EPROM were themselves a reproduction in a material form of the Applesoft and Autostart source programs.

10. The appellants challenge each of the findings of Fox J. and Lockhart J. In the first place, it is argued that the source programs were not original literary works because in order to satisfy that description the work must afford either information and instruction or pleasure in the form of literary enjoyment to human beings. Reliance is placed on the recent decision of the Court of Appeal in England in Exxon Corporation v. Exxon Insurance Ltd. (1982) Ch 119 and the approval expressed in that decision of the observations of Davey L.J. in Hollinrake v. Truswell (1894) 3 Ch 420, at p 428:

"Now, a literary work is intended to afford either
information and instruction, or pleasure, in the
form of literary enjoyment".

Even if this statement was adopted, for the purpose of applying the current legislation, as an exhaustive definition of a literary work, it seems to us that the appellants would be hard pressed to deny that the source programs, expressed in 6502 Assembly Code, were intended to afford information and instruction. In any event, we agree with the view expressed by all the members of the Full Court that nothing said in the cases we have mentioned was intended to establish a comprehensive or exhaustive definition of a literary work for copyright purposes. Certainly, the actual decision in each of those cases affords no assistance to the appellants. In Hollinrake v. Truswell, copyright protection was denied to words and figures "inscribed on and necessarily forming part of an apparatus or tool" (per Lord Herschell L.C., at p.424). In Exxon Corporation, the Court of Appeal denied the character of a literary work to the single word "Exxon" standing alone. The short reason for that conclusion was expressed by the learned judge at first instance, Graham J., whose decision was affirmed by the Court of Appeal, when his Lordship said at p.130:

"It is a word which, though invented and therefore
original, has no meaning and suggests nothing in
itself".

11. The Act defines a literary work to include "a written table or compilation" (s.10(1)). In University of London Press, Limited v. University Tutorial Press, Limited (1916) 2 Ch 601, Peterson J., after referring to a similar reference to "literary work" in s.35 of the Copyright Act 1911 (U.K.) said, at p.608:

"In my view the words 'literary work' cover work
which is expressed in print or writing,
irrespective of the question whether the quality or
style is high. The word 'literary' seems to be
used in a sense somewhat similar to the use of the
word 'literature' in political or electioneering
literature and refers to written or printed
matter".

Later in his judgment (at pp.609-610), his Lordship added:

"The objections with which I have dealt do not
appear to me to have any substance, and, after all,
there remains the rough practical test that what is
worth copying is prima facie worth protecting".

In the result, examination papers were held to be the subject of copyright. The observations of Peterson J. were referred to with approval by Lord Reid, Lord Hodson and Lord Pearce in the football coupon case - Ladbroke (Football) Ltd. v. William Hill (Football) Ltd. (1964) 1 WLR 273, at pp 285, 291; (1964) 1 All ER 465, at pp 475, 479.

12. We have no hesitation in coming to the conclusion that each of the source programs was an original literary work. The question is to be answered by reference to their form and character at the time of their completion by Messrs Wigginton and Arkley respectively. Whether or not a literary work must be in writing, these programs were written. Although the substance of the program in each case was expressed in 6502 Assembly Code, this is a language which was readily intelligible to anyone versed in computer science. Each program was the product of skill, time and effort. It was a particular kind of vehicle for the communication of useful information to persons who may desire it. The fact that its creation was a step towards the goal of facilitating the operation of a computer does not warrant its dismissal by the appellants as no more than a mere adjunct to the operation of a mechanical device, by analogy with the sleeve which was the subject of the decision in Hollinrake v. Truswell. On any view, in the form in which it was created and before it was transformed into another medium, each source program had an existence which was entirely independent of the machine. It was capable of conveying meaning as to the arrangement and ordering of instructions for the storage and reproduction of knowledge. In that form it was entitled to copyright protection.

13. The second submission of the appellants is that even if the source programs are the subject of copyright, the object programs are not. It is said that the members of the Full Court who formed the majority erred in holding each object program to be a "translation" of the relevant source program and therefore an adaptation of a literary work within the meaning of that term as defined in s.10(1). The arguments advanced in support of the submission may be summarized as follows:

(a) the object programs are merely sequences of

electrical impulses within the computer;

(b) a translation, in order to constitute an

adaptation, must be a translation from one human language to another;

(c) a translation of a literary work must itself be a

literary work;

(d) a translation of a literary work must convey a

clear idea of the literary work to human beings.

14. It is not correct to describe an object program as merely a sequence of electrical impulses within the computer. Electrical impulses there are, but those impulses serve to identify a set of instructions in machine readable language designed to guide the machine in its basic operations. They do not form part of the computer itself, notwithstanding that they may be embodied in a ROM or ROMs located permanently in silicon chips in a machine. They might equally well be contained in a magnetic disc or tape which can be inserted into or removed from the machine at will. The language may be recorded in visible form in binary or hexadecimal notation. The same result as is achieved by the use of an object program could be achieved by the manual operation of a computer in accordance with instructions written out in ordinary English and contained in a manual. The only problem would be that such a process would be impossibly tedious and therefore wholly impracticable.

15. The next question is whether the object program is a translation of the source program. The word "translation" is not defined in the Act. One of the meanings given in the Shorter Oxford English Dictionary is "the expression or rendering of something in another medium or form". In the Federal Court, Sheppard J. was of the opinion that a translation did not encompass something which could neither be seen nor heard. With all respect to his Honour, we are unable to share that view. True it is that the object program can be neither seen nor heard, but there can be no doubt that it exists and one knows where it can be found and how it can be activated. It seems to us that the conversion of a source code into an object code, albeit by the mechanical operation of a computer, is a process which effects a "translation" of the source code from one language into another language. It is immaterial that the last- mentioned language takes a form which can only be read and understood in that form by a machine. It is an adaptation of a literary work. This conclusion finds support in the statement from the 1979 report of the United States National Commission on New Technological Uses of Copyright Works at p.21, n.109, which was cited by Fox J. at p.235 of his Honour's judgment in the Federal Court:

"A source code is a computer program written in any
of several programming languages employed by
computer programmers. An object code is the
version of a program in which the source code
language is converted or translated into the
machine language of the computer with which it is
to be used".

We therefore reject the argument of the appellants that the only relevant translation is that which translates one human language into another human language. We also find no warrant for the argument that a translation must convey a clear idea of the literary work to human beings although, if it matters, there is no reason why an object program may not be reduced to writing and understood by those skilled in the art.

16. The question whether an adaptation of a literary work must itself be a literary work is more difficult of resolution. Fox J. (at pp.235-236) considered it to be plain from the language of the Act itself that the question must be answered in the negative, and he did not find it necessary to consider whether the Apple II object codes could be regarded as literary works. Lockhart J. did not address the question. His Honour, however, did confess to having considerable reservations about the correctness of the view that object programs are original literary works (a question which he did not have to decide), but we would understand that to be a different question from the one we are presently considering. Section 31(1)(a)(vii) provides that copyright, in relation to a literary work, is the exclusive right, inter alia, to reproduce in a material form a work that is an adaptation of that literary work. "Work" is defined in s.10(1), unless the contrary intention appears, to mean inter alia a literary work. Prima facie, therefore, an adaptation of a literary work must itself satisfy the description of a literary work. Of course, one would expect that an adaptation, in this case a translation of the original literary work constituted by the source codes, would continue to display, in its new form, the characteristics which enabled the original work to satisfy that description. But it is argued for the appellants that that is not so in the present case because a literary work must be in writing. Ordinarily and traditionally it is no doubt true that a literary work would take a written form. But the Act does not require it to be so. Indeed, s.22(1) of the Act identifies the time when a work is made as the time when "the work was first reduced to writing or to some other material form" (our emphasis). See also s.21 of the Act. There seems to be no reason to doubt that a literary work is made and entitled to copyright protection from the time it is first recorded on tape, if that be the first material form that the work takes. In our opinion, an object code, although brought into existence by mechanical means, takes on the same literary character as is possessed by the source code from which it is derived. This conclusion seems necessarily to follow, if the protection secured by the Act to the source programs as original literary works is to be effective. If there is no copyright in the object programs which are a natural and necessary derivative of the source programs then there is no point in protecting the source programs.

17. The third submission of the appellants is that the Federal Court erred in holding that the Wombat ROMs and EPROM were reproductions of the Apple II ROMs which contained the Applesoft object program and the Autostart object program. Once it is accepted, first, that the object programs contained in the Apple II ROMs are adaptations of original literary works and, secondly, that the object programs contained in the Wombat ROMs and EPROM were copied from those adaptations, it would seem to follow that the Wombat computers contain a reproduction in a material form of an adaptation of an original literary work. It is sufficient if the reproduction relates not to the whole but to a substantial part of the work: s.14(1). The fact that the Apple II programs are contained in six ROMs in the Apple II computer but in only two ROMs and one EPROM in the Wombat computer is immaterial. The distinction merely represents an advance in computer technology. It is argued for the appellants that the Wombat ROMs and EPROM are not reproductions but are merely manufactured in accordance with instructions which can be derived from the literary work. But this is to misunderstand what it is that is alleged to have been reproduced. It is unnecessary to decide whether the Wombat ROMs and EPROM could be said to be reproductions in a material form of the original literary works, namely, the Applesoft and Autostart source programs. The allegation here is that the appellants have reproduced directly the Applesoft and Autostart object programs themselves. Those programs are entitled to protection as adaptations of the original literary works. The appellants sought support for the submission in the decision of Pape J. in Cuisenaire v. Reed [1963] VicRp 96; (1963) VR 719, and referred particularly to pp 734-736, where his Honour was dealing with a claim by the plaintiff for infringement of literary copyright in certain tables and compilations set out in a book published by the plaintiff which described a new method of teaching primary school children mathematics with the aid of rods of varying lengths and colours. The infringement was alleged to rest, not in the reproduction of the tables and compilations themselves, but in the production of the rods. The plaintiff failed. At p.734, Pape J. said:

"In my view, a set of written directions is not
'reproduced' by the construction of an article made
in accordance with those directions. A
reproduction must reproduce the original, and here
the original is in one case a set of words in the
form of a table and in the other case a series of
plain and coloured circles which are numbered and
which are set out in the form of a chart. In my
view, the defendants' rods reproduce neither".

The facts of which his Honour was speaking bear no resemblance to the facts of the present case and we are unable to derive from the decision any assistance which is material to the present case. Here the Wombat ROMs and EPROM contain in substance the same sequence of electrical impulses as are contained in the Apple ROMs. In our opinion, the Federal Court was correct in its decision on the reproduction issue.

18. It remains now to consider the cross-appeal. The respondents (cross-appellants) argue that on the evidence before the trial judge the only proper conclusion is that the first appellant through its managing director the second appellant knew from the time of first importation that if the Wombat ROMs and EPROM had been made in Australia they would have constituted an infringement of the copyright in the Applesoft and Autostart object programs. The difficulty is that the primary judge did not make any findings of fact on the question of the second appellant's knowledge. The answer to that question could well turn on the credibility attaching to the evidence given by the second appellant. Both Fox J. and Lockhart J. considered that in the absence of the necessary findings of the trial judge on credibility it would not be right for an appellate tribunal to draw the inferences necessary to sustain the respondents' argument. We agree.

19. We would dismiss both the appeal and the cross-appeal.

BRENNAN J.: The first respondent, Apple Computer, Inc. ("Apple") is a Californian corporation which makes and sells computers. The second respondent, a New South Wales company, is the Australian distributor of computers made by Apple. The computer model which is relevant to these proceedings is a micro-computer sold under the name "Apple II". The circuitry of the Apple II includes six silicon chips which, in conjunction with the other parts of the computer, give it a capacity to perform particular functions. The silicon chips are known, in computer terminology, as ROMs (an acronym for "read only memory").

2. The first appellant, Computer Edge Pty. Limited ("Computer Edge"), is a Victorian company which imported and sold in Australia computers made in Taiwan. These computers were sold under the name "Wombat". The circuitry of the Wombat includes three silicon chips which, in conjunction with the other parts of the Wombat, give it a capacity to perform functions identical with those performed by Apple II. The electrical operation of the silicon chips in the Wombat is very similar to, if not identical with, the electrical operation of the six ROMs in the Apple II. Although there is no direct evidence as to the way in which the Wombat silicon chips were produced, it is probable that they were made by electrical machinery which reproduced on blank silicon chips the same circuitry as the circuitry on the Apple II ROMs. At all events, there is no doubt that the circuitry of the Wombat silicon chips was derived from the circuitry of the Apple II ROMs. Two of the Wombat silicon chips are ROMs, the other is an EPROM (an acronym for "erasable programmable read only memory"). Nothing turns on the difference in number or description of the silicon chips to be found respectively in the Apple II and the Wombat. That is because the critical question in this appeal is whether, if Computer Edge had produced the Wombat silicon chips in Australia, their production would have constituted an infringement of Apple's copyright under the Copyright Act 1968 (Cth) ("the Act"). That question arises under each of ss.37 and 38 of the Act, upon which the respondents relied in proceedings brought against Computer Edge and its managing director, the second appellant, in the Federal Court of Australia.

3. Section 37 of the Act reads as follows:

" The copyright in a literary, dramatic, musical
or artistic work is infringed by a person who,
without the licence of the owner of the
copyright, imports an article into Australia for
the purpose of -

(a) selling, letting for hire, or by way of
trade offering or exposing for sale or
hire, the article;

(b) distributing the article -

(i) for the purpose of trade; or

(ii) for any other purpose to an extent
that will affect prejudicially the
owner of the copyright; or

(c) by way of trade exhibiting the article in
public,

where, to his knowledge, the making of the
article would, if the article had been made in
Australia by the importer, have constituted an
infringement of the copyright."

Section 38 contains a similar provision relating to infringement by sale of, and by other dealings with, imported articles.

4. Apple claims copyright in what are respectively called "source programmes" and "object programmes" (the nature of which will be examined presently) which are said to be "literary works" within the meaning of that term in the Act. At the time relevant to these proceedings, that is, before 15 June 1984 when the Copyright Amendment Act 1984 came into operation, s.10 of the Act defined a "literary work" to include "a written table or compilation". (The 1984 Act amended the definition of "literary work" to include "a computer program or compilation of computer programs".) Beaumont J. at first instance held that none of the source programmes and object programmes was a literary work. He dismissed Apple's claim. On appeal, the Full Court (Fox, Lockhart and Sheppard JJ.) held unanimously that the source programmes were literary works in which copyright subsisted. Fox and Lockhart JJ. did not find it necessary to decide whether copyright subsisted in the object programmes, while Sheppard J. thought it did not.

5. A source programme is compiled in writing by a computer scientist or technician for the purpose of making components of a computer that will give the computer a capacity to function in a desired way. It is written in letters of the alphabet. When it is compiled, it is keyed into a computer. Each keystroke generates a particular series of electrical impulses. The computer into which a source programme is keyed is equipped with an assembler which converts each series of electrical impulses into a different series of electrical impulses. The latter series consist in a sequence of "on" or "off" electrical charges which cause a computer's micro-processor to function. The sequences of "on" or "off" electrical charges produced by an assembler when a source programme is keyed into the computer is called an "object programme". The electrical charges constituting an object programme can be recorded as magnetic fields on a magnetic tape or disc. The conventional way of describing a sequence of electrical charges on most computer components is by using the numerals "0" and "1" to denote "on" and "off" (binary notation). A sequence of four "ons" and "offs" may be in any of 16 possible combinations (0000, 0001, 0010, etc). A shorthand notation has been developed which allows each combination to be identified by a numeral in the series "0" to "9" or by a letter in the series "A" to "F". There being sixteen characters - 10 numbers and 6 letters - in this alphanumeric series, it is described as "hexadecimal notation". By using hexadecimal notation the lengthy sequences of "0's" and "1's" which are descriptive of the "on" or "off" electrical charges in an object programme can be shortened. Binary notation and hexadecimal notation are conventional ways of representing in writing an object programme, but the object programme is not the writing. The object programme is the sequence of electrical charges and its existence and use are not dependent on a written representation.

6. The relationship between the source programme and the object programme is fixed by the assembler, and that relationship is known to and used by the compiler of the source programme who so compiles the source programme that the desired object programme will be produced when the source programme is keyed in. The assembler used in creating the object programmes from the source programmes in which Apple claims copyright is known as a "6502 assembler", and the source programmes were therefore written in "6502 Assembly Language", to adopt the terminology of computer science. The source programmes for which copyright is claimed are known as Applesoft and Autostart. They were compiled by computer scientists employed by Apple. If those compilations are susceptible of copyright, Apple became the owner of that copyright when the compilations were made. These source programmes were in writing.

7. When the computer into which the source programmes were keyed produced the sequences of electrical charges which constituted the object programmes, no written compilation of the object programmes in hexadecimal or binary notation was made. At all events, there was no evidence to suggest that the making of the Wombat ROMs and EPROM involved any step derived from a written compilation of the object programmes.

8. The ROMs which are included in the circuitry of Apple II were made by a process which commenced with the keying in of the Applesoft and Autostart source programmes into a computer and the creation of the electrical charges which constituted the corresponding object programmes. The object programmes may have been recorded on magnetic tapes or discs. Then the object programme (either directly or from a magnetic tape or disc on which it had been recorded) was recorded upon blank silicon chips. The process of recording the object programme involves the application of the sequence of electrical charges to the chip, so that minute fusible connectors in the chip are burned out or charges are applied to minute insulated capacitors in the chip. The recording of the programme is practically irreversible. A printout in hexadecimal or binary notation of an object programme contained in the ROMs can be recovered from a computer in which the ROMs are installed. A printout of the text of a source programme (other than instructions which are not keyed in) can also be recovered.

9. As the circuitry of the Wombat ROMs and EPROM was derived from the Apple II ROMs, probably by an electrical process which reproduced the object programme recorded on the Apple II ROMs, the question for determination is whether, if that process had been carried out by Computer Edge in Australia, it would have constituted an infringement of Apple's copyright in the Applesoft and Autostart source programmes or in the corresponding object programmes. Both source programmes were first published in the United States of America. The corresponding object programmes were first produced there. The provisions of the Act apply, in relation to literary works first published in the United States of America in like manner as they apply in relation to literary works first published in Australia (s.184(1) of the Act and reg.4 of the Copyright (International Protection) Regulations. Computer Edge submits that the source programmes are no more than instructions for the operation of a computer and that neither the source programmes nor the corresponding object programmes are "literary works" susceptible of copyright under the Act.

10. A literary work need not have literary merit (Robinson v. Sands & McDougall Proprietary Ltd. [1916] HCA 51; (1916) 22 CLR 124, at p 133). The words "literary work", as Peterson J. pointed out in University of London Press, Limited v. University Tutorial Press, Limited (1916) 2 Ch 601, at p 608, "cover work which is expressed in print or writing, irrespective of the question whether the quality or style is high". A "literary work", according to Davey L.J. in Hollinrake v. Truswell (1894) 3 Ch 420, at p 428, is a work "intended to afford either information and instruction, or pleasure, in the form of literary enjoyment". Although that case was decided under the Copyright Act of 1842, the observation of Davey L.J. was adopted by the Court of Appeal in Exxon Corporation v. Exxon Insurance Ltd. (1982) 1 Ch 119, at pp 143,144; (1982) RPC 69, at pp 88,89. The observation is not unduly restrictive. If the print or writing in which the work is expressed conveys information or instruction, albeit to a limited group with special knowledge,it is immaterial that the information or instruction is not expressed in the form of words, phrases or sentences. Thus a telegraphic code has been held to be a literary work though the words of the code were meaningless in themselves: D.P. Anderson & Co.Limited v. Lieber Code Company (1917) 2 KB 469. The source programmes were the product of substantial originality and skill, they were prepared as instructions for the manufacture of Apple II ROMs, they were in writing and they conveyed meaning at least to computer scientists and technicians. That is sufficient to bring them within the scope of literary copyright. The argument that a source programme is not a literary work proposes a dichotomy between literary works and what are "merely adjuncts to the operation of a mechanical device". The dichotomy is false, for it cannot be said that every book of instructions for the operation of a machine is outside the scope of copyright protection. Copyright might not protect every minor direction for use of a machine or device, but that is not the present case. Here, the source programmes are a set of complex and precise instructions for the manufacture of Apple II ROMs. The source programmes are substantial original compilations. It would be an infringment of Apple's copyright in a source programme to do in relation to that programme in Australia without Apple's licence any of the acts prescribed in s.31(1)(a) of the Act.

11. The corresponding object programmes, on the other hand, are not in writing. The electrical charges which constitute them may be represented in writing, using the conventional binary or hexadecimal notations. However, the written representation must not be confused with what is represented but not written. Copyright protection under Part III of the Act is given only to works; under Part IV of the Act it is given to subject matter other than works. The object programmes do not come within Part IV. To be a "work" within Part III, an object programme must be one of the four classes of works as defined by s.10(1) - literary, dramatic, musical or artistic. Clearly an object programme is not a dramatic, musical or artistic work. To come within the scope of "literary work" as that term was defined in the Act prior to the 1984 amendment, a literary work had to be in print or writing. Peterson J. in University of London Press, at p.608, said:

" Copyright Acts are not concerned with the
originality of ideas, but with the expression of
thought, and, in the case of 'literary work,'
with the expression of thought in print or
writing. The originality which is required
relates to the expression of the thought."

And in Ladbroke (Football) Ltd. v. William Hill (Football) Ltd. (1964) 1 WLR 273, at p 287; (1964) 1 All ER 465, at p 477, Lord Hodson regarded it as the undoubted truth that "copyright is not concerned with the originality of ideas, but with the expression of thought, in the case of literary work with the expression of thought in print or writing".

12. Counsel for Apple nevertheless submitted that print or writing is not essential to the existence of a literary work, arguing that it is sufficient that the work be reduced to some material form. Some support for that submission may be found in s.22 of the Act which prescribes the time of the making of "a literary, dramatic or musical work" as the time when "the work was first reduced to writing or to some other material form". It may be that the draftsman of s.22 regarded Braille as "some other material form" rather than as "writing", for writing is ordinarily read by sight but Braille is read by touch. It is unnecessary to speculate on that point. Section 22 implies that the form in which a literary work is expressed is writing or some other material form. A material form is a form which can be perceived by the senses. That is not to say that copyright subsists in a document on which the writing of an original literary work appears (In re Dickens, Dickens v. Hawksley (1935) Ch 267). Copyright in a work is a right in respect of the form to which the work has been reduced, not a right of property in the object in or from which the form is perceptible. But a form from which the words, letters or figures of a literary work cannot be perceived by sight or touch (or, possibly, hearing) is not a material form to which the work has been reduced. The electrical charges which constitute the object programmes cannot be seen or touched or heard or, if they can, they do not communicate the letters of the original literary work, the source programmes. Nor, for that matter, do the electrical charges communicate the letters or figures by which an object programme may be represented. The object programmes are not literary works. Apple's right to relief, if any, must depend on its ownership of the copyright in the Applesoft and Autostart source programmes.

13. As the owner of the copyright in those programmes, Apple had the exclusive right in Australia to do all or any of the acts specified in s.31(1)(a) of the Act, that is to say -

" (i) to reproduce the work in a material form;

(ii) to publish the work;

(iii) to perform the work in public;

(iv) to broadcast the work;

(v) to cause the work to be transmitted to
subscribers to a diffusion service;

(vi) to make an adaptation of the work;

(vii) to do, in relation to a work that is an
adaptation of the first-mentioned work, any
of the acts specified in relation to the
first-mentioned work in sub-paragraphs (i)
to (v), inclusive".

Sub-paragraphs (i) and (vii) are relied on by Apple to establish infringement of its copyright in the source programmes. The argument which succeeded in the Full Court was that the making of the Wombat ROMs and EPROM was a reproduction in a material form of the object programmes which are adaptations of the corresponding source programmes. The term "adaptation" is defined by s.10(1)(c) of the Act to mean, in relation to a literary work, either -

" ...

(i) a translation of the work; or

(ii) a version of the work in which a story or
action is conveyed solely or principally by
means of pictures".

A majority of the Full Court (Fox and Lockhart JJ, Sheppard J. dissenting) held that the object programmes were translations and therefore adaptations of the source programmes. If the object programmes are adaptations (that is, translations) of the corresponding source programmes, Apple's claim could be put on alternative bases: Apple could seek to show that the making of the Wombat ROMs and EPROM was a reproduction either of the adaptations of the source programmes or of the source programmes themselves. The majority of the Full Court found that there was a reproduction of the "adaptations". It was therefore unnecessary for their Honours to decide whether there was a reproduction of the source programmes themselves.

14. In the Oxford Dictionary the relevant meanings of translation are given as follows:

" The action or process of turning from one
language into another; also, the product of this;
a version in a different language"; and

" The expression or rendering of something in
another medium or form, e.g. of a painting by an
engraving or etching".

The action or process of turning the source programme in "6502 Assembly Language" into the object programme in what is called "machine readable language" is said to be a translation. It strains the meaning of "language" to include within its denotation the alphabetic symbols in which the source programmes were written. Although the alphabetic symbols of the source programmes may be classified as literary in form for the purposes of the Act, I would not describe those symbols as a language, the primary meaning of which is "the whole body of words and of methods of combining them used by a nation, people or race" (Shorter Oxford Dictionary). It is difficult to divorce "language" from human speech, and a means of communicating ideas which does not consist of words is not properly to be described as a language. A chemical formula or a mathematical equation are apt to communicate ideas, but I should not classify either as a language. "6502 Assembly Language" is not a language; it is a code. Even if the code were thought to be susceptible of "translation" into another language, it was not so translated: the electrical charges which constitute the object programmes are clearly not a language. To describe such electrical charges - no doubt helpfully enough for the purposes of computer science - as "machine readable language" is to make metaphor serve as reality. The machine has no comprehension of thought which it is the essential purpose of language to convey, and the fact that a micro-processor is activated by a sequence of electrical charges in a predictable way does not mean that it has understood and executed some command.

15. The secondary meaning of translation comes closer to describing the connection between the sequences of electrical charges constituting the source programmes and corresponding object programmes. Like the messages sent by a morse telegraphist, the object programmes might be regarded as an expression of the original text in a different medium: a translation of a written text into a sequence of electrical charges. The fact that an object programme will so operate upon a suitably programmed computer as to cause it to print out either the original written text (the source programme) or a binary or hexadecimal notation of the electrical charges (the object programme) makes it arguable (I need put it no higher) that an object programme is an expression of a source programme in another medium and is a "translation". This approach, freed from the confusion arising from the metaphorical use of "language", might command acceptance if it were not for the necessity of the "adaptation" being itself a "work". In the Full Federal Court, the majority did not think it necessary to decide whether an "adaptation" (that is, a "translation") was a "work" but I respectfully disagree. The language of s.31(1)(a)(vii) is clear: "a work that is an adaptation". We have seen that a sequence of electrical charges constituting an object programme is neither a literary work nor any other kind of "work". An object programme is therefore not "a work that is an adaptation" of the source programme in which copyright subsists.

16. The alternative basis of Apple's claim rests on sub-par.(i) of s.31(1)(a) of the Act. If the recording of the object programmes on the silicon chips included in the Wombat computers "reproduced" the copyright source programmes in a material form, it would have been an infringement by Computer Edge of Apple's copyright in the source programmes to make the Wombat ROMs and EPROM in Australia without the licence of Apple (s.36(1)).

17. An exclusive right to reproduce a work in which copyright subsists has been an aspect of copyright created by successive Copyright Acts. Each Act differed from the others, of course, in the terms and context by and in which that exclusive right was conferred. Under the Fine Arts Copyright Act 1862 (Imp.), the exclusive right "of copying, engraving, reproducing and multiplying" an original painting or drawing "by any means and of any size" was held not to be infringed by the representation of a painting by a grouping of figures in a tableau vivant: Hanfstaengl v. Empire Palace (1894) 2 Ch 1. The Copyright Act 1911 (U.K.), adopted subject to modifications by the Copyright Act 1912 (Cth), adopted a different formulation of the right. Section 1(2) of the 1911 Act conferred "the sole right to produce or reproduce the work ... in any material form whatsoever". That provision was thought to give wider protection to authors than that given by the 1862 Act and to remedy the effect of the decision in Hanfstaengl v. Empire Palace: see Bradbury, Agnew & Company v. Day (1916) 32 TLR 349; Chabot v. Davies (1936) 3 All ER 221; King Features Syndicate, Inc. v. O. & M. Kleeman, Ld. (1940) Ch 806, at pp 816,827; (1941) AC 417, at pp 432,435. Under the 1911 Act, the sole right of reproduction of a work is infringed by its unlicensed reproduction in a material form different from the form in which the original work is made. The widening of protection was particularly significant for the protection of artistic works. The sole right to reproduce an artistic work was infringed by the unlicensed making, in any material form, of an object visually similar to and derived from the original work. The protection thus accorded to original artistic works, which included mechanical drawings, was thought to be too wide and a limitation was enacted (see the Copyright Act 1956 (U.K.), s.9(8), and its analogue in s.71 of the Act). The extent of the limitation was considered by this Court in S.W. Hart &. Co.Pty.Ltd. v. Edwards Hot Water Systems [1985] HCA 59; (1985) 59 ALJR 729, at p 738; [1985] HCA 59; [1985] HCA 59; 61 ALR 251, at p 267, where I drew attention to the difficulty of applying the test of visual similarity to a written specification - a literary work. Although copyright in an original work may now be infringed by unlicensed reproduction whatever the material form of the infringing article may be, the form of the original work may limit the forms in which it is possible to reproduce it. A building may resemble a drawing (as in Chabot v. Davies), but it is difficult to conceive of a building resembling a piece of writing or a musical score.

18. In Cuisenaire v. Reed [1963] VicRp 96; (1963) VR 719, the plaintiff claimed infringement of copyright in a book which he had written describing a system of teaching children mathematics with the aid of rods of varying lengths and colours. The alleged infringement was the manufacturing of sets of rods in accordance with the plaintiff's description. The plaintiff failed, Pape J. saying (at p.734):

" When the chart or compilation referred to in
paragraph 1C of the statement of claim is looked
at, it is so entirely dissimilar to the
appearance of the defendants' rods that it is
impossible to conclude that they are a
reproduction of that chart or compilation. In my
view, a set of written directions is not
'reproduced' by the construction of an article
made in accordance with those directions. A
reproduction must reproduce the original, and
here the original is in one case a set of words
in the form of a table and in the other case a
series of plain and coloured circles which are
numbered and which are set out in the form of a
chart. In my view, the defendants' rods
reproduce neither."

That decision was followed in similar litigation in Canada: Cuisenaire v. South West Imports Limited (1967) 37 Fox Pat.C.93; (1969) 40 Fox Pat.C. 81. Similarly, in Brigid Foley Ltd. v. Ellott (1982) RPC 433, where the plaintiff alleged that her copyright in written instructions for knitting a garment had been infringed by knitting garments in accordance with the instructions, Megarry V.-C. held there was no reproduction. He said (at p.434):

" ... it seems to me quite plain that there is no
reproduction of the words and numerals in the
knitting guides in the knitted garments produced
by following the instructions. The essence, I
think, of a reproduction (and I do not attempt to
be exhaustive) is that the reproduction should be
some copy of or representation of the original.
I do not see how anyone looking at the knitted
garment could then say 'Well, that is a copy of,
or a reproduction of, the words and numerals to
be found in the knitting guide'. By a process of
counting up the number of stitches, and so on, in
the knitted garment one might be able to work
back and produce the knitting instructions; but
that is a very different matter from saying that
the garment is a reproduction of those
instructions."

19. Reproduction in any material form is a notion wider than the notion of copying which defined the scope of copyright protection before the 1911 Act, but the notion of reproduction still connotes a resemblance between the work in which copyright subsists and the work which is copied from it. The material form to which the respective works are reduced need not be the same, but if the forms are so dissimilar as to deny resemblance between the works, one cannot be said to be a reproduction of the other. As Gibbs C.J. pointed out in Hart v. Edwards Hot Water Systems (at p 730; p 253) -

" The notion of reproduction, for the purposes
of copyright law, involves two elements -
resemblance to, and actual use of, the copyright
work, or, to adopt the words which appear in the
judgment of Willmer L.J. in Francis Day & Hunter
Ltd v Bron (1963) Ch.587 at 614, 'a sufficient
degree of objective similarity between the two
works' and 'some causal connection between the
plaintiffs' and the defendants' work'."

20. The chief difficulty in the way of Apple's claim that the making of the Wombat ROMs and EPROM "reproduced" the Applesoft and Autostart source programmes lies in the want of resemblance between the written source programmes and the Wombat ROMs and EPROM. Perceiving this difficulty, counsel for Apple sought to identify the literary work in which Apple's copyright subsists not as the written compilation but as that collocation of ideas expressed in the written compilation. If the collocation of ideas be the work in which copyright subsists, so the argument runs, the expression of that collocation of ideas in any material form constitutes a reproduction of the original work: it is immaterial whether the form in which the collocation of ideas is originally expressed resembles the form in which they are expressed by the copyist. The argument runs counter to a formidable body of authority.

21. The distinction between writing and the ideas therein expressed was drawn by Erle J. in Jefferys v. Boosey [1854] EngR 816; (1854) 4 HLC 815, at p 869 [1854] EngR 816; (10 ER 681, at p 703), where his Lordship held that a claim of copyright "is not to ideas, but to the order of words". Lindley L.J. stated the principle in Hollinrake v. Truswell (at p 427):

" Copyright, however, does not extend to ideas,
or schemes, or systems, or methods; it is
confined to their expression; and if their
expression is not copied the copyright is not
infringed."

That principle was applied by Peterson J. in University of London Press (at p.608) and by Lord Hodson in Ladbroke (Football) Ltd. (at p.287; p.477). It was applied in Canada in Moreau v. Roland St. Vincent (1950) 10 Fox Pat.C.194, where Thorson P said (at p 200):

" Just as an author has no copyright in the ideas
he has expressed even although they are original,
but only in his expression of them, so also no
person has any copyright in an arrangement or
system or scheme or method for doing a particular
thing even if he devised it himself. It is only
in his description or expression of it that his
copyright subsists."

The same view was taken by Holmes J. in the Supreme Court of the United States in White-Smith Music Co. v. Apollo Co. [1908] USSC 66; (1908) 209 US 1, at p 19 (52 Law.Ed.655, at p 663):

" The ground of this extraordinary right is that
the person to whom it is given has invented some
new collocation of visible or audible points, -
of lines, colors, sounds, or words. The
restraint is directed against reproducing this
collocation ...".

If copyright subsisted in the collocation of ideas expressed rather than in the collocation of "lines, colors, sounds, or words" in and by which ideas are expressed, copyright protection would be tantamount to the protection given by the grant of a patent. In Baker v. Selden [1879] USSC 9; (1879) 101 US 99 (25 Law.Ed.841) Bradley J., speaking for the Court, said (at p 102; p 843):

" ... no one would contend that the copyright of
the treatise would give the exclusive right to
the art or manufacture described therein. The
copyright of the book, if not pirated from other
works, would be valid without regard to the
novelty, or want of novelty, of its subject-
matter. ... To give to the author of the book an
exclusive property in the art described therein,
when no examination of its novelty has ever been
officially made, would be a surprise and a fraud
upon the public. That is the province of
letters-patent, not of copyright."

The exclusive right to reproduce a work in which copyright subsists is not calculated to prohibit the use of ideas expressed in the work, though the form in which those ideas may be expressed without the licence of the owner of the copyright is limited. But copying in the sense of producing something which embodies the same ideas is not, without more, an infringement; the form of the thing produced must be such as to resemble the work in which the copyright subsists. Here there is no resemblance between the Wombat ROMs and EPROM and the written compilations of the Applesoft and Autostart source programmes. It is immaterial that the former were derived from the latter when there is no resemblance between them, for both resemblance and derivation are essential to reproduction.

22. As neither sub-par.(i) nor sub-par.(vii) of s.31(1)(a) applies in relation to the making of the Wombat ROMs and EPROM, there was no infringement of Apple's copyright in the Applesoft and Autostart source programmes. It is unnecessary to consider what was the knowledge of Computer Edge and Mr Suss as to the making of the Wombat ROMs and EPROM - a question raised by the cross-appeal. Knowledge is an element of liability under ss.37 and 38 but neither of those sections could impose liability on Computer Edge or Mr Suss. I would therefore allow the appeal, set aside the whole of the order of the Full Court, and restore the order of Beaumont J. The respondents should pay the appellants' costs on this appeal and before the Full Court.

DEANE J.: The factual and technical context within which this appeal falls to be determined and the relevant legislative provisions are set out in the joint judgment of Mason and Wilson JJ. and, in somewhat more detail, in the judgments in the Federal Court of Australia. Except to the extent necessary for the purposes of discussion, I shall seek to avoid repetition. There remains in dispute the issue of the appellants' alleged infringement of copyright under ss.37 and 38 of the Copyright Act 1968 (Cth) ("the Act"). That issue falls to be determined by reference to the Act in the form it took before the introduction of special provisions dealing with copyright in computer programmes (see Copyright Amendment Act 1984 (Cth)). It was decided in the appellants' favour by the learned trial judge (Beaumont J.) and, subject to the outstanding question of the requisite knowledge under those sections, in the respondents' favour by a majority of the Full Federal Court (Fox and Lockhart JJ.; Sheppard J. dissenting). It is apparent that a finding that the appellants were guilty of infringement under ss.37 and 38 can be sustained only if, as required by each of those sections, the actual making, as distinct from the use or functioning, of the Wombat microcomputer's ROMs and EPROM would have constituted an infringement of some copyright owned by the respondents, or either of them, in some literary work relating to one or both of the Applesoft and Autostart programmes.

2. There is an inevitable element of ambiguity in any reference to the Applesoft and Autostart "programmes" unless one takes care to distinguish between a written programme in source or object code and the actual series of electrical impulses or charges which constitutes an operational computer programme. The written programme in source code is directed to a human reader and not to a machine. Its essential function is to record and communicate programming instructions. Thus, the written expression of each of Applesoft and Autostart in source code consisted essentially of instructions, in 6502 Assembly Code mnemonics with accompanying labels and comments, to be read and followed by a human reader in his or her operation of an assembler to generate the series of electrical impulses or charges constituting the relevant operational programme. A written expression of Applesoft or Autostart in object code consists of the designation in a notation code (e.g. binary or hexadecimal notation) of the relevant sequence of electrical impulses or charges. The actual series of electrical impulses or charges constituting the operational object programme may be in a transient state or "stored" in any one of a number of ways including temporarily in the RAM of an activated computer and semi-permanently or permanently on a magnetic tape or disc or in a ROM. In the present case, the electrical impulses or charges generated by the operation of an assembler in accordance with the instructions of the written Applesoft and Autostart source programmes were, after some intervening steps, stored by mechanical process in silicon chips designed to serve the function of programmed ROMs in the Apple II microcomputer. Once the first set of charged or programmed ROMs had been produced, their functional qualities could be directly reproduced or copied by mechanical means. It is not, however, suggested that a written expression of a programme in object code actually played any part in the chain of events leading from the composition of the written Applesoft and Autostart programmes in source code to the manufacture of the ROMs and EPROM of a Wombat microcomputer.

3. It was submitted on behalf of the respondents that the actual sequences of electrical charges, which constitute the operational Applesoft and Autostart object programmes "stored" in the ROMs, were themselves original literary works in which the first respondent owned the copyright and that that copyright would have been infringed by the unauthorized making in Australia of the ROMs and EPROM of the Wombat microcomputer. That argument will be considered subsequently in this judgment. For the moment, it is convenient to examine what appears to me to be the more strongly arguable basis of the respondents' claim for relief, namely, that the making in Australia of the Wombat ROMs and EPROM by the importer would have constituted an infringement of copyright owned by the first respondent in the written expression of the Applesoft and Autostart programmes. Apart from ss.37 and 38, the legislative provisions upon which the respondents primarily relied for the purposes of this argument are to be found in s.36 and s.31(1)(a)(i),(vi) and (vii). For present purposes, the effect of those provisions is that copyright in a literary work is infringed by the unauthorized doing in Australia of any act comprised in the copyright and that the making of a reproduction in a material form or an adaptation of a literary work (or of a reproduction in a material form of an adaptation of the literary work) is an act comprised in the copyright in the work. That being so, the respondents' argument can prevail only if the making in Australia of the Wombat ROMs and EPROM would, for relevant purposes, have involved the making of a reproduction or adaptation (or a reproduction of an adaptation) of an actual literary work constituted by the written expression of the Applesoft and Autostart programme in source or in object code. The Act does not define what constitutes reproduction in a material form. That must be determined in the not always clear light of the authorities. Section 10(1) of the Act defines "adaptation" as including "a translation of the work". It is in the sense of making a translation that the respondents have relied upon alleged adaptation.

4. The programming of a ROM is done electrically. Its effect is essentially functional rather than visual and the programme state of a ROM is ascertained not by visual observation but by actual electrical operation. As I followed the argument in this Court, it appeared to be common ground that one does not - and, so far as the evidence goes, cannot - visually observe, let alone visually discern the pattern of, the overall distribution of electrical charges in a programmed ROM. The evidence does not disclose the precise technology used in charging the two ROMs and the EPROM in a programmed Wombat microcomputer ("the Wombat ROMs") or the extent of any correspondence between the spatial distribution of electrical charges in the Wombat ROMs and the distribution of charges in the six ROMs of a programmed Apple II microcomputer ("the Apple II ROMs"). It is, however, clear that, functionally, the three Wombat ROMs were a reproduction of the six Apple II ROMs. Indeed, the appellants do not dispute that the main step in the original production of a set of the Wombat ROMs probably involved reproduction, by mechanical copying, of the functional qualities (in the sense of electrical operation) of the Apple II ROMs. That being so, the question whether the electrical charging of silicon chips to produce the Wombat ROMs involved the making of a reproduction or adaptation of the written programmes in source or object code corresponds, for practical purposes, with the question whether the electrical charging of silicon chips to produce the Apple II ROMs involved the making of such a reproduction or adaptation. The answer to the question lies, in each case, in the nature of the written programmes in source and object codes and of the programmed ROMs and in the distinction between reproducing or adapting an actual literary work (or reproducing an actual adaptation of a literary work) on the one hand and utilizing or applying the information conveyed by the literary work or the adaptation to produce a functional object.

5. Where the written description of, or specification for, a functional article constitutes an original literary work, copyright prevents the unauthorized reproduction in a material form or adaptation of the description or specification. It does not however, of itself, provide the foundation of an exclusive right to make the designated article. Thus, the "description of the art in a book, though entitled to the benefit of copyright, lays no foundation for an exclusive claim to the art itself. The object of the one is explanation; the object of the other is use. The former may be secured by copyright. The latter can only be secured, if it can be secured at all, by letters patent" (Baker v. Selden [1879] USSC 9; [1879] USSC 9; (1879) 101 US 99, at p 105 (25 Law Ed 841, at p 844)). The point was tellingly made by Pape J. in the oft-cited passage from his judgment in Cuisenaire v. Reed [1963] VicRp 96; (1963) VR 719, at pp 735-736:

" This present case is of course so far as this
aspect is concerned not one relating to artistic
copyright, but to literary copyright, and I realize
that under the Act the copyright in a book may be
infringed by a performance in public of the
substance of that book if it be converted into a
dramatic work. But that is because the Act in
s.1(2)(c) specifically so provides. But still the
dramatic work must reproduce the literary work.
Where, as here, you have a literary copyright in
certain tables or compilations, there is in my view
no infringement of the copyright in those tables or
compilations unless that which is produced is
itself something in the nature of a table or
compilation which, whether it be in two dimensions
or three dimensions, and whatever its material
form, reproduces those tables. Were the law
otherwise, every person who carried out the
instructions in the handbook in which copyright was
held to subsist in Meccano Ltd. v. Anthony Hordern
and Sons Ltd. (1918), 18 SR (NSW) 606, and
constructed a model in accordance with those
instructions, would infringe the plaintiff's
literary copyright. Further, as Mr. Fullagar put
it, everybody who made a rabbit pie in accordance
with the recipe in Mrs. Beeton's Cookery Book would
infringe the literary copyright in that book."

(See also Cuisenaire v. South West Imports Ltd. (1968) 40 Fox Pat. C. 81, at p 85; Boosey v. Whight (1899) 1 Ch 836, at p 842 and, on appeal, (1900) 1 Ch 122, at p 123; White-Smith Music Publishing Company v. Apollo Company [1908] USSC 66; (1907) 209 US 1, at pp 12-14 (52 Law Ed 655, at pp 659-660); Moreau v. St. Vincent (1950) 10 Fox Pat. C. 194, at p 200; Purefoy Engineering Co. Ltd. v. Sykes Boxall & Co. Ltd. (1955) 72 RPC 89, at 99; Brigid Foley Ltd. v. Ellott (1982) RPC 433, at p 434).

6. The physical alteration within a silicon chip when it is charged to serve the function of a ROM in a microcomputer consists, in the case of a ROM with many addresses where simple fusing is likely to be inappropriate, of a redistribution of the electrons or electrical charges within the chip. The redistribution of electrons in the original Apple II ROMs was, albeit indirectly after initial "storage" of the sequence of charges on a magnetic disc or tape, the result of electrical charging in pursuance of the written directions constituting the source programmes. In that sense, an Apple II (and, more remotely, a Wombat) charged ROM was the outcome of action taken in compliance with information conveyed by the written expression of the relevant programme in source code. A charged Apple II or Wombat ROM did not however, of itself, reproduce or adapt (in the sense of making a translation of) that written expression in whole or in part. To the contrary, the relationship between the written expression of the programme in source code and the programmed Apple II or Wombat ROM was conceptually similar to the relationships - i.e. between the handbook instructions and a Meccano model and between the recipe book and a rabbit pie - to which Pape J. referred in Cuisenaire v. Reed in the above extract from his judgment: the one resulted from the following of the directions of the other; it was not, however, a reproduction or adaptation of the literary work embodying those directions. The point can be illustrated by reference to the situation which would have arisen if the programming instructions constituting the two written programmes in source code had been published in a book comprising a collection of such source programmes. To reproduce in a material form or adapt the actual written instructions or directions constituting a source programme would, plainly enough, involve infringement of the copyright in the written programme. To follow those instructions or directions to produce a programmed disc, tape, ROM or RAM would however, equally plainly, consist of carrying out the instructions or directions to produce the designated article as distinct from reproducing or adapting the actual written programme in which copyright subsisted.

7. Nor was a programmed Apple II or Wombat ROM a reproduction or adaptation (in the sense of a translation) of the written expression of the relevant programme expressed in object code. The written object code denotes the series of electrical impulses produced by following the directions of the source code. The charged silicon chip reflects the application of that series of electrical impulses. The written denotation of the electrical impulses in object code - as distinct from the electrical impulses themselves - played no part whatever in the production of the Apple II or Wombat ROMs however and it would be a distortion to regard the latter as representing a reproduction or adaptation of the former. It is not to the point that an Apple II or Wombat ROM is capable of functioning as a switching device in an appropriately instructed operating microcomputer to reproduce the series of electrical impulses and to display or print out their denotation in object code. Indeed, it would be an inversion of logic to regard that consideration as supporting the conclusion that the silicon chip was an adaptation or reproduction of the written denotation of the programme in object code.

8. In the light of the foregoing, I find it unnecessary to determine whether a written expression of the Applesoft and Autostart programmes in source or in object code constituted or would have constituted literary works in which the first respondent owned the copyright or whether a written programme in object code constituted an adaptation of a written programme in source code. Even if all those matters be assumed in favour of the respondents, the making in Australia of the ROMs by the importer would not have constituted an infringement of copyright in any of those literary works in that it would not have involved the making of a reproduction or adaptation (or reproduction of an adaptation) of any of them.

9. There remains to be considered the separate argument, to which reference has already been made, based on the proposition that the actual series of electrical charges constituting each operational programme was itself an original literary work when "embodied" in a ROM or ROMs. I am unable to accept that proposition. The re-arrangement of electrons in a programmed ROM is not visible to the human eye. The programmed ROM may be used as a switching device in a functioning computer to produce a print out or visual display of something which could, for relevant purposes, properly be called a "literary work". Of itself however, and regardless of how widely one construes the phrase, the arrangement (or series) of electrons or electrical charges in the silicon chip does not constitute a "literary work". It is not written. It is not in a comprehensible language. It cannot be read. It cannot even be seen. Nor is it designed or produced to be read or seen. It is, and was designed and produced to be, an attribute of a functioning part of an operating machine.

10. It follows that the making of the Wombat ROMs in Australia would not, of itself, have constituted an infringement of the respondents' copyright in a literary work. It may well be that the operation of the programmed Wombat computer to produce a visual display or print out of the programme burnt into the Wombat ROMs would constitute such an infringement of some copyright owned by the first respondent in a literary work constituted by the written expression of the Applesoft or Autostart programme in source or object code. As has been seen however, that would not suffice to establish infringement pursuant to ss.37 and 38 of the Act since those sections require that the actual making, as distinct from the operation or functioning, in Australia of the imported or traded article would have constituted an infringement of copyright.

11. It may well be thought that it would be unfortunate if the law were to provide no protection for the respondents against conduct of the kind engaged in by the manufacturers of the Wombat microcomputers. Obviously, however, general questions of the existence and extent of copyright protection in a new field cannot be determined by reference to notions of what may or may not be fair in the circumstances of a particular case. In particular, the question whether the protection of literary copyright should be extended to provide protection in a field to which the terms of existing legislative provisions are inappropriate and inapplicable involves the assessment and balancing of a wide range of policy matters and considerations of a type with which Australian courts, following the adversary system, are ordinarily ill-equipped to deal (cf. Lahore, Dworkin and Smyth, Information Technology: The Challenge to Copyright (1984) p.110; Borking, Third Party Protection of Software and Firmware (1985) p.253; J.O. Webster, "Copyright Protection of Systems Control Software Stored in Read Only Memory Chips: Into the World of Gulliver's Travels", Buffalo Law Review, vol.33 (1984), 193, at pp.223-224). Those matters and considerations include the determination of the appropriate boundary between patent law, with its requirement of novelty and its fixed and more limited period of protection, and copyright law, with its limited requirement of originality and its much greater period of protection. They also include the resolution of the competing legitimate claims of different interests - the claim of an author, a designer or an inventor to the protection of his work and the claims of the public to the unrestricted use of information, technology and other material. The assessment and balancing of those matters and considerations and the question whether some, and if so what, extension of copyright protection to a new field to which existing legislative provisions are inapplicable is warranted are matters for the Parliament and not for the courts.

12. I would allow the appeal and restore the judgment of the learned trial judge. The question involved in the cross-appeal does not arise and the cross-appeal should be dismissed.

ORDER

Appeal allowed with costs.

Cross-appeal dismissed.

Set aside the judgment and order of the Full Court of the Federal Court of Australia dated the twenty-ninth day of May 1984 (except for paragraph 5 of that order) and in lieu thereof order that the appeal to that Court be dismissed with costs and that the judgment and order of Beaumont J. be restored.