Dyason v Autodesk Inc

Re: MARTIN PATRICK DYASON; CHRISTINE DYASON and PETER VINCENT KELLY
And: AUTODESK INC. and AUTODESK AUSTRALIA PTY. LTD
No. V G300 of 1989
FED No. 499
Intellectual Property
18 IPR 109/96 ALR 57

COURT

IN THE FEDERAL COURT OF AUSTRALIA

VICTORIA DISTRICT REGISTRY
GENERAL DIVISION
Lockhart(1), Sheppard(2) and Beaumont(3) JJ.

CATCHWORDS

Intellectual Property - copyright - meaning of "computer program" in Copyright Act 1968 - infringement action - respondents' device designed to prevent its programs being run on a computer without the device - whether this device is also a computer program, or part of a computer program - appellants sold to licensed users of the respondents' programs a device performing the same function - whether the appellants' device is a reproduction or adaptation in material form of the respondents' device - meaning of "version" in the Act - whether appellants authorised a breach of respondents' copyright by selling their device.

Copyright Act 1968 (Cth), ss.10, 13, 14, 31(1)(a)(i), (vi) and (vii), 36(1).

HEARING

SYDNEY

#DATE 14:9:1990

Counsel for the appellants: Mr. J.W.K. Burnside, QC

and Mr. I.G. Waller

Solicitors for the appellants: Nevett Ford of Melbourne

Counsel for the respondents: Dr. J. McL. Emmerson QC and

Mr. A. Panna

Solicitors for the respondents: Stephens of South Melbourne

ORDER

Appeal allowed.

Cross-appeal dismissed.

Set aside the injunction (Order 1) granted on 15 September 1989; in lieu thereof, order that the claims made in paras. 3, 4 and 5 of the statement of claim be dismissed; and that the whole application be dismissed, with costs.

Order that respondents pay the costs of the appellants of the appeal and of the cross-appeal.

Note: Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.

JUDGE1

The protection of the form of expression, not the ideas embodied in it, is fundamental to the law of copyright. But, as Lord Hailsham said in B. Plastics Limited v Swish Products Limited (1979) RPC 551 at 629:

"As the late Professor Joad used to observe, it all depends on what you mean by 'ideas'."

So it is with function which is not itself protected by copyright, and it is the concept of function that this case is all about. Modern technology and its challenges must make us pause to think about underlying assumptions. Take, for example, the principle that reproduction in copyright law necessarily requires there should be "a sufficient degree of objective similarity between the two works": see Francis Day and Hunter Limited v Bron (1963) Ch 587 at 614; Lend Lease Homes Pty. Limited v Warrigal Homes Pty. Limited (1970) 3 NSWR 265 at 273; S.W. Hart and Co. Pty. Limited v Edwards Hot Water Systems (1985) 159 CLR 466 at 472; and Computer Edge Pty. Limited v Apple Computer Inc. (1986) 161 CLR 171 at 186 and 207.
With literary, dramatic, musical or artistic works as generally understood traditional tests such as the one just mentioned are still relevant and correct, but what about computer programs? Can their protection really be left, as some writers would have it, to the law of patents or breach of confidence? Many countries do not think so and have legislated to bring computers under the umbrella of copyright, not just by squeezing them into traditional frames, but by specifically legislating for them. The amendments to the Copyright Act 1968 ("the Act") introduced by the Copyright Amendment Act 1984 ("the Amendment Act") represent Australia's endeavours so far to achieve this objective together with the reference to the Copyright Law Review Committee in October 1988.
The Act, so far as it relates to computer programs is a patchwork dealing basically with copyright in its traditional setting and grafting onto it, albeit as an interim measure, computer programs to which traditional notions of reproduction and adaptation are applied; though the notion of adaptation suggests an inherent capacity for change, possibly greater than reproduction. Computer technology changes overnight. The law does not but the application of its principles is not intractable.
Computers present many problems to copyright law. A computer program is a functional entity the sole purpose of which is to make a piece of electronic equipment work in a particular way. Computer programs are primarily intended for use in the operation of computers, but the use of computer programs will not necessarily involve a reproduction of the programs. The new definition of "literary work" introduced by the Amendment Act to include a "computer program" is designed to ensure, not only that translations of human languages constitute adaptations, but that translations of computer languages into different levels also constitute adaptations.
But even following the 1984 amendments, what constitutes a literary work being a computer program protected by the Act? On one extreme it may be said to be the marks on the paper that constitute the source code. Another view is that the marks on the paper or some other material record simply evidence the true work which is the product of the mind of its author. For there to be an infringement of copyright the act of storing a copyright work inside a computer system must be more than the representation of the material form. There must be also a reproduction of the work in a material form: see ss. 31(1)(a)(i) and 36 and the definition of "material form" in s. 10. Notwithstanding the new definition of "computer program" as a literary work (s. 10) there is a question as to whether electronic circuitry answers the description of "the conversion to another language, code or notation" within the definition of "computer program" in s. 10.
The amending Act did not deal with the vital question of reproduction of a work and leaves still for resolution the fundamental question of whether electro-mechanical and other functional devices can constitute reproduction of literary works for purposes of infringement. The principles expounded in copyright cases in the nineteenth century and the present century demonstrate the concern of the Courts to ensure that in holding one work to be a reproduction of an original work the monopoly is not conferred on the owner of the original work to prevent everyone else from using a process or system which may be seen as representing an embodiment of the work. The decision of the High Court in the Apple Computer Case, especially the judgment of Gibbs C.J., draws a sharp distinction between making a copy of the writing being the source code and making a copy of the electronic circuitry. The majority of the High Court in the Apple Computer Case established a test of visual similarity between the original work and the offending work for the purposes of reproduction under the Act. Does it follow that an embodiment of object code which is not visible on a disk, tape or silicon chip does not reproduce the computer program? These and other considerations give rise to the question whether there should be a new form of protection for computer programs which either abandon altogether the concept of copyright or, whilst retaining copyright in computer programs, separate them from the notion of literary works and establish a self contained statutory regime within the Act.
This case presents a difficult question. At its heart is the concept and role of function in copyright law with respect to computers and computer programs.
The respondent, Autodesk Inc., an American company, developed AutoCAD which is a computer program for the drafting of architectural and engineering plans and designs. AutoCAD is sold in Australia as a package which consists of a number of disks and a lock or key known as a hardware lock and to which I shall refer as "the AutoCAD lock". The program is expensive, its retail price in Australia being more than $5,000.
The AutoCAD lock is essential to the operation of the AutoCAD program. Only one AutoCAD lock is sold with each AutoCAD program to ensure that only one program can be used at any one time. Unless the AutoCAD lock is in place the program cannot function. The AutoCAD lock is attached to the serial port of the computer. The AutoCAD disks can be copied by methods of disk duplication and the respondents encourage buyers to do this, but it is not possible to operate more than one at a time because of the necessity to use the AutoCAD lock. The AutoCAD lock is the method used by Autodesk to prevent piracy of the AutoCAD program and this is the sole function of the AutoCAD lock. It performs the function by generating a sequence of binary digits in response to incoming digital information.
The appellant, Mr. Kelly, is an electronic designer. During 1988 he approached one of AutoCAD's dealers and agreed to investigate the feasibility of selling AutoCAD in certain parts of Melbourne. Mr. Kelly worked with the personnel of the AutoCAD dealer for four months during which he learnt the technical and commercial details of AutoCAD including details of the operation of the AutoCAD lock. Mr. Kelly was given an AutoCAD program and AutoCAD lock to use on his own personal computer. He was also given the technical drawings which related to the operation of the AutoCAD program and lock. Mr. Kelly understood the function of the AutoCAD lock and the fact that the AutoCAD program could not function without the lock in place. He did not dismantle the AutoCAD lock to examine its parts; but by use of an oscilloscope he observed the workings of the AutoCAD lock and by looking at electronic signals passing from the computer to the AutoCAD lock and from that lock to the computer whilst AutoCAD was being run he compared the stimuli going to the lock and the responses from the lock back to the computer. He discerned a pattern based on the "negative edge", being the occasion on which the electronic charge went off the line. He prepared his own program to produce a standard stimulus to the lock based on the AutoCAD program so that he could look at the response from the lock simplified by the application of a standard stimulation.
In the result he produced a prototype of the device which could be used with a computer which enabled the AutoCAD program to be run without the AutoCAD lock in place. Mr. Kelly thus "cracked the code" used between the AutoCAD lock and the computer. He was able to duplicate the function of the lock and to produce an alternative device enabling the AutoCAD program to be run without the lock.
Mr. Kelly sought advice from a solicitor and a patent attorney and was advised by the solicitor that he could market his alternative device without infringing the law. Mr. Kelly decided to do this under the name "Auto-Key". He obtained the assistance of the other appellants to market Auto-Key. The Auto-Key differs from the AutoCAD lock in that it is larger and is connected to the computer somewhat differently. It was, however, designed and is produced and marketed for one purpose only, namely, to allow the operation of the AutoCAD program without the necessity of using the AutoCAD lock. The saving of cost to users of AutoCAD was considerable and the demand for Auto-Key became heavy.
The respondents commenced these proceedings alleging that the appellants had infringed their copyright in the computer program contained within the AutoCAD lock and that in marketing Auto-Key the appellants had "authorised" the respondents' customers to infringe the respondents' copyright in the AutoCAD lock pursuant to s. 36 of the Act.
Northrop J. who heard the case at first instance held that each of the AutoCAD lock and the Auto-Key lock is a computer program as defined in the Act and thus a "literary work" within the meaning of that phrase in the Act. He held that the appellants had infringed the copyright of the respondents in the AutoCAD lock by reproducing the AutoCAD lock in a material form and granted injunctions to restrain the appellants from reproducing in any material form the computer program contained in the AutoCAD lock or any adaptation thereof. Having found that the conduct of the appellants constituted a reproduction of the computer program contained within the AutoCAD lock he did not find it necessary to deal with the question of adaptation. The appellants appealed to a Full Court of this Court from Northrop J.'s judgment.
I have stated the basic facts in summary form. The facts are fully stated by Sheppard J. in his reasons for judgment of which I have had the advantage of reading. In particular, Sheppard J. has stated the way in which the AutoCAD and Auto-Key locks work including the AutoCAD program and the set of instructions written into it known as WIDGET.C.
I agree with Sheppard J. that the WIDGET.C program and the AutoCAD lock constitute an integrated system and that the two in combination answer the description of a "computer program" within the meaning of that expression as defined in s. 10 of the Act. I agree also with the reasons which his Honour gives for reaching that conclusion. The AutoCAD lock itself, considered in isolation from WIDGET.C, is not a computer program; but the two in combination do constitute the relevant computer program in this case.
The relevant provisions of the Act as they stand after the amendments made by the Amendment Act in 1984 are as follows:-
. Copyright in relation to a literary work is the exclusive right to reproduce the work in a material form (s. 31(1)(a)(i)) and to make an adaptation of the work (s. 31(a)(vi)).
. "Literary work" is defined in s. 10, the definition section, to include a computer program. "Computer program " is defined as follows:-

"'Computer Program' means an expression, in any language, code or notation, of a set of instructions (whether with or without related information) intended, either directly or after either or both of the following:

(a) conversion to another language, code or notation;

(b) reproduction in a different material form; to cause a device having digital information processing capabilities to perform a particular function;"

. The Act does not contain a definition of "reproduction", but the Amendment Act inserted a definition of "material form", namely:

"'material form', in relation to a work or an adaptation of a work, includes any form (whether visible or not) of storage from which the work or adaptation, or a substantial part of the work or adaptation, can be reproduced."

. The word "adaptation" is defined in s. 10 as meaning:

"(a) in relation to a literary work in a

non-dramatic form - 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;

(ba) in relation to a literary work being a

computer program - a version of the work (whether or not in the language, code or notation in which the work was originally expressed) not being a reproduction of the work;

(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;"

. Section 13 provides that a reference in the Act to an act comprised in the copyright in a work shall be read as a reference to any act that under the Act the owner of the copyright has the exclusive right to do.

. Section 14 provides that a reference to the doing of an act in relation to a work or other subject matter is to be read as including a reference to the doing of that act in relation to a substantial part of the work or other subject matter; and a reference to a reproduction, adaptation or copy of a work shall be read as including a reference to a reproduction, adaptation or copy of a substantial part thereof.

The first question is whether the Auto-Key lock is a reproduction of the AutoCAD program. Mr. Kelly did not copy the AutoCAD program. I have already traced the steps which he followed enabling him to produce the Auto-Key lock. Northrop J. found that the Auto-Key lock was a reproduction of the program in the AutoCAD lock. His Honour referred to the judgment of Gibbs C.J. in the Apple Computer Case where the Chief Justice said at 186 that the notion of reproduction involved two elements: first that the infringing work sufficiently resembled the copyright work and second, that the infringing work was produced by the use of the copyright work. Northrop J. found that the latter element was established.

"Mr Kelly had access to the AutoCAD lock. He made a close study of it. He made an analysis of its response to the challenge from AutoCAD. Based on the results of his studies he prepared a computer program to determine the information to be placed in the EPROM which forms the basic component of the Auto-Key lock. Based on the results of his study he produced a computer program which was designed to perform and in fact performed the same function as the AutoCAD lock."

Northrop J. then turned to the question of whether the first of the two elements mentioned by Gibbs C.J. was satisfied. His Honour said:

"There is a difference in form between the hardware within the AutoCAD lock and the Auto-Key lock, namely the shift-register and the EPROM. But the Court has concluded that each lock constitutes a computer program. Each performs the same function. It is this function which is the essential aspect of each lock. Function has a particular importance in the definition of a computer program and regard must be given to this concept of function in considering the question of whether there is a 'sufficient degree of objective similarity' between the two locks. Having regard to the definition of computer program, it seems that the particular method by which 'a device having digital information processing capabilities to perform a particular function' is not really relevant in considering whether the Auto-Key lock 'sufficiently resembles' the AutoCAD lock. In any event, on the findings made, in my opinion the processes adopted in the Auto-key lock to produce the same function as the processes adopted in the AutoCAD lock do sufficiently resemble the literary work of the AutoCAD lock as to come within the first element of the test propounded by Gibbs C.J.

. . .

In the present case, the form of the reproduction is, in my opinion, irrelevant. There is no doubt the whole of the function of the AutoCAD lock has been reproduced in a material form by the Auto-Key lock. Thus it is not necessary to consider whether a substantive part of the work has been produced. Many of the submissions put to the Court on this issue by counsel for the respondents were based upon the source code Widget C being part of AutoCAD and another source code being part of AutoCAD and described as SPTEST.C. These would be accepted as being works within the Copyright Act but the submissions proceeded on the basis that the AutoCAD lock was not a work. The argument then was that the AutoCAD lock was not a reproduction in material form of either Widget C or SPTEST.C. It was argued that there was no resemblance between any part of those programs and the Auto-Key lock. Reliance was placed upon the fact that different methods were employed between those programs and the method used in the Auto-Key lock. There may be similarity in logic, but not appearance, and therefore there could be no resemblance. further, it was put, and is accepted, that Mr. Kelly had no access to either Widget C or SPTSET.C. . . .

On the facts of this case a finding has been made that the AutoCAD lock and the Auto-Key lock each constitutes a computer program under the Copyright Act and thus each is itself a literary work under that Act. In this context, regard must be had to the function of the computer program in determining resemblance. Physical appearance is immaterial. The hardware or physical equipment within which the expressions of the sets of instructions are contained, is immaterial. The fact that Mr. Kelly did not see Widget C, nor for that matter, the internal mechanisms of the AutoCAD lock, are immaterial. Mr. Kelly knew of the function of the AutoCAD lock, he discovered how it performed that function, and prepared the expression for his own set of instructions for the Auto-Key Lock to enable it to perform the identical function. This constitutes the resemblance between the two computer programs."

His Honour concluded that the Auto-Key was a reproduction of the AutoCAD program.
There is I think a fundamental difficulty in a finding of reproduction on the facts of this case. The element which was held in the Apple Computer Case to be an essential ingredient in the notion of reproduction is that there should be "a sufficient degree of objective similarity between the two works" (Francis Day and Hunter Limited v Bron) or, to use the language of Gibbs C.J.: "The infringing work sufficiently resembles the copyright work." As Brennan J. put it at 207:

"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."

With respect to the learned primary Judge, once it is recognised, as I think it must from the reasons of the majority in the Apple Computer Case, that there must be a sufficient degree of objective similarity between the copyright work and the infringing work, which in this case involves a comparison between the AutoCAD program and the Auto-Key program, the conclusion that the latter is a reproduction of the former cannot be sustained. The composition of WIDGET.C and the program in the Auto-Key are different. The algorisms (that is the procedures for solving the particular problem in a finite number of steps) employed by each are different and the implementation of each is different.
I agree with Northrop J. that function performs a critical role in the definition of computer programs since the 1984 amendments. Nevertheless, in the light of the facts as I have outlined them earlier it cannot be concluded that there is a sufficient degree of similarity between the copyright work and the impugned work. If this reveals a lacuna in the Act following the 1984 amendments it can only be cured by Parliament.
I turn to the question of whether the program in the Auto-Key lock is an adaptation of the AutoCAD program. The starting point is the definition of "computer program" which was introduced by the amending Act in 1984, the terms of which are set out earlier. That definition demonstrates that it is the expression of the requisite set of instructions that constitutes the program. This establishes, consistently with accepted thinking in copyright law, that the Act intends to confer copyright protection, not on abstract ideas or mathematical principles, but on the particular expression of the ideas or principles in the form of sets of instructions. The words "in any language, code or notation" are intended to include all levels, "high level" and "low level", of computer language programs.
The word "version" is not defined in the Act. It appears in paragraphs (a), (b), (ba) and (c) of the definition of "adaptation". In paragraph (a) it means, in relation to a literary work in a non-dramatic form, a version of the work in a dramatic form, a change which may be considerable. Also in paragraph (b) the adaptation means, in relation to a literary work in a dramatic form, a version of the work in a non-dramatic form. In paragraph (c)(ii) adaptation means, in relation to a literary work whether in a non-dramatic form or in a dramatic form, a version of the work in which a story or action is conveyed solely or principally by means of pictures. The word "version" in that context is necessarily capable of application to cases where the form of the work is itself greatly changed. The literary work is a story which is then conveyed principally in pictures and which becomes an adaptation of the work. This illustrates the wide connotation to be attributed to the word "version" in the definition of "adaptation" in s. 10. Paragraph (ba) was introduced by the amending Act and is designed to cover many possibilities. A clear case is where the adaptation is from one code to another; for example, from source code to object code.
The word "version", where appearing in the definition of "adaptation", bears in my view its ordinary and natural meaning as a special form or variant of something.
The amendments introduced in 1984 were intended to be remedial and capable of application to a wide variety of circumstances to encompass the many advances which occur in the field of computer technology.
Copyright in computer programs resides in the expression of a set of instructions which cause a specified device to perform a particular function; but copyright does not reside in the function of the device itself. Function is nevertheless critical because, when considering whether something answers the description of a program within the definition in s. 10, it is only when the expression of the set of instructions is intended to cause the specified device to perform a particular function that a "computer program" exists for the purposes of Australian copyright law.
It is true that the law of copyright does not protect function; it protects the form of expression of ideas. The essential difference between computer programs and other literary works is that the former take the form which they do to produce a particular result or function. The programs are not prepared primarily for the purpose of reading or seeing them, but for the purpose of achieving a particular result in a device, in this case the AutoCAD lock and the Auto-Key lock. Hence the reference to function as an essential part of the definition of "computer program" in s. 10.
A computer program is a purely functional entity, the sole purpose of which is to make a piece of electronic equipment work in a particular way. This is demonstrated by the definition of "computer program" following the amending Act. Literary, artistic, musical and dramatic works are not useful in the functional sense of a computer program. It is not in my view necessary for there to be an adaptation of a copyright work that the impugned version of the work embodies or preserves the original intrinsic qualities of the copyright work. There are indicia within the definition of "adaptation" itself which demonstrate that a work may be an adaptation of another work notwithstanding considerable differences between the two works. Indeed, the word "version", meaning a variant of something, necessarily requires that there be change or variation. Though it goes too far to say that something is a version of another thing if the two are essentially different.
Mr. Kelly not only had access to the AutoCAD lock, but he derived the program for Auto-Key from the AutoCAD lock. He chose Auto-Key because it acts on transitions, gives the precise sequence of responses as is given by the AutoCAD lock and has what is described in the evidence as the same "wrap-around" facility which takes effect at the same stage of the cycle. Mr. Kelly's program for Auto-Key was reduced to written form and was an exhibit at the trial.
What Mr. Kelly did was to observe with the aid of an oscilloscope the passage of voltages or transitions between the AutoCAD program and the AutoCAD lock. He observed the regularity of the pattern of the voltages or transitions, in particular the sequence and the eventual wrap-around after the 127 bit sequence had concluded. This enabled him to devise the Auto-Key lock. Mr. Kelly did not, however, copy any part of the WIDGET.C program itself which formed the major component of the combined program and which enables the operator of the AutoCAD program to continue to run it.
Mr. Kelly observed a repeating pattern in the input and output of the negative transitions of the AutoCAD lock operations. He then prepared a source program consisting of a series of numbers based upon the repeating pattern of the AutoCAD lock operations. This was a tabular representation of pattern and was the program to be used for the preparation of a "look-up" table to be coded into an EPROM. The program for the "look-up" table was prepared from the tabular representations. The EPROM in the Auto-Key contains a set of digits in a sequence which is identical with that contained in the original state machine table contained in the AutoCAD program and used in the AutoCAD lock shift register. When the input from the AutoCAD program comes to the Auto-Key lock, the transitions on the input are counted by a simple binary counter. Whilst the AutoCAD lock counts these in a pseudo-random sequence generator, the Auto-Key lock counts them in a plain and simple binary counter. The AutoCAD lock, unlike the Auto-Key lock, does not contain an EPROM. The shift register in the AutoCAD lock is hard wired. Each is a piece of hardware performing a similar function in moving digits.
The AutoCAD lock is produced simply to limit the use of the AutoCAD program to ensure that it is run on the terms on which Autodesk sells it and is intended to inhibit the making of unauthorised copies of the program. The Auto-Key lock was devised to circumvent the precautions taken by the respondents to prevent unauthorised use of the AutoCAD program.
The composition of WIDGET.C and the program in the Auto-Key differ, but they produce the same result and the achievement of that result is the essential purpose of the programs within the respective locks. The forms of expression of the two programs with which this case is concerned differ but each is designed to and in fact achieves the same function. The two programs have different algorithms and the implementation of each of those algorithms is different.
I have so far put the case for the respondents at its highest. I then pause and ask whether a case of adaptation has been established. It has not. Ultimately the only element that WIDGET.C and the AutoCAD lock have in common with Auto-Key is function. There is not a shred of similarity between the two programs or sets of instructions.
In my opinion the program contained in the Auto-Key lock is not a version of what is contained in the combination of the WIDGET.C program and the AutoCAD lock. I would therefore allow the appeal.
I agree with the reasons for judgment given by Sheppard J. and Beaumont J. on the cross appeal subject to one matter, namely, that I prefer to leave open the question whether the operation of the AutoCAD program on a computer results in there coming into existence a reproduction or adaptation of the program.
I would allow the appeal and dismiss the cross-appeal, set aside the injunction granted by the primary Judge on 15 September 1989 and order that the application be dismissed with costs. The respondents must pay the costs of the appellants of the appeal and cross-appeal.

JUDGE2

This is an appeal from a Judge of this Court (Northrop J.) in which his Honour granted injunctive relief which had the effect of restraining the appellants from reproducing in any material form the whole, or a substantial part of, or any part of, a computer program contained within a device described as "the autoCAD lock" or any adaptation thereof or selling or letting for hire, or by way of trade, offering or exposing for sale or hire, any article being a reproduction of the whole, or a substantial part, of the work, that is the computer program within the AutoCAD lock. The order also restrained the appellants from marketing a device known as "Autokey" which the respondents claimed contained a computer program which infringed their copyright in the program contained within the AutoCAD lock. His Honour did not deal with the question of damages but stood the matter over to a date to be fixed for directions for the further conduct of it. His Honour dismissed a further claim brought by the respondents for injunctive relief based on a submission that the appellants, by marketing their computer lock, had, without the licence of the owner of the copyright, authorised the doing in Australia of an act comprised in the copyright, namely, the running of the AutoCAD program without the necessity to use the AutoCAD lock. His Honour gave each party leave to appeal and reserved costs. The respondents have cross appealed against the dismissal of their claim based upon the appellants' alleged unlicensed authorisation of the doing of an act comprised in the respondents' copyright. Counsel for the respondents also sought some variation of the terms of his Honour's order.

Outline of the Facts and Issues

The account of the facts which I am about to give is taken substantially from his Honour's judgment. The respondent Autodesk Inc. is a company incorporated under the laws of the State of California in the United States of America. The respondent Autodesk Australia Pty. Limited is a company incorporated in Victoria and is a wholly owned subsidiary of Autodesk Inc. Autodesk Inc. is the owner of the copyright subsisting in the United States of America and other parts of the world, including Australia, in a computer program known as "AutoCAD". Autodesk Australia Pty. Limited is, throughout Australia, the sole and exclusive licensee of Autodesk Inc. with respect to its copyright in AutoCAD. I shall hereafter refer to the respondents as "Autodesk".
AutoCAD is a sophisticated computer program designed to assist in the process of drafting. It is apparently used in a variety of drafting offices including the offices of professional architects and engineers. The letters CAD in the name AutoCAD represent "Computer Assisted Drafting". In Australia AutoCAD is sold by the Australian company to dealers who then sell it to persons for use in conjunction with their own computers and related equipment. Thus sales are made by dealers and not by Autodesk.
When a user purchases AutoCAD he or she acquires a package which comprises a number of disks. The computer program is contained in the disks. A person having those disks is able to make copies of them and, as a result, the program contained in the disks can be transferred from one disk to another thereby enabling the program to be used by a number of persons at any one time. These persons may, but need not, be employees or associates of the original purchaser of the program. A user may have a drafting office in which a large number of members wish to make use of the program on different computers at the same time. This could be achieved by the user making a number of copies of the program, a process which is cheap, but which may, subject to what is said later about licensing and authorisation, be illegal. The user would not need to purchase more than one AutoCAD program to enable this to be done. But although illegality may be involved, the problem is detection and enforcement. For this reason, a supplier of a computer program may attempt to limit the use of the program to any one computer at any one time irrespective of whether it be at work stations at the user's office or at other places.
It is Autodesk's attempt to enforce such a limitation which is at the heart of this case. In order to enforce it, it developed and produced a hardware lock which is known as the AutoCAD hardware lock. It is designed to prevent the AutoCAD program being used unless the lock is in position. By limiting the supply of the lock, Autodesk sought to limit the use of AutoCAD. Even if the program were copied on to other disks, those disks could not be used and the program could not be run without utilising the lock. Only one lock was supplied with each AutoCAD package. When a user purchased AutoCAD, the package supplied included a lock. Each lock had a serial number. Unlike the AutoCAD program itself, it was not easily copied.
In appearance the lock is black and rectangular in shape. It is about 75 mm. long, 43 mm. wide and 15 mm. deep. On one end is fitted a plug containing 25 holes. The plug is fitted into the outward or serial port of the computer. On the other end of the lock is fitted a socket containing 25 pins on to which can be plugged a lead connecting the computer to peripheral equipment such as its screen or printer. This socket contains 25 holes to take 25 pins but an adaptor is supplied in the package which reduces the outlet from 25 to 9. The lock has signs on it which include the word "AutoCAD", the logo of Autodesk and the serial number of the lock.
The AutoCAD lock was designed and produced for one purpose only, namely, to limit the use of the AutoCAD program. The program could be run if, and only if, the lock were in position and operating effectively. Although copies of the program could be made, the copies could not be run unless the computer being used was fitted with a lock.
I should pause at this stage to say that what is called throughout the evidence in the proceedings a lock is not really a lock at all. It is a key. The computer is locked unless the AutoCAD lock is used. If it is, the program is unlocked. Another point that should be made clear is that what really is locked is the screen and any printer attached to the computer. Unless the lock is attached, no display will appear on the screen and no printout of any drawings or calculations can be made.
The appellant, Mr. Kelly, is an electronic designer. He was familiar with the AutoCAD program and its lock. Using an oscilloscope he looked at the electronic signals passing from the computer to the lock and from the lock to the computer while AutoCAD was being run. When in operation, electrical charges of low voltage (5 volts) are sent along a line connecting the computer to the lock and from the lock to the computer. The oscilloscope showed when the charge was present. The oscilloscope had a dual channel and Mr. Kelly was able to compare the stimuli going to the lock and the responses from the lock back to the computer. Mr. Kelly carried out his examination over a period of some six to eight weeks. At that stage, his investigation was done out of curiosity. He discovered that there was output from the lock only when there occurred what he described as "a coincident negative edge on the output from the computer to the lock". He said that there appeared to be a direct "tie up" between a negative transmission from the computer and a charge in response from the lock. He noticed that there was no response from the lock except when there was a negative transmission. From this he deduced that the process was a repetitive one.
Using other electronic equipment, Mr. Kelly was able to ascertain a pattern based on "the negative edge", being the occasion on which the electronic charge went off the line. He prepared his own program to produce a standard stimulus to the lock based on the AutoCAD program to enable him to look at the response from the lock simplified by the application of a standard stimulation. As a consequence, Mr. Kelly produced a prototype of a device which could be used with a computer to enable the AutoCAD program to be run without the AutoCAD lock being used. Mr. Kelly thought that there might be a market available for the sale of his device to the users of AutoCAD. He sought legal advice about the matter and was advised that the sale of such a device would not be unlawful. He decided to manufacture the device and to test the market. To this end he sought financial and marketing advice from the other two appellants, Mr. and Mrs. Dyason. The three decided that the marketing would be done in such a way as to conceal the identity of the maker and seller of the device until the success of the project was established. They feared the reaction of Autodesk should it discover what was happening. Mr. Kelly made a number of his devices which he called the "Autokey hardware lock".

The Autokey lock is larger than the AutoCAD lock. It has a shiny metal-like plug containing 25 points to be inserted into the outward port of the computer fitted to one end and a socket containing provision for 25 pins fitted at the other. The only relevant word on the device is "Autokey".
Late in 1988 the appellants implemented their plan to manufacture and sell the Autokey lock. They advertised the lock in an advertisement which said that the Autokey hardware lock was guaranteed to run all releases of AutoCAD. The cost was $499 and a post office box number was given as the address which was to be contacted. The appellants went to extreme lengths to prevent their identity becoming known to purchasers. They did not use their surnames. They engaged in a number of procedures in relation to invoices, delivery and payment all designed to conceal their identity. In their evidence, each said that this was done deliberately, but his Honour found, nevertheless, that each of the appellants thought that the manufacture and sale of the Autokey lock was not in breach of the law.
With each AutoCAD program package was a document headed, "AutoCAD Hardware Lock, Reference Sheet". Amongst other things the Reference Sheet said that the AutoCAD licence agreement allowed the purchaser to use one copy of AutoCAD from as many computers as he or she wished provided that it was used on only one computer at any one time. If AutoCAD were used in more than one computer, the user would need to move the hardware lock from one computer to another. The lock, so the sheet said, did not prevent the user from making back-up copies of the AutoCAD software disks, nor did it prevent the user from loading AutoCAD on to a hard disk within a computer. In fact, users were encouraged to do these things.
His Honour said that the AutoCAD program could be copied easily from the soft disks contained in the package to a hard disk permanently installed within a computer. Once the AutoCAD program was stored in a hard disk, the soft disk could be removed. The computer was then capable of performing its functions pursuant to the instructions contained in the AutoCAD program in accordance with directions given to it by the operator of the computer. In performing this function, the information on the hard disk was transmitted to the random access memory (RAM) of the computer. Of course, the program was of no use unless the operator could see the information given by the computer on the screen or perhaps on a printout. These were attached to the computer through the serial or outward port of it. As mentioned, no display would appear on the screen and no printout could be made unless a lock were in place.
Written into the AutoCAD program is a set of instructions described as WIDGET.C. WIDGET.C is a program which is part of the AutoCAD program but which is present only because it forms an integral part of the locking and unlocking device which is involved in the AutoCAD lock. In lay language WIDGET.C gives instructions to the computer requiring it to send out a challenge to the AutoCAD lock. On receipt of the challenge, the lock processes it and transmits its answer to the computer. The computer receives the answer, checks it against its own instructions and, if the answer is correct, continues to operate the program. If the answer is not correct, the computer ceases to run.
The AutoCAD Lock
The actual way in which the AutoCAD and Autokey locks work are important for the determination of the two principal issues in the case, namely, whether any process or device within the AutoCAD lock or a combination of the program WIDGET.C and any process or device within the AutoCAD lock is a computer program for the purposes of the Copyright Act 1968 ("the Act"), and for the question whether, if there is such a program, the Autokey lock is a reproduction of it so as to constitute an infringement of the Autodesk copyright in the AutoCAD program. The account which follows is technical. It is taken, not from the judgment, but from a statement of facts which was given to us by counsel for Autodesk. It was hoped that it would be an agreed statement of facts but agreement was not achieved. However, it does not seem to me that the differences which the parties have indicated to us about a number of matters are of substantial significance for the outcome of the case. Accordingly, I propose to use the statement and to indicate the differences which there are between the parties about the verbiage of it.
In the case of the AutoCAD lock, information (the appellants would say voltage transitions) is transmitted from the AutoCAD program to the lock. If the lock is not in place, the AutoCAD program does not run. If the lock is in position and gives the correct response to the information (voltage transitions) received, the program continues to run. Written into the AutoCAD program is the set of instructions which I have described as WIDGET.C. This comprises instructions to the computer requiring it to send out the information (voltage transitions, sometimes described as a challenge) to the AutoCAD lock. On receipt of the challenge, the lock processes (the appellants would say responds to) the challenge and transmits its answer (response) to the computer. The computer receives the answer (response) and checks it against its own instructions. The appellants would say that it compares it with the response which it expects. If the answer is correct, the program continues to run. If the answer is not correct, the computer ceases to run. The appellants' account of what occurs goes into some detail which I do not believe to be incorrect but which may not be significant for the outcome of the case. It says that WIDGET.C assumes a different "key" and recalculates its expected response. If, after trying each of the available keys, its calculated response does not match the response received it causes the program to cease operation.
Autodesk says that the basic form of WIDGET.C is a 127 state machine table which generates a series of ones and zeros. The appellants did not agree to this statement going into the statement of facts but it is a statement taken directly from the judgment. An alternative, but less technical name, for a state machine table in this context is a look-up table.
The challenge which is sent out is a sequence of electrical impulses which, in binary notation, can be represented as a series of ones and zeros. They may also be referred to as situations where there is either present a charge or no charge. The lock responds only to transitions either from one to zero or from zero to one - it is not clear which. It is these transitions which cause the lock to operate. For this reason the binary number 10001 will give rise to the same response as the binary number 101. The response to each transition takes the form of an electrical signal which may be interpreted as either a 1 or a 0. The string of ones and zeros sent back by the lock is different from the string of ones and zeros sent by the computer to the lock. For each lock, the string of ones and zeros is identical and is the same as the string of ones and zeros expected by WIDGET.C in the AutoCAD program. Autodesk contends that it is correct to say that the string is identical with that contained in the original state machine table. That may also be correct. I do not think it matters which way the matter is put.
When WIDGET.C receives the response of the lock to the challenge, it compares it with the correct response which it extracts from its 127 state machine table. If the response is correct, it continues to execute AutoCAD; otherwise it stops. That is Autodesk's way of putting matters. In the way that the matter is put by the appellants, it is said that WIDGET.C calculates, by assuming "the key value from a limited number of available key values". If the response is correct it allows AutoCAD to continue, otherwise it causes AutoCAD to stop. The way the matter is put by Autodesk is the way that it was put by his Honour and, in any event, I see little practical distinction between the two ways of stating the matter.
A final spot in the WIDGET.C state machine is vacant but, in the way that the program works, this has the effect of what is described in the evidence as a "wrap-around". This results in a repeat search from the beginning of the state machine table. The locks also repeat their cycle after 127 states. The number of possible 127 state machines is very large being, in binary terms, one raised to the power of 127, which in the decimal system is about 10 raised to the power of 38. The resulting figure would be many billions. The AutoCAD lock uses one of these possible sequences of 127 bits. However, it is possible, by making observations of the AutoCAD lock, to discover that it operates only on transitions and that the sequential output repeats itself after 127 transitions, and also to determine the precise sequence of digits in the 127 state machine table.
The AutoCAD lock includes the following components:-
(a) A device described as a clock;
(b) a shift register;
(c) an exclusive OR gate;
(d) wiring and switching circuitry.
The function of the clock is to detect transitions and to activate the rest of the circuitry when it does so. According to the appellants the clock is an integral part of the shift register. That statement is not agreed to by Autodesk and there may be a question about it. Whether there is or not, I cannot think that the matter is of any significance. The shift register performs two functions. First it can store up to 8 binary digits. Secondly, when instructed to do so (the appellants would say stimulated), it shifts the binary digits so stored one place to the left. The exclusive OR gate (XOR gate) has two input lines and one output line. Each of the input lines can have an electrical signal corresponding to one or zero applied to it. If the two numbers applied to the input lines are different, then an electrical signal corresponding to the number 1 appears on the output line. If the two input numbers are the same, then the electrical signal corresponding to the number 0 appears on the output line.
These components are wired together in such a way that, initially, 7 binary digits of the state machine table are automatically placed (or are present) in the shift register. When each transition is detected, the contents of the shift register are moved one place to the left and the right hand cell of the shift register loaded with the digit obtained by applying the signals from cells 6 and 7 to the input lines of the exclusive OR gate and placing the digit generated by the exclusive OR gate into cell number 1. This digit is the next digit in the state machine table. In this way all 127 digits pass successively through the shift register. At each step the signal from cell 6 appears on the output line of the lock. The shift register is not in itself a pseudo random number generator. However, the combination of shift register, exclusive OR gate and associated circuitry includes (the appellants would say constitutes) a pseudo random number generator.
The individual components referred to above can be purchased commercially, but the combination cannot, that is, unless one purchases an AutoCAD lock.
The Autokey Lock
Autodesk says that the Autokey lock also has a clock. The appellants describe it as a counter. Be it clock or counter, it detects transitions and causes the electrical circuitry of the lock to move on one step at each transition. The appellants say that it causes the processes of the lock to advance one step at each transition. Again, I do not regard the difference as material.
In the case of the Autokey lock, the 127 digits are stored in an EPROM (Electrically (or Erasable) Programmable Read Only Memory). According to Autodesk, the numbers stored in the EPROM have two functions. First the least significant digits are used to generate the output. At each step the circuit advances to the next memory location and the least significant digit from that location in the EPROM is the output. The most significant digit of each number is used to control the wrap-around function. If the digit is a zero then the lock returns to the first number in the EPROM, that is, the wrap-around occurs. If the digit is a one, then the device moves on to the next memory location in sequence. The sequence is arranged so that the wrap-around function occurs after 127 steps. In the way that the appellants would put the matter, the least significant digits comprise the output. At each step the counter advances to the next memory location and the least significant digit from that location in the EPROM is output. Secondly, the most significant digit of each number is used to control the resetting of the counter. If the digit is a zero, then the lock returns to the first address in the EPROM, i.e. the counter resets. If the digit is a one, then the counter advances to the next memory location in sequence. The sequence is arranged so that the resetting function occurs after 127 steps.
The Relevant Provisions of the Copyright Act
It is next appropriate to refer to the relevant provisions of the Act. In s.10 "literary work" is defined to include, inter alia, a computer program or compilation of computer programs. "Computer program" is defined as follows:-

"'Computer program' means an expression, in any language, code or notation, of a set of instructions (whether with or without related information) intended, either directly or after either or both of the following:

(a) conversion to another language, code or notation;

(b) reproduction in a different material form, to cause a device having digital information processing capabilities to perform a particular function;"

By s.31, copyright, in relation to a work, is the exclusive right, in the case, inter alia, of a literary work, to reproduce the work in a material form and to make an adaptation of the work; see sub paras. (1)(a)(i) and (vi).
The definition of "adaptation" in s.10 of the Act is as follows:-

"'adaptation' means

(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;

(ba) in relation to a literary work

being a computer program - a version of the work (whether or not in the language, code or notation in which the work was originally expressed) not being a reproduction of the work;

(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;"

The expression "material form" is defined in s.10 as follows:-

"'material form', in relation to a work or an adaptation of a work, includes any form (whether visible or not) of storage from which the work or adaptation, or a substantial part of the work or adaptation, can be reproduced;"

There are some further provisions which should be noticed. Section 13 provides that a reference in the Act to an act comprised in the copyright in a work shall be read as a reference to any act that, under the Act, the owner of the copyright has the exclusive right to do. By s. 14, a reference to the doing of an act in relation to a work shall be read as including a reference to the doing of that act in relation to a substantial part of the work and a reference to a reproduction, adaptation or copy of a work shall be read as including a reference to a reproduction, adaptation or copy of a substantial part of a work. Finally, s.36 provides for infringement. 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 authorises the doing in Australia of, any act comprised in the copyright.
Is there a Computer Program within the AutoCAD Lock?
The first question to be determined is whether the Autodesk lock, whether alone or in conjunction with the WIDGET.C program, part of the AutoCAD program, is a computer program within the meaning of the Act. His Honour did not deal with the question whether the program which he found to exist was a composite of the WIDGET.C program and a program in the AutoCAD lock. He concluded that the AutoCAD lock itself contained within it a computer program. In relation to this matter he said:-

"Each lock is a device designed to perform a particular function namely to respond to a challenge from AutoCAD. To do this, each lock has a digital information processing capability in the sense explained earlier in these reasons. Each lock depends upon that capability to give the correct response to the challenge.

. . . . . . . . . . . . . . . . . . . . . . . . The information sent to the locks, of necessity, constitutes instructions and in the context constitutes a set of instructions. Those instructions require the lock to give an answer. In the event, the answer is implemented by a negative or possibly a positive transition. Thus on every instruction being sent by AutoCAD, there would be an even chance that the response would be the correct transition. This chance is too great to satisfy the integrity of a computer program like AutoCAD. There should be no room for chance at all. Hence, there is a process encoded within each lock to ensure that the even chance is converted almost to a certainty. This was done in the AutoCAD lock in an attempt to make the AutoCAD lock effective and difficult, one cannot say impossible, to duplicate. Thus the performance of the digital information processing capabilities within each lock is to ensure the correct answer to each challenge. These challenges occur frequently and many transitions occur with respect to each challenge. . . . . . . . . . . . . . . . . . . . . . . . The information from AutoCAD is in the nature of an instruction: tell me whether to stop or to proceed. Responding to that instruction the lock executes its digital information process capabilities in accordance with the information, being instructions, encoded in its shift register or EPROM (a reference to the Autokey lock) as the case may be, and sends information to AutoCAD. This information is in the nature of an instruction to stop or to proceed. This is received by AutoCAD which translates the information received and acts accordingly."

His Honour then referred to the judgment of Mason J. (as he then was) and Wilson J. in Computer Edge Pty. Limited v. Apple Computer Inc. (1986) 161 CLR 171 where their Honours said (p 194):-

"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."

His Honour continued:-

"What is said has application to the electrical impulses within the shift register of the AutoCAD lock and the EPROM of the Auto-Key lock.

In my opinion each of the AutoCAD lock and the Auto-Key lock is a computer program as defined in the Copyright Act and thus each is a literary work within the meaning of that phrase as used in the Copyright Act. Put another way, a computer program is contained within each lock and thus each lock is a literary work."

I should mention at this point that, as was said during the argument, it would not be correct to say that the locks themselves were literary works within the meaning of the Act. I do not think that his Honour meant to say that they were; so much is borne out by the form of his order. If Autodesk is entitled to succeed, it must show that its copyright in a computer program, as defined in the Act, has been infringed. A more correct statement, on the assumption that Autodesk is right, is that each lock contains a computer program. Nothing turns on this, but I mention it because it was a matter raised in the argument.
Counsel for Autodesk submitted that each lock processes digital information as follows. The processing can take either a transmission mode or an active mode. In the transmission mode, it transmits information down the line. In the active mode it carries out the following steps:-
(a) first, it identifies a transition;
(b) on each transition it steps forward one state and, according to the position it has reached in the program, generates a 0 or a 1 according to the rules of the program;
(c) when it has stepped through the complete program, it returns to its initial state and the cycle begins again.
It followed, so counsel submitted, that each lock is a device having digital information processing capabilities which is programmed to perform a particular function.
The governing word in the definition of "computer program" is the word "expression". For the purpose of this case, the relevant words of the definition are, "an expression, in any language, code or notation, of a set of instructions . . . intended, either directly or after . . . reproduction in a different material form, to cause a device having digital information processing capabilities to perform a particular function". The definition of "material form", in relation to a work or an adaptation of a work, includes any form (whether visible or not) of storage from which the work or adaptation can be reproduced. It may be that one does not need to have regard to the phrase "reproduction in a different material form" in the definition of "computer program"; but if one does it is, in my opinion, clear beyond doubt that the expression of a set of instructions will fit the definition if that set of instructions is reproduced in a form of storage in a device having digital information processing capabilities and it causes the device to perform a particular function.
It is to be observed that his Honour found that the computer program was contained in the AutoCAD lock. He did not find it to be contained in the WIDGET.C program or in a combination of the WIDGET.C program and the responses of the lock. There is no sign in his Honour's judgment that the case was put to him in any way other than as one in which it was claimed that the program was to be found in the lock alone. But we were informed by counsel for Autodesk that the case had been put to him in an alternative way which was put to us as the primary way in which Autodesk put its case. As I finally understood the submissions which were addressed to us by counsel for Autodesk, there were three ways in which the argument that there was a program contained in the lock was put. In descending order of priority they were as follows:-

(1) The program was in WIDGET.C, but the Court had to bear in mind that WIDGET.C and the AutoCAD lock were to be likened to two interlocking parts of a piece of machinery. It was essential that they worked together.

(2) WIDGET.C was itself the program but a substantial part of that program was reproduced in the lock. Thus the computer had the program WIDGET.C in it and the lock had part of WIDGET.C in it. In other words WIDGET.C was a program and each of the locks was a reproduction of part of that program so that the comparison was WIDGET.C and each of the locks.

(3) The device, i.e. the lock, was designed with a program built into it. This was the way his Honour looked at the matter.

In the submission of counsel for the appellants there could be no program in the lock itself. The program could only consist of the 127 bits of the string or so much of it as was contained in the shift register at any one time or, on the other hand, the underlying electronic characteristics of the devices which made up each of the locks. These consisted primarily of the shift register and the XOR gate, both of which were pieces of hardware readily available from computer suppliers. After referring to some passages from his Honour's judgment, counsel said that there could be no program in the lock itself because there was not, within the meaning of the definition of "computer program" in the Act, "an expression, in any language, code or notation of a set of instructions . . . " The set of instructions was in the WIDGET.C program within the AutoCAD program itself. The look-up table (or the 127 state machine table) was in the WIDGET.C program. It was this which generated voltages which required responses from the lock and it was this which analysed those responses for the purpose of seeing whether they were in accordance with what it anticipated receiving.
In relation to this matter, his Honour said:-

"The real difference between the parties on this issue is whether the information contained in the locks constitutes a set of instructions. Counsel for the respondents expressed their submissions on this aspect in a number of different ways. It was argued that the locks do not make calculations, indeed, it was said, they do not have the ability to do so. They do not record or store the number of 1's and 0's received from the computer or sent to the computer. There is nothing in the locks which analyses the information received. The AutoCAD does not contain any memory whether an EPROM or micro-silicon chip. The locks do not make comparisons between information received and information stored. It was said that the only thing the locks did was to respond slavishly in a pre-determined manner to stimuli received, a negative transition, to give another stimuli to the computer. It was said neither lock fetched, interpreted or executed instructions."

and:-

"Widget C is a sub-set of instructions within AutoCAD. There may be sub-sets of instructions within the set of instructions constituting Widget C. The challenge resulting from Widget C is transmitted to the AutoCAD lock. In receiving, processing and responding to that challenge, there appears to be no logical difficulty in concluding that the sub-set of instructions within the AutoCAD lock, or for that matter, the Auto-Key lock, constitutes an expression of a set of instructions, if otherwise the locks come within the definition."

In later passages to which I have already referred his Honour came to the question of function. The set of instructions must cause a device having digital information processing capabilities to perform a particular function; otherwise the definition will not apply. His Honour said that each lock was a device designed to perform a particular function, namely, to respond to a challenge from AutoCAD. Each lock, his Honour said, had a digital information processing capability and each depended upon that capability to give the correct response to the challenge. There was no argument before us that the various devices did not have digital processing capabilities. Eventually (again in one of the passages already quoted) his Honour said each lock executed its digital information processing capabilities in accordance with the instructions encoded in its shift register or EPROM as the case may be, and sent information to AutoCAD. Critically his Honour said that this information was in the nature of an instruction to stop or to proceed. In my opinion that was where his Honour took a step which led him inevitably to the conclusion that what was in the lock constituted a program within the meaning of the definition.
I think the question which is raised for decision is one of not inconsiderable difficulty. I agree with his Honour that the construction of the provisions of the Copyright Amendment Act 1984 should be approached upon the basis that they were remedial. They should not be construed narrowly. They should be construed so as to give effect to the object and purpose which the legislature had in mind in the changes which it made to the law. Nevertheless, one must give proper weight to the words which the legislature has used. It would be quite undesirable if one took such a wide view of their operation that one detected computer programs in devices which in reality could not be said to fall within the words of the definition. In my opinion the evidence establishes that the way in which the device in the AutoCAD lock works is by way of automatic response. The intricacy of the locking system is not in the lock but in the WIDGET.C program in the AutoCAD program itself. With respect, I do not agree with his Honour's conclusion that the information - the voltages - sent to AutoCAD is in the nature of an instruction to stop or to proceed. As his Honour says in the sentence which follows, the reality is that voltages or information are transmitted to the WIDGET.C program but it is that program which interprets the information and determines whether or not the AutoCAD program is to continue to operate. It follows that I am of opinion that there is not in the lock itself, looked at in isolation, a device which constitutes itself a computer program.
But I have a different view of the matter if one does, as counsel for Autodesk said should be done, that is, to look at the WIDGET.C program and the device in the lock as an integrated system. I think that is the correct way in which the matter should be approached. If it is, there can be no doubt that the two viewed in combination do constitute a computer program within the meaning of the Act.
Is the Autokey Lock a Reproduction or an Adaptation of the Autodesk Program?
The next question is whether the appellants have reproduced or made an adaptation of the whole or a substantial part of that program. It is common ground that they have not reproduced any part of the WIDGET.C program but, for the purposes of this aspect of the case, s.14 of the Act is important. It provides that a reference to the doing of an act in relation to a work or other subject matter is to be read as including a reference to the doing of that act in relation to a substantial part of the work or other subject matter and a reference to a reproduction, adaptation or copy of a work shall be read as including a reference to a reproduction, adaptation or copy of a substantial part thereof. Also important is the definition of "adaptation" which, in relation to a literary work being a computer program, means a version of the work which is not a reproduction thereof. Subparagraph 31(1)(a)(vi) of the Act confers upon the owner of copyright in a computer program the exclusive right to make an adaptation of the work. That is one of the exclusive rights which Autodesk says was infringed. Read in conjunction with the definition of "adaptation" in s.10 and the provisions of s.14 of the Act, the exclusive right to make an adaptation of a computer program which is a literary work includes the exclusive right to make versions of a substantial part of such a work.
The appellant, Mr. Kelly, did not copy the work in which I have found copyright to exist. It is common ground that what he did was to observe with the aid of an oscilloscope the passage of voltages or transitions between the AutoCAD program and the AutoCAD lock. Critical for his deductions were those which passed from the lock to the AutoCAD program. It was the regularity of the pattern which he discerned, particularly the sequence and the eventual wrap-around after the 127 bit sequence had concluded, that led him to devise the Autokey lock. The question is whether what he achieved was a reproduction or a version of a substantial part of the entirety of the literary work comprised of the combination of the WIDGET.C program and the device in the AutoCAD lock.
His Honour did not consider the question whether the Autokey lock or the component parts of it constituted an adaptation of a computer program he found to exist in the AutoCAD lock or the reproduction of an adaptation thereof. He was satisfied that the Autokey lock was a reproduction of a program in the AutoCAD lock. As earlier noted, his Honour had found that the Autokey lock, or component parts of it, was itself a computer program. That is not a question which arises necessarily for determination in the proceedings. The question is whether, be the Autokey lock a program or not, it is a reproduction or an adaptation of the program comprised in the combination of the WIDGET.C program in the AutoCAD program itself and what is in the AutoCAD lock. His Honour did not approach the matter in this way because he found that there was a computer program independently contained within the AutoCAD lock itself.
With that preamble, I come to analyse his Honour's reasons for concluding that the Autokey lock was a reproduction of the program he found to be contained in the AutoCAD lock. His Honour began by referring to the judgment of Gibbs C.J. in the Apple Computer case. The Chief Justice there said (161 CLR at p 186) that the notion of reproduction involved two elements, namely, that the infringing work sufficiently resembled the copyright work and that was reproduced by the use of the copyright work. After saying that the latter element was satisfied in the Apple case, the learned primary Judge said that the former element required that there should be a sufficient degree of objective similarity between the two works.
After this reference, his Honour continued:-

"The second element mentioned by Gibbs C.J. has been established in this case. Mr. Kelly had access to the AutoCAD lock. He made a close study of it. He made an analysis of its response to the challenge from AutoCAD. Based on the results of his studies he prepared a computer program to determine the information to be placed in the EPROM which forms the basic component of the Auto-Key lock. Based on the results of his study he produced a computer program which was designed to perform and in fact performed the same function as the AutoCAD lock."

That conclusion of his Honour is the subject of challenge in this appeal.

His Honour then turned his attention to the first of the elements mentioned by Gibbs C.J.. He said:-

"There is a difference in form between the hardware within the AutoCAD lock and the Auto-Key lock, namely the shift-register and the EPROM. But the Court has concluded that each lock constitutes a computer program. Each performs the same function. It is this function which is the essential aspect of each lock. Function has a particular importance in the definition of a computer program and regard must be given to this concept of function in considering the question of whether there is a 'sufficient degree of objective similarity' between the two locks. Having regard to the definition of computer program, it seems that the particular method by which 'a device having digital information processing capabilities to perform a particular function' is not really relevant in considering whether the Auto-Key lock 'sufficiently resembles' the AutoCAD lock. In any event, on the findings made, in my opinion the processes adopted in the Auto-Key lock to produce the same function as the processes adopted in the AutoCAD lock do sufficiently resemble the literary work of the AutoCAD lock as to come within the first element of the test propounded by Gibbs C.J."

His Honour said that he found support for his conclusion in the definition of "material form" inserted into the Act by the 1984 amendments. He referred to the fact that, before the amendments were made, no definition of "material form" was included in the Act. He added, "Now, those words, when used in relation to a work include 'any form (whether visible or not) of storage from which the work . . . or a substantial part of the work . . . can be reproduced.'"
His Honour continued:-

"In the present case, the form of the reproduction is, in my opinion, irrelevant. There is no doubt the whole of the function of the AutoCAD lock has been reproduced in a material form by the Auto-Key lock. Thus it is not necessary to consider whether a substantive part of the work has been produced. Many of the submissions put to the Court on this issue by counsel for the respondents were based upon the source code Widget C being part of AutoCAD and another source code being part of AutoCAD and described as SPTEST.C. These would be accepted as being works within the Copyright Act but the submissions proceeded on the basis that the AutoCAD lock was not a work. The argument then was that the AutoCAD lock was not a reproduction in material form of either Widget C or SPTEST.C. It was argued that there was no resemblance between any part of those programs and the Auto-Key lock. Reliance was placed upon the fact that different methods were employed between those programs and the method used in the Auto-Key lock. There may be similarity in logic, but not appearance, and therefore there could be no resemblance. Further, it was put, and is accepted, that Mr. Kelly had no access to either Widget C or SPTEST.C."

The reference to SPTEST.C is a reference to another program which is contained within the AutoCAD lock. It is not necessary to refer to the nature of it.

His Honour went on to say that the submissions did not take notice sufficiently of the 1984 amendments to the Act and added:-

"On the facts of this case a finding has been made that the AutoCAD lock and the Auto-Key lock each constitutes a computer program under the Copyright Act and thus each is itself a literary work under that Act. In this context, regard must be had to the function of the computer program in determining resemblance. Physical appearance is immaterial. The hardware or physical equipment within which the expressions of the sets of instructions are contained, is immaterial. The fact that Mr. Kelly did not see Widget C, nor for that matter, the internal mechanisms of the AutoCAD lock, are immaterial. Mr. Kelly knew of the function of the AutoCAD lock, he discovered how it performed that function, and prepared the expression for his own set of instructions for the Auto-Key lock to enable it to perform the identical function. This constitutes the resemblance between the two computer programs."

Accordingly, his Honour concluded that the infringement alleged by the respondents had been established.
It will have been observed that his Honour placed great reliance on the expression "to perform a particular function" in the definition of "computer program". He said that it was the same function which was the essential aspect of each lock. He also said that function had a particular importance in the definition of a computer program and regard must be had to this concept of function in considering the question whether there was a sufficient degree of subjective similarity between the two locks. That conclusion is, as counsel for the appellants submitted, central to his Honour's reasons for his conclusion that there was a reproduction of the program contained in the AutoCAD lock. The submission was challenged by counsel for the appellants as being erroneous. Counsel said that one could not draw from the form of the language used in the definition of "computer program" the conclusion that the legislature had intended, in the case of computer programs, to put aside the long established principle that copyright affords protection for the form of language and no protection for ideas or concepts. Accordingly, it becomes important to consider the implications of the form and definition of "computer program" in the Act and particularly the significance of the use of the word "function" therein.

In my opinion, there is considerable force in the appellants' argument. Sheppard J. has held that WIDGET. C and the lock, viewed in combination, constituted a computer program within the meaning of the Act. But, for reasons to be given shortly, it is not necessary for me to decide, and I do not decide, the point. I will assume, for the purposes of the argument, that WIDGET. C and the lock together constituted a program.
(3) The respondents' third contention: The device, that is the lock, was designed with a program built into it.
According to this view, the lock, regarded in isolation, was itself a computer program. For the reasons already given, I cannot accept this proposition.
8. The third question on the appeal: Was the Auto-Key lock a reproduction, or adaptation, of WIDGET. C?
As has been said, it is common ground that WIDGET. C was a computer program for the purposes of the Act. This is the first way the respondents put their case.
It is also common ground that the appellants have not reproduced the whole or a substantial part of WIDGET. C (cf. s.31(1)(a)(i) of the Act).
Section 31(1)(a)(vi) of the Act confers upon the owner of copyright the exclusive right to make an adaptation of the work. In relation to a literary work being a computer program, "adaptation" means a version of the work (whether or not in the language, code or notation in which the work was originally expressed) not being a reproduction of the work (s.10(1)). It is provided by s.14(1) of the Act that a reference to an adaptation of a work shall be read as an adaptation of a substantial part of the work.
The Macquarie definitions of "version" include the following:
"3. a particular form or variant of anything."
In my opinion, "version" is used in s.14(1) in this sense. The version in question must be a version of "the work", that is to say, a version of the expression of the relevant set of instructions. The only similarity between the respondents' lock and the appellants' lock was the fact that each performed the same function: each responded to incoming voltage transitions by producing as output the same string of 127 bits. Yet the algorithm employed by each was different, and the implementation of those algorithms was different. It follows, in my view, that the appellants did not adapt the work of the respondents.
9. The fourth question on the appeal: If a computer program was partly contained in WIDGET. C and partly contained in the hardware lock, was the Auto-Key lock a reproduction, or adaptation, of that program?
I have already expressed the view that the only relevant computer program was contained in WIDGET. C. I am also of the opinion that the Auto-Key lock is not a reproduction or adaptation of WIDGET. C. It must follow, in my view, that no infringement of copyright has been made out on this branch of the respondents' argument.
10. The fifth question on the appeal: If a computer program was contained entirely within the device being the hardware lock, was the Auto-Key lock a reproduction, or adaptation, of that program?
Since I am of the view that no computer program is contained entirely within the device being the hardware lock, it must follow, in my opinion, that the infringement alleged by the respondents has not been established.
Conclusion on the appeal
In the result, I am of the view that no infringement of copyright has been made out by reason of the sale of Auto-Key. It is another question, which does not now arise, whether the law should provide another form of protection in a case such as this (see, e.g. David A. Einhorn, "Copyright and Patent Protection for Computer Software: Are they mutually exclusive?", IDEA, The Journal of Law and Technology, (1990) Vol. 30(3) at p 265: see also the observations of Deane J. in the Apple Computer Case at pp 215-6).
In my opinion, the appeal should be allowed and the injunction granted at first instance set aside. The appellants should have their costs on the issues arising in the appeal both at first instance and on the appeal.
B. The issues arising on the cross-appeal
At first instance, the respondents (now the cross-appellants but for convenience here referred to as the respondents) also alleged that the appellants (now the cross-respondents) in selling the Auto-Key device were authorising a breach of copyright in the AutoCAD program by users of the program and inducing the breach of a "shrink-wrap" licence by AutoCAD program licensees. Northrop J. dismissed these claims. As summarised in the IPR headnote (at p 3), his Honour held that the appellants did not, by the sale of the Auto-Key device, authorise the reproduction of the AutoCAD program, either by reproduction in the temporary memory of a computer (RAM) by use without the AutoCAD lock or by making copies for use without the AutoCAD lock. The AutoCAD software licence, even if its restrictions were binding on the user, made no mention of the AutoCAD lock. It was held that the mere supply of the device did not, by itself, result in the user committing a breach of the licence to use the copyright in the AutoCAD.
1. The case pleaded at first instance by the respondents
Two claims were pleaded by the respondents. First, the respondents claimed that the appellants, not being the owners of the alleged computer program, and without the licence of the respondents, authorised the reproduction of the whole, or a part of, the computer program (para.6). The following particulars of that claim were given. It was said that, in order for persons to use the computer program, or any part of it, they must "load" the computer program into the memory of a computer by instructing the computer to "load" the program into its memory. A computer does this by reproducing the program in its memory. The program is designed so that upon a part of its being loaded into the memory of a computer, the program will search for the presence of the AutoCAD lock. If the computer fails to locate the lock, the program then instructs the computer to cease loading the balance of the program and to erase from the computer's memory that part of the program which has already been loaded. The sole use of the Auto-Key is to enable the computer program to be reproduced in this way. Then it was claimed that by providing the Auto-Key to these persons, they are able to load the full computer program into the memory of a computer and thereby utilize it in breach of the software licence agreement to be described below, in cases where unauthorised copies of the computer program have been made by such persons.
Secondly, the respondents claimed that, to the knowledge of the appellants, each copy of the computer program was sold subject to a software licence agreement between the respondents and the purchaser of the program (para.9). It was further claimed that it was a term of the software licence agreement that a purchaser may make a copy of the computer program for security purposes only, and may use a copy of the program, or part of it, on only one computer at a time (para.10). It was said that the appellants knew, or ought to have known, and intended, that the provision and sale of the Auto-Key lock would induce, and did induce, the said purchasers to breach the terms of the software licence agreement by (a) making a copy of the computer program other than for security purposes; (b) using a copy of the computer program, or parts of it, in more than one computer at a time (para.12). Then it was claimed that the appellants, wrongfully and with intent to injure the respondents, conspired and agreed together to induce purchasers of the computer program to breach the software licence agreement (para.13).
The case sought to be made by the respondents at the trial was based on the provisions of ss.31(1)(a)(i) and 36 of the Act. By s.36 of the Act, an infringement of copyright occurs if a person, not being the owner of the copyright and without the licence of the owner, authorises the doing in Australia of any act comprised in the copyright. By s.31(1)(a)(i) of the Act, the relevant act relied on for present purposes was the exclusive right to reproduce the work in material form. According to the respondents, the relevant work was the computer program being AutoCAD; and the person who reproduced the work in a material form was said to be the user of AutoCAD being a person who had purchased AutoCAD from a dealer of the respondents.
2. The evidence and findings at the trial
There was no dispute about the primary facts in this area.
Evidence was given by a user of AutoCAD that he purchased an Auto-Key lock from the first and second appellants, and in the presence of the first appellant, he removed the Auto-CAD lock from the computer and plugged in the Auto-Key lock. Northrop J. said that implicit in this was that, in the process, information was transcribed from the hard disk in the computer to the RAM in the computer and that this constituted a reproduction of part of the program from the form contained in the hard disk to the form contained in the RAM. His Honour said that implicit also was the fact that the user would have seen that AutoCAD operated by a representation on peripheral equipment connected to the computer.
There was also evidence that, as a general practice, dealers selling AutoCAD explain to prospective purchasers the existence and purpose of the AutoCAD lock. It is explained that AutoCAD can be run only when the AutoCAD lock is fitted to the computer, that the user can make as many copies of AutoCAD as desired but only one program can be run at any one time because only one AutoCAD lock is supplied with each package.
Further, there was evidence that among the articles contained in the AutoCAD package purchased from dealers are several small packages, each sealed in a clear plastic package. In one of the packages is the document entitled "Software licence". This agreement makes no reference to the hardware lock.
Within the outside covering within which the AutoCAD program is supplied there is set out the following notice:

"WARNING - THIS SOFTWARE IS COPYRIGHT Read this notice carefully before opening and using this package.

It is unlawful to load this software into a computer without the licence of the Copyright Owner named below. As purchaser of the authorised copy of the software contained within this package you are offered a licence by the Copyright Owner on the conditions set out below:

- to store this software in any single computer which is under your control; and

- to make copies of the software for security purposes only."

Several conditions are then set out, including the following:

"1. You may not:

- use this software, or any part or copy of it, on more than one computer at a time;

- make copies of it except for security purposes;

- loan, rent, lease or sub-licence this software, or any copy or part of it, to anyone else; - adapt this software, by decompiling it or otherwise; - or make copies of its associated documentation.

After the conditions there appears the following: "Sign and mail this registration card today "ONLY REGISTERED PRODUCT will be eligible for future upgrade. Your Authorised AutoCAD Dealer is your first point of contact for technical enquiries. However, if you need to contact Autodesk directly, you will be requested to provide your product serial number. "Completion and return of the registration card will also ensure that you will receive quarterly issues of Autodesk newsletter for AutoCAD users, Final Draft Australia."

With each AutoCAD program package was an "AutoCAD hardware lock, reference sheet." On the sheet, it was stated that the licence agreement allowed the purchaser to use one copy of AutoCAD on as many computers as the purchaser wished, provided that it was used on only one computer at any one time; that if AutoCAD were used on more than one computer, the user would need to move the lock from one computer to another; and that the lock did not prevent the user from making back-up copies of the AutoCAD software disks, nor did it prevent the user from loading AutoCAD on to a hard disk within a computer.
There was no direct evidence that purchasers of the AutoCAD package signed and returned the licence registration form, but his Honour inferred that a majority would.
There was nothing in the evidence to suggest that users of AutoCAD purchased Auto-Key to enable them to make extra copies of AutoCAD to be disposed of and to be used together with the Auto-Key lock, at a profit to third persons. There was no evidence that any person, other than a purchaser of the AutoCAD program, was using AutoCAD. There was nothing in the evidence to suggest that any AutoCAD user had used the program on more than one computer at any one time.
3. The reasoning of Northrop J. in rejecting the claims made in paras. 6 to 13 of the statement of claim
Northrop J. said that the issue was whether purchasers of the AutoCAD package had a licence to "run" AutoCAD without the AutoCAD lock. It was contended by the respondents that, in the circumstances of the case, it should be inferred or implied that the licence to use AutoCAD was subject to a condition that a purchaser would use AutoCAD only with an AutoCAD lock. His Honour rejected the contention for the following reasons:

"In my opinion, the facts of the present case do not give rise to an implied limitation on the use of the copyright in AutoCAD to the use of an AutoCAD lock. The software licence does not refer to any such limitation. There is no express licence to use copyright in the AutoCAD lock, but at least an implied licence arises but that implication does not carry with it a mandatory requirement that AutoCAD will not be used with any other device. The licence granted to use AutoCAD, at the most, is limited by the conditions set out in the software licence. There is no evidence that any user of AutoCAD has committed a breach of any of those conditions. At the time the user of the Auto-Key lock tested it, he was using AutoCAD on one computer only. On this conclusion, the principles enunciated in University of New South Wales v. Moorhouse (1975) 133 CLR 1 apply. The selling and supplying of the Auto-Key lock to users of AutoCAD does not, by itself, result in the user committing a breach of the licence to use the copyright in AutoCAD. It does not follow that AutoCAD will be used in more than one computer at any one time: see also WEA International Inc. v. Hanimex Corp (1987) 10 IPR 349; 77 ALR 456."

Northrop J. then held that the respondents had failed to establish their claim based upon the provisions of the Act. (His Honour further held that the respondents had not made out their claim that the appellants had induced the users of AutoCAD who purchased the Auto-Key lock to breach the terms of their licence agreement with the respondents, but no challenge has been made to this finding.)
4. The respondents' submission on the issue of "authorisation" on the cross-appeal
The respondents' submissions on the question of "authorisation" may be summarised as follows:
(1) There are two issues: (a) did the appellants' customers require a licence from the respondents to do what they did?;
(b) if so, and if they did not have a licence, did the appellants authorise the breach of copyright?
(2) The respondents contend that: (a) all users of AutoCAD require a licence to use AutoCAD; (b) they are not licensed to use AutoCAD except with the AutoCAD lock; and (c) the appellants, by providing the Auto-Key, have authorised their own customers to make an unlicensed use of AutoCAD and hence to infringe copyright.
On behalf of the respondents, it is said that if a person runs the AutoCAD program on a computer, he or she necessarily reproduces a substantial part of the program by copying the program from the disk into RAM. They say that, in order to run a program, there is a need to reproduce it in a material form. Both a hard disk, or a floppy disk, and the RAM, are forms of storage into which the AutoCAD program can be stored, and from which it can be reproduced. This satisfies the definition of "material form" in the statutory definition of "computer program". The respondents argue that there is no implied licence to use the component parts of the AutoCAD package in any way that the buyer wishes; in particular, there is no licence to use the AutoCAD software with the Auto-Key lock or any other such device. For to do so, they say, would destroy the very purpose for which the AutoCAD lock was introduced. The respondents rely upon the fact that the licence agreement contained within the AutoCAD package, together with the AutoCAD hardware reference sheet, referred to one hardware lock to run with each AutoCAD program. They say that this made it clear that there was no intention to grant a licence to run AutoCAD other than with the AutoCAD lock. On the question whether the appellants authorised the use of the Auto-Key by their customers, the respondents rely upon the evidence that the Auto-Key lock was in fact used by every purchaser, and the fact that the whole point of buying Auto-Key was to use it with the AutoCAD program. Then the respondents submit that the word "authorised", where used in s.36 of the Act, means to "sanction, approve or countenance" and they say that the appellants authorised the infringement of Autodesk copyright by advertising the Auto-Key lock, supplying it, encouraging the use of the lock, sanctioning, approving, countenancing and permitting such use.
5. The first question on the cross-appeal: Is a "reproduction" of the AutoCAD program in a "material form" within the meaning of the Act involved in "running" the AutoCAD program on a computer?
In the case of copyright claimed for works first made on computer and stored on disk or tape or silicon chip, there may be some degree of circularity in the definition in the Act of "material form" by reference to the concept of "reproduction" (see James Lahore, Intellectual Property in Australia, Copyright Law, at pp 2164-5). It will be recalled that, although there is no statutory definition of "reproduced", in 1984, the following definition of "material form" was inserted:

"'material form', in relation to a work or an adaptation of a work, includes any form (whether visible or not) of storage from which the work or adaptation, or a substantial part of the work or adaptation, can be reproduced;"

In the Explanatory Memorandum accompanying the Copyright Amendment Bill 1984, the following explanation of the definition was given:

"(g) 'Material form'

28. The definition of 'material form' is new and makes it clear that material form includes such methods of fixation as (sic) storage or (sic) reproduction on magnetic tape, read only or random access computer memory, magnetic or laser disks, bubble memories and other forms of storage which will doubtless be developed." (Emphasis added)

Counsel suggested that there may have been a printing error in the explanation and that it should read as follows:

"(g) 'Material form'

28. The definition of 'material form' is new and makes it clear that material form includes such methods of fixation or storage as reproduction on magnetic tape, read only or random access computer memory, magnetic or laser disks, bubble memories and other forms of storage which will doubtless be developed." (Emphasis added)

On behalf of the appellants, it is submitted that loading a program into RAM in order to run it does not constitute a "reproduction" of the program "in material form". They contend that running a program is analogous to reading a book or hearing a musical work. They say it is the very use intended to be made of the work.

There is force in the appellants' submission. However, for reasons which will be given shortly, it is not necessary to decide the point. Instead, I will assume, for the purposes of the argument, that running AutoCAD should be treated as a reproduction in a material form for the purposes of ss.31(1)(a)(i) and 36 of the Act
6. The second question on the cross-appeal: Assuming that running AutoCAD constituted a reproduction of the program, does the running of the program in conjunction with "Auto-Key constitute an infringement of copyright?
Before considering the question whether the appellants have "authorised" any infringement, it is necessary to inquire whether a primary infringement has been made out (see RCA Corp v. John Fairfax and Sons Ltd. (1981) 1 NSWLR 251; Wea International Inc. v. Hanimex Corporation Ltd. (1987) 77 ALR 456 at p 469).
Was there an infringement here? There is no infringement if a purchaser copies the program into as many computers as he or she wishes: the program is explicitly sold on that footing. That copying can be done without using the lock. The restriction sought to be imposed by the respondents at the time of sale is that only one copy of the program may be used at any one time. This is an express restriction but the present question is whether, in the absence of any express restraint on the activity, a prohibition on the use of a device similar in operation to the AutoCAD lock, such as the Auto-Key lock, should be implied.
As Mason J. pointed out in Secured Income Real Estate (Australia) Limited v. St. Martins Investments Proprietary Limited (1978) 144 CLR 596, at p 607, it is easy to imply a duty to co-operate in the doing of acts which are necessary to the performance by the parties, or by one of the parties, of fundamental obligations under the contract. But as his Honour went on to say (at p 607):

"It is not quite so easy to make the implication when the acts in question are necessary to entitle the other contracting party to a benefit under the contract but are not essential to the performance of that party's obligations and are not fundamental to the contract. Then the question arises whether the contract imposes a duty to co-operate on the first party or whether it leaves him at liberty to decide for himself whether the acts shall be done, even if the consequence of his decision is to disentitle the other party to a benefit. In such a case, the correct interpretation of the contract depends, as it seems to me, not so much on the application of the general rule of construction as on the intention of the parties as manifested by the contract itself."

In the present case, as Northrop J. pointed out (at p 31), the software licence makes no mention of an AutoCAD lock. Further, the Hardware Lock Reference Sheet does not state that no other device is to be used as a substitute for the AutoCAD lock.
With respect, I agree with Northrop J., for the reasons he gives, that a prohibition on the use of a device such as Auto-Key ought not be implied in the contract between the second respondent and a purchaser of the AutoCAD software.
It follows, in my opinion, that the first respondent's copyright in its computer program was not infringed if a purchaser of its software used the Auto-Key device.
7. The third question on the cross-appeal: Did the appellants "authorise" an infringement of copyright as a consequence of the alleged breach by a purchaser of Auto-Key of an implied term that the AutoCAD lock, and no other device, would be used to run the program?
Since I am of the view that there was no relevant infringement of copyright, it must follow that no question of "authorisation" can arise.
8. The fourth question on the cross-appeal: Did the appellants authorise an infringement of copyright by encouraging purchasers of Auto-Key to breach the express terms of their licence agreements with the second respondent by using the AutoCAD program on more than one computer at any one time?
An alternative way in which the respondents put their case is as follows. They say that that the proper inference to be drawn from the evidence is that Auto-Key was purchased by persons who had already made, or been given, unauthorised copies of AutoCAD, or were proposing to make, or obtain, unauthorised copies in the future.
I cannot accept the submission. Northrop J. found facts which contradicted the submission. He concluded (at p 31) that -

"there was no evidence to suggest that any AutoCAD user has used the program on more than one computer at any one time."

There may be other answers to the respondent's argument but it is not necessary to consider them.
9. Conclusion on the cross-appeal
It follows, in my opinion, the cross-appeal should be dismissed, with costs.
C. Orders proposed on the appeal and cross-appeal
I propose the following orders to dispose of the matter:
1. Appeal allowed.
2. Cross-appeal dismissed.
3. Set aside the injunction (Order 1) granted on 15 September 1989; in lieu thereof, order that the claims made in paras. 3, 4 and 5 of the statement of claim be dismissed; and that the whole application be dismissed, with costs.
4. Order that respondents pay the costs of the appellants of the appeal and of the cross-appeal.

[1990] FCA 499

14 SEPTEMBER 1990