FEDERAL COURT OF AUSTRALIA

EVIDENCE - admissibility on grounds of relevance - computer software - allegation of copyright infringement in relation to source code - allegation of mis-use of confidential information - whether expert evidence utilising a pseudocode analysis of the relevant software is admissible in relation to either allegation.

INTELLECTUAL PROPERTY - Copyright - computer software - allegation of copyright infringement in relation to source code - admissibility of expert evidence utilising a pseudocode analysis of the relevant software - whether pseudocode analysis is a means by which one may determine whether one source code is a substantial reproduction, publication or adaptation of another source code.

Copyright Act 1968 (Cth)  s 10(1)(a), s 29(1)(a), s 31(1)(a), s 36

Evidence Act 1995 (Cth)  s 55

Copyright, Designs and Patents Act 1988 (UK)

Powerflex Services Pty Ltd v Data Access Corporation (1997) 37 IPR 436

Autodesk Inc v Dyason (1992) 173 CLR 331

Autodesk Inc v Dyason (No 2) (1993) 176 CLR 300

Ibcos Computers Ltd v Barclays Mercantile Highland Finance Ltd (1994) 29 IPR 25

ADMAR COMPUTERS PTY LTD v EZY SYSTEMS PTY LTD & ORS

No VG 90 of 1995

GOLDBERG J

MELBOURNE

29 AUGUST 1997

RULING

I have been asked to rule on the admissibility of the evidence the applicant seeks to lead from Professor Robert John Willis and Dr Chung-Hsing Yeh in relation to the carrying out, and the results, of their investigation of the two software systems in issue in this proceeding which investigation can be described generally as a pseudocode analysis.  The case has been opened by counsel for the applicant and at the conclusion of the opening counsel for the respondents submitted that I should rule on the admissibility of such evidence.  Counsel for the applicant submitted that I should defer the making of a ruling until the conclusion of the trial but having regard to the impact which the ruling may have on the future conduct of the trial particularly in relation to cross‑examination I consider it desirable to do so at this early stage.

The applicant seeks declaratory and injunctive relief and claims damages in relation to the software comprised in a series of computer programs known collectively as the Admar Winery System (“the Admar system”) which is designed for use in the wine industry.  The applicant claims to be the owner of the copyright subsisting in those programs which include the source code and object code of the programs.  The relevant version of the Admar system relied upon in the proceedings is the February 1993 version.

The fourth respondent was employed by the applicant as a computer programmer from February 1986 to February 1993 and between September 1992 and March 1993 the second respondent contracted with the applicant to supply the services of the thirdnamed respondent to the applicant as a computer programmer.  In February 1993 the fourth respondent resigned from his employment and in March 1993 the third respondent terminated his consultancy arrangement with the applicant.  They established a software business and in May 1993 they procured the incorporation of the first respondent.  The first respondent has developed and marketed a computer system comprising a series of computer programs known as the Ezy Wine Management System (“the Ezy system”) which is designed for use in the wine industry and competes with the Admar system.  The relevant version of the Ezy system relied upon in the proceedings is the October 1994 version.

The applicant alleges that the software in its system and in particular its source code and certain documentation (not relevant for this ruling) comprises valuable confidential information belonging to the applicant which the respondents have used in developing the Ezy system, the first commercial release of which is said to have occurred in or about October 1994.  In its amended statement of claim the applicant says that the respondents have used the software in the Admar system and in particular the source code in such a manner as to constitute a breach of the obligation of confidence they owe to the applicant in relation to the confidential information of the applicant which came into the possession of the individual respondents; they also claim that the respondents’ Ezy system programs reproduce the applicant’s software or a substantial part of it and are an adaptation of it.  The applicant says that the respondents have infringed the applicant’s copyright in its software by the use and reproduction of the applicant’s source code of the software in the Ezy system or a substantial part of it.  There are also associated claims for contravention of s 52 of the Trade Practices Act 1976 (Cth) which are not relevant for the purposes of this ruling.  In their defence the respondents put in issue the applicant’s copyright in its software and they say that the Ezy system programs were developed by the first respondent as original programs and were not copied from any program of the applicant.

In support of its case that the respondents have infringed the applicant’s copyright in the source code and have used the applicant’s confidential information the applicant has tendered affidavit evidence from Professor Willis and from Dr Yeh.  Professor Willis is the Associate Dean of the Faculty of Computing and Information Technology at the Clayton campus of Monash University and Head of the Department of Business Systems and Dr Yeh is a member of that faculty.

Section 36 of the Copyright Act 1968 (Cth) (“the Act”) provides that the copyright in a literary work is infringed by a person who, not being the owner of the copyright and without the licence of the owner, does or authorises the doing in Australia of any act comprised in the copyright.  Section 31(1)(a) of the Act provides that copyright in relation to a work, inter alia, is the exclusive right in the case of a literary work to reproduce the work in a material form, to publish the work and to make an adaptation of the work.  Section 29(1)(a) provides that a literary work is deemed to have been published if reproductions of it have been supplied to the public.

The definitions in the Act of “adaptation” and “computer program” are relevant for present purposes.  Section 10(1)(a) defines “adaptation” as meaning, inter alia:

“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”.

In the context of computer programs a version, not being a reproduction, is a reference to a translation of the computer program:  Powerflex Services Pty Ltd v Data Access Corporation (1997) 37 IPR 436, 454.

Section 10(1)(a) of the Act defines “computer program” as meaning:

“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 thrust of the applicant’s case is that the respondents have copied and used the applicant’s source code and have thereby been able to release a commercial version of their Ezy system much earlier than they would have otherwise been able if they had not copied or used the applicant’s source code.  The applicant says that if the respondents had started from scratch in the development of their system it would have taken them much longer than in fact occurred to develop their system and the applicant invites the Court to infer that they have used the confidential information of the applicant thereby compressing the time within which they were able to develop their system. 

When I refer to the evidence of the applicant’s experts Professor Willis and Dr Yeh I will refer only to Professor Willis but by this reference I intend to include the work carried out by both experts and in conjunction with each other.  There is no challenge to the qualifications of the applicant’s experts to give their expert evidence.  Professor Willis was asked to compare the Admar system with the Ezy system and in order to do so he said it was necessary to examine the source code of the two systems.  Professor Willis said that:

“the term ‘source code’ is used by computer programmers to refer to any computer program written in a computer language that can be readily understood and interpreted by humans.  When the computer programmer refers to writing a computer program, the code which the programmer writes is described as source code.”

Professor Willis then noted that a computer does not generally run a source code directly and that before a computer runs software the source code is converted to object code.  Once this conversion is achieved by a compiler it is not necessary to have a source code from which the object code came to run the program but it is necessary to have access to the source code if changes are to be made to the software.

Because the source code of each of the Admar system and the Ezy system is in excess of 90,000 lines of coding in length Professor Willis examined the core modules of both systems.  Professor Willis noted that the applicant’s system used the FoxBASE+ language and the Ezy System used the dBMAN language both of which are evolved from dBASE III+ and are in the category of the X base language.

In a report which was exhibited to his first affidavit Professor Willis set out a number of similarities and differences in the programming, conventions and coding style, system structure, user interfaces, menu programs, master file maintenance programs, and other programs of the two systems.  In his conclusion (page 20) he noted that the Ezy system “as a whole” did not appear to be a line for line conversion of the Admar system and said that it used the same system, knowledge and programming techniques as the Admar system for major system features and modules.  He noted that the different coding files used by the two systems result in different detailed coding implementations and he said:

“However, the similar program design and programming logic found in some source programs of the two systems such as file naming conventions and certain module function handling procedures suggest that the same system knowledge and programming techniques are used by both systems for some system modules and programs.”

He noted that both systems had the common functionality that a winery management system would require.  Professor Willis said that:

“On the basis of the observations made by comparing two sets of selected source programs, similar program design and programming logic have been identified in some source programs of both systems”.

However, it was only in the case of the screen report viewer program of the Ezy system that he concluded that there was most likely a direct copy (with minor modifications) of a program in the Admar system.  However, he noted that:

“the similarities in design and programming logic for some modules and programs of both systems suggest that the knowledge and experience of developing the Admar system have been used in developing the Ezy system”.

He then recommended that further investigations be undertaken centred around a “pseudocode” analysis of the two systems.  It is that analysis which is the subject of the objection to relevance and admissibility.

In his first affidavit Professor Willis reiterated that the source code of the selected modules “exhibited strikingly similar program design and logic” which led him to the conclusion that the persons who developed the Ezy system “had an in depth knowledge and experience in the development of the” Admar system.  It was only in respect of the screen report viewer report program that he ventured the opinion that the Ezy program was a direct copy (with additions and minor modifications) of the Admar program.  For present purposes his important conclusion was that while his investigations had led him to believe that the Ezy system “is based” upon the Admar system, his investigation did not provide conclusive evidence apart from the screen report viewer program of copying by the respondents of the Admar system.

Professor Willis then undertook “an examination and comparison of the underlying programming logic of both systems”.  He carried out this investigation by converting the source code of various modules in both systems to “pseudocode” which Professor Willis said was:

“a method of describing the mechanism of a computer program by the use of statements that are written in free english text, consisting of nouns and verbs (for example ‘get employee work details’).  The statements are linked together either sequentially or by simple controls (for example ‘if the employee is Claude Smith return to the main menu’).  The generation of these statements is usually the final step before a computer program is written in the chosen computer language.”

He said that using pseudocode was an accepted practice in the computer industry and that it was often used in preference to flow charts as a method for designing and defining the structure of a computer program.

Professor Willis undertook his pseudocode investigation by a detailed analysis of the program logic of the two systems in which “a step by step comparison of the logic flow of both systems was examined”.  He first identified four program types used by all or most system modules in both systems, namely master file maintenance, reporting, transaction entry and transaction update.  His investigation was limited to a program logic comparison of these four program types and their representative programs namely - the grower master file maintenance program, the debtor master listing program, the debtor cash receipts entry program, the accounts payable invoice/adjustment (creditor transaction) update program and the general ledger interface update program.

His conclusion was that:

“by comparing the logic flow of selected programs, both systems as a whole use the same logic structure for performing common program tasks”.

He noted that the Ezy system had additional features which did not result in a “fundamental distinction in program functionality”.

Professor Willis continued by expressing the opinion that:

“it is extremely likely that the Admar system was used as the basis for the creation of the Ezy System.  Further there is some exact reproduction, for example in the Screen Report Viewer Program.”

Professor Willis concluded his pseudocode analysis as follows:

“In our previous report, it was determined the Ezy System was not a duplicate of the Admar System.  However, examination of the source code for the two systems reveal startling similarities which strongly indicated that the two systems were closely related.  The pseudocode analysis of the two systems clearly demonstrates that the two systems are so closely related to each other that it is extremely likely that the Admar System was used as the basis for the creation of the Ezy System.  In effect, the pseudocode of the Admar System (as examined) has been copied and incorporated into the Ezy System.  In our opinion, the programmer(s) creating the Ezy System used the Admar System as the basis for creating the Ezy System and, using the same functioning modules, altered (and sometimes improved) the Admar System so as to create the Ezy System.”

There were appended to the report of the pseudocode analysis a number of appendices which set out “the logic flow” of each of the four Admar and Ezy programs which had been examined.  These appendices incorporated the pseudocode of each of the Admar and Ezy programs analysed.

There was also in evidence hard copy print outs of each of the Admar and Ezy source codes for each of the four Admar and Ezy programs.  Each of the logic flows produced as an appendix was in fact the pseudocode of the source code which was being examined.  Dr Yeh also produced a reproduction of each of the logic flows or pseudocode used for the four respective programs annotated by the inclusion of references to the particular lines of source code upon which the pseudocode analysis was based.  Those reproductions also included a description of the logical link between the lines of the source code and the corresponding section of pseudocode.

Having produced the results of his investigation Professor Willis said in his first affidavit:

“In my opinion, the findings of the investigation conclusively demonstrated that the AWS was used as the basis for the EWMS.  The logic and flow of the modules examined were so similar as to be virtually impossible for it to be created independently.

In my opinion, the results of the pseudocode analysis definitively and conclusively show that the EWMS was written using the AWS as the basis for the EWMS.  In effect, I believe that the pseudocode of the AWS has been copied and incorporated into the EWMS.”

The respondents have filed affidavits by Dr David Bellair who challenges the evidence of Professor Willis and Dr Yeh.  However, for the purposes of determining the admissibility of the evidence of Professor Willis and Dr Yeh in relation to their pseudocode investigation and analysis and the conclusions they draw from it I have not had regard to Dr Bellair’s evidence.  I have proceeded on the basis of accepting the evidence of Professor Willis and Dr Yeh as uncontradicted in the same manner in which I would accept the truth of the allegations in a statement of claim where a respondent, on a pleading motion, would seek to strike out a statement of claim for failure to disclose a cause of action.  Further for the purposes of this ruling, I have accepted that Professor Willis’ evidence is admissible in the form in which it is given and I have not sought to deal with objections to the form of the affidavit raised by the respondents if the evidence is otherwise relevant.

Mr Golvan, who appears for the respondents, submitted that I should reject Professor Willis’ and Dr Yeh’s evidence of their pseudocode analysis and the conclusions they drew from it as the evidence was not relevant to any issue before the Court on either infringement of copyright or misuse of confidential information.  He submitted that the claim for infringement of copyright was based upon the use and reproduction of the source code of the Admar system or a substantial part of it in the Ezy system and that such infringement of copyright could only occur, in the circumstances of this case, by either reproduction, publication or adaptation.  He submitted that the pseudocode evidence did not go to any issue whether the applicant’s source code had been copied, reproduced or published nor did it go to whether there had been an adaptation of it.  By concentrating on, or directing attention to, the programming logic of the two systems the experts were, according to Mr Golvan, straying into the area of comparing the functions of the two programs and were not directing their attention to any issue which would assist the Court in determining whether or not there had been a copying of Admar’s “expression” of its source code.  Mr Golvan submitted that admitting the expert evidence would fly in the face of the principles and dicta in Autodesk Inc v Dyason (1992) 173 CLR 331 and Powerflex (supra).  

Mr Clarke, who appears for the applicant, submitted that the applicant was not saying there was an identicality of programs.  He accepted that there were differences between the two source code programs and said that the issue was one of similarity.  He relied upon s 55 of the Evidence Act 1995 (Cth) and submitted that the evidence of Professor Willis went to the probability of the existence of facts in issue, namely access to the source code of Admar and its copying.  However there is no substantial issue as to whether Mr Cugura or Mr Ellis had access to Admar’s source code.  Rather the substantial issue is whether they have used, copied and reproduced Admar’s source code.  The thrust of Professor Willis’ evidence is directed to that issue.  Mr Clarke said he was referring to access in the sense that Mr Ellis and Mr Cugura had the source code in their possession at the time they were writing the Ezy system source code.  That may be, but the issue of infringement of copyright ultimately depends upon a consideration and evaluation of the expression of the work said to be the infringing work. 

Mr Clarke submitted that the pseudocode was tendered as a tool of analysis by experts to prove that there had been reproduction, publication and adaptation of Admar’s source code.  Mr Clarke also submitted that Mr Golvan’s submissions went to weight rather than admissibility having regard to the criterion of relevance in s 55 of the Evidence Act 1995 (Cth) being whether the evidence could rationally affect, directly or indirectly the assessment of the probability of the existence of a fact in issue in the proceeding.  For the reasons to which I shall refer I reject this submission as I consider that the issue relating to the pseudocode analysis is an issue of admissibility and that if I were to accept that analysis it could not rationally affect either directly or indirectly, the assessment of the probability that the Ezy source code has substantially reproduced or published the Admar source code or that that the Ezy source code is an adaptation of the Admar source code.

In Autodesk (supra) Dawson J, who delivered the leading judgment (with which the other members of the Court agreed) emphasised the distinction between the function performed by a computer program and the expression of a computer program.  In Autodesk Inc  v Martin Peter Dyason (1989) 15 IPR 1 at first instance Northrop J concluded that the processes adopted in the competing software locks produced the same function, that the form of reproduction was irrelevant and that the whole of the function of the autoCAD lock had been reproduced in material form by the Auto Key lock.  However, as Dawson J noted, the fact that the two locks performed the same function did not mean that there was any similarity between the two sets of instructions in them and he said (344):

“Indeed, the significance placed by Northrop J upon the function of the two locks would appear to be in disregard of the traditional dichotomy in the law of copyright between an idea and the expression of an idea”.

Dawson J then referred (344) to the judgment of Lindley LJ in Hollinrake v Truswell [1894] 3 Ch 420, 427 where His Lordship said:

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

His Honour then noted that there was a particular difficulty in distinguishing an idea from its expression in the case of a utilitarian work such as a computer program and referred to an American authority to the effect that a utilitarian work “is its purpose or function and that the method of arriving at that purpose or function is the expression of the idea”. 

Ultimately, Autodesk was decided on the basis that the Auto Key lock reproduced the look-up table in the Widget C program of the autoCAD lock.  It as an integral part of the reasoning of Dawson J that in considering whether there has been an infringement of the copyright in a computer program one does not concentrate on the function performed by the computer program but rather concentrates on the two sets of instructions in the relevant programs.

An application was subsequently made to the High Court to vacate the judgment in Autodesk which application was dismissed by a majority of the Court (Autodesk Inc v Dyason (No 2) (1993) 176 CLR 300).  The reasons why the application was dismissed are not pertinent to this ruling but in the course of their reasons the Court made a number of observations relevant to the issue of the nature of protection of copyright in computer programs.  In particular Mason CJ who would have set aside the judgment, and was therefore in the minority, noted (303‑4) that the decision in Autodesk confirmed two fundamental principles:

“First, the definition of a ‘computer program’ by reference to ‘an expression ... of a set of instructions’ should be understood as conferring protection upon the set of instructions itself - which must be identified with some precision - but as doing so in a way which is adapted to the nature of copyright.  Thus the protection of computer programs is to conform to the dominant principle of copyright law that protection is given not for ideas, but only for the form of expression

...

The second fundamental proposition confirmed in Autodesk derives from the first.  Functionality is not the proper object of copyright protection.  As Dawson J stated in Autodesk, the purpose or function of utilitarian work is its idea, while the method of arriving at that purpose or function is the expression of the idea ... (citation omitted)”

The principles in Autodesk were adopted and followed in Powerflex (supra) and (at 450) the Court said:

“The definition of ‘computer program’ in the Act focuses upon the need for there to be an expression of a set of instructions.  That set of instructions must cause the computer to perform a particular function.  Consistent with copyright law, the emphasis is not upon the idea which underlines the set of instructions, but upon the expression of that idea in whatever form it may take.”

In my view admitting the evidence of Professor Willis and Dr Yeh in relation to their pseudocode investigation and analysis would be inconsistent with the principles in Autodesk and Powerflex.  Apart from the screen report viewer program Professor Willis does not venture an opinion that any particular part of the Ezy source code substantially reproduces the Admar source code nor does he give any evidence from which that conclusion could be drawn.  Indeed Mr Clarke said it is for that reason that Professor Willis goes to the next step of undertaking a pseudocode analysis.  The whole thrust of his evidence is that the Ezy source code “is based upon” the Admar source code and he uses the pseudocode to remove “the embellishments a programmer may incorporate into a program to determine whether the Admar source code has been used as a basis for writing the Ezy source code”.  His conclusion provides no basis for a finding that in the Copyright Act sense the Admar source code has been copied, reproduced or published or that the Ezy source code is an adaptation of the Admar source code.  His conclusion that the Admar pseudocode has been copied and incorporated into the Ezy source code or system is not a conclusion about the Admar source code which is the subject of the claim for copyright.

Professor Willis noted in his second affidavit (16 December 1996) that there are six ways in which software may be “copied”.  The first relates to the source code which may be copied, he says, “at the command statement level”.  I take this to be a reference to what I will call actual copying of the source code as it is expressed in hard copy.  His pseudocode analysis is not directed to this form of copying.  Another way of copying is “program logic (program design)” which he says is “the copying of program structure and its logic flow required to implement the program functionality”.  I take this to be a reference to the copying he found in his pseudocode analysis.  However, this is not a reproduction, publication or adaptation of a substantial part of the source code in the Copyright Act sense which is the subject matter of these proceedings.

In paragraphs 11 to 17 of his third affidavit (19 May 1997), Professor Willis expands upon his use of the expressions “the function, functionality, structure and program logic of computer programs”.  He says that:

“An examination of the program logic of a program reveals exhaustively the task which the program seeks to achieve.  But in order to understand what the task of the program is, it is not necessary to know the details of each and every step by which the program operates, or their sequence or structure.  A less detailed description of the task of a computer program, as understood by reference to how the end result is sought to be achieved, can be described as a functionality of the program.  The function of a program can be achieved by employing various and different program logics.”

Professor Willis’ pseudocode analysis was in fact an examination of the program logic of the two source codes.  This observation of Professor Willis does not lie easily with the principles by which one determines whether there has been infringement of copyright, that is to say whether one computer program or source code, as the case may be, substantially reproduced, published or adapted the source code in respect of which copyright is asserted.  This observation also is directed to the functions performed by the program but as Mason CJ noted in Autodesk (No 2) (supra, 304) when identifying the second fundamental proposition confirmed in Autodesk, “‘Functionally’ is not the proper object of copyright protection”.

Pseudocode analysis may be “the means” by which program logic is identified as stated by Professor Willis but it does not constitute or pretend to be a means by which one determines whether the Ezy source code is a substantial reproduction, publication or adaptation of the Admar source code in the Copyright Act sense.  That is not to say that the pseudocode analysis does not have a function.  I am of the opinion that a pseudocode analysis is relevant to the issues which arise in the claim for misuse of confidential information as they bear upon the issue whether the respondents used or had regard to the Admar source code in designing or developing the Ezy source code and used it as a springboard in that development.

The manner in which Professor Willis describes why it is necessary to use the pseudocode analysis demonstrates why it is not relevant to any issue of infringement of copyright.  He says

“In my opinion, it is necessary in order to properly compare the programs in terms of their program logic, it is necessary (sic) to perform a method of analysis which filters out matters of coding styles, techniques and special language features.  The pseudocode analysis performed by Dr Yeh under my supervision is effective, in my opinion, to achieve this result.”

As I understand Professor Willis what he is saying in this passage is that he is not comparing the expression of one source code with the expression of another source code but is rather removing a number of matters from the source code in respect of which copyright is claimed not for the purpose of determining whether there is an infringement of a substantial part of the source code but for a quite different purpose, namely the purpose of analysing what he calls the underlying program logic which analysis he undertakes by expressing the source code in a different manner.  Professor Willis’ final conclusion in his third affidavit demonstrates that he has not undertaken any task required for the purpose of establishing an infringement of copyright.  He says:

“The degree of similarity displayed between the corresponding ‘pairs’ of programs the subject of the pseudocode analysis referred to above is such that in my opinion it is extremely likely that the Admar programs were used as the basis of the creation of the corresponding Ezy Systems programs.”

Mr Clarke relied on the reasoning in Ibcos Computers Ltd v Barclays Mercantile Highland Finance Ltd (1994) 29 IPR 25 in support of his submission that in determining whether there has been an infringement of copyright in a computer program one can have regard to the structure of the program and that whether there has been a copying of the structure is a matter of degree.  It was implicit in this submission that as the pseudocode analysis analysed the structure, sequence and organisation of the source code it was accordingly providing relevant and admissible evidence.  In my view there is no principle or dicta found in Ibcos Computers which supports the proposition that the pseudocode analysis undertaken by Professor Willis provides evidence relevant to the issue of infringement of copyright in the Admar source code. 

In Ibcos Computers Jacob J was considering the relevant events by reference to the English Copyright, Designs and Patents Act 1988 under which copyright subsists in “original .... literary works” and where the expression “literary work” is defined as including, inter alia, “a computer program” which expression is not otherwise defined.  It was claimed that copyright had been infringed in computer programs (called the ADS program) relating to an accounts payroll package by the use of the source code of the programs.  The ADS program was a composite of interrelated programs, routines and sub‑routines comprising 335 program files, 171 record layout files and 46 screen layout files.  Copyright was claimed in each individual program and also in the package as a whole as a compilation.  In considering the question of originality Jacob J considered that it was a question of degree whether an “idea”, either functional, visual or literary, was sufficiently detailed so as to make it susceptible of copyright protection.  However Jacob J was not saying that simply because a work or program is functional it is capable of copyright expression.  The particular work has to be more than a general idea and by saying that the work has to be detailed, Jacob J was saying no more, in my view, than it has to be expressed in some manner.

In any event one must be careful in applying English decisions on the copyright protection afforded by the 1988 English Act which does not define “computer program” because s 10(1)(a) of the Act defines “computer program” as meaning “an expression ... of a set of instructions ...”  As Dawson J pointed out in Autodesk (supra, 343):

“The definition of ‘computer program’ by reference to a ‘set of instructions’ requires the identification of the set of instructions in question with some precision”.

Mr Clarke also relied on a passage in Ibcos Computers where Jacob J said (40):

“The consequence of regarding ADS as a whole, as a copyright work, is that it is right to consider what Mr Turner called the program structure as a whole and the design features as a whole, as part of the work, in addition to the literal bits of code and the program structure within an individual program.”

However this passage does not support the proposition that in determining whether copyright in a source code is infringed it is relevant to undertake a pseudocode analysis of the type undertaken by Professor Willis.  Jacob J was responding to the submission that there was copyright in the compilation of the interrelated program routines and sub‑routines as a whole independently of the copyright in the individual programs.  Accordingly it became necessary to look at the structure and layout of the whole of the compilation.  But that did not mean that the Court looked beyond the expression of the source code and into what Professor Willis called the underlying program logic of the source code.

Jacob J found that there were a number of similarities between the two programs which could only have resulted from disk-to-disk copying of the ADS source code.  He bore in mind that there were also many lines in the two programs which were not the same but at all times he was considering the source code and not any underlying program logic.  A comparable approach in this case would be undertaken by analysing the terms, layout and arrangement of the Admar and the Ezy source codes but not by undertaking a pseudocode analysis of them.

Finally Mr Clarke relied on a passage in Ibcos Computers where Jacob J agreed with the observation of Ferris J in the United States case Computer Associates International Inc v Altai (1992) 982 F 2d (2nd Cir); 23 IPR 385 that:

“Consideration is not restricted to the text of the code ...”

Jacob J went on (49):

“That must be right:  most literary copyright works involve both literal matter (the exact words of a novel or computer program) and varying levels of abstraction (plot, more or less detailed, of a novel, general structure of a computer program).

I therefore think it right to have regard in this case, not only to what Mr Turner called ‘literal similarities’, but also to what he called ‘program structure’ and ‘design features’.  Some of the similarities relate to the data divisions of the programs.  In Total, Judge Baker said (speaking of the data division):

The part copied can be likened to a table of contents.  It would be very unusual that that part of a book could be described as a substantial part of it.  The specification in high level language of fields and records in the data division tells one little or nothing about the costing program and so, in my judgment, cannot be regarded as a substantial part of it within the meaning of s 16(3).

I do not agree that it would be ‘very unusual’ that a table of contents would be a substantial part of a book.  It is all a question of degree.  Very often the working out of a reasonably detailed arrangement of topics, sub‑topics and sub‑sub‑topics is the key to a successful work of non‑fiction.  I see no reason why the taking of that could not amount to an infringement.  Likewise there may well be a considerable degree of skill involved in setting up the data division of a program.  In practice this is done with the operating division in mind and its construction may well involve enough skill, labour and, I add, judgment, for it to be considered a substantial part of the program as a whole.”

These observations must be read in the light of the fact that copyright was being claimed in the compilation of the many programs as well as in the individual programs themselves.  To that extent the structure of the compilation as a whole and the features of the design of the compilation as a whole and its lay out and arrangement were relevant in determining whether there had been an infringement of copyright by copying.  However there is nothing in these observations which suggests that one looks beyond the text of the source code, for example, by way of pseudocode analysis or exposition of an underlying program logic to determine whether, and to demonstrate how, there has been a reproduction, publication or adaptation of the source code.

To the extent to which one is entitled to move away from the literal text of a literary work to determine whether the copyright subsisting in it has been infringed and to look at its structure and design features, one is nevertheless limited to analysing the contents and arrangement within the four corners of the literary work.  It may be that a pseudocode analysis describes the structure, sequence and organisation of the source code and that it articulates the program logic of the source code.  But that description does not mean there has been a reproduction, publication or adaptation of the source code itself in the Copyright Act sense.  It may mean, or provide evidence, that the respondents have had recourse to, and used, the source code as a springboard for the purpose of the development of the Ezy software (on which issue I express no view at the present time as it is yet to be considered) but in such circumstances the cause of action which makes such evidence relevant is couched in terms of breach of confidence and mis-use of confidential information.

I therefore do not accept that Ibcos Computers is authority for the proposition that Professor Willis’ evidence in relation to his pseudocode analysis is relevant to and admissible in respect of any issue of infringement of copyright in a computer program.

Mr Clarke also relied on Zeccola v Universal City Studios Inc (1982) AIPC ¶90‑019 but there is nothing in the first instance decision or in the decision on appeal which supports the proposition that Professor Willis’ pseudocode analysis is relevant to any issue involved in a cause of action based on copyright infringement.

In essence Professor Willis’ pseudocode analysis concludes that the four program types analysed of each of the Admar and Ezy source codes perform the same function.  But that conclusion does not lead to or result in a conclusion of copyright infringement of the Admar source code:  Autodesk (supra), 343, 344), Autodesk (No 2) (supra, 304); Powerflex (supra, 455).

My ruling is therefore that the evidence of Professor Willis and Dr Yeh in relation to their pseudocode investigation and analysis is not admissible on the issue whether the respondents have infringed the applicant’s copyright in the applicant’s software by using and reproducing the source code of that software or a substantial part of it in the Ezy system.  That evidence is substantially found in paragraphs 16 to 22 of Professor Willis’ first affidavit sworn 14 February 1996 and exhibit RJW‑4 thereto, paragraphs 14 to 17 of Professor Willis’ third affidavit sworn 19 May 1997 and exhibit RW‑3 thereto and paragraphs 2 to 6 of Dr Yeh’s affidavit sworn 16 December 1996.

However, as I have stated earlier in this ruling I am of the opinion that that evidence is relevant to the issue whether the second, third and fourth respondents have breached the obligations they owed to the applicant as pleaded in the amended statement of claim by developing the Ezy system by using the Admar software and in particular the source code of the Admar software in its development and whether, by its knowledge of relevant matters, the first respondent has also breached the obligations it owed to the applicant.  Mr Golvan submitted that the conclusions expressed by Professor Willis did not justify a breach of confidence claim.  However, that submission goes to the weight of the evidence rather than its admissibility.  Mr Clarke seeks to draw inferences from what he says is the degree of similarity between the two programs and he relies upon the pseudocode analysis to assist in the drawing of those inferences.  The confidential information upon which the applicant relies is the source code of the Admar system and the contents of two exhibits to an affidavit of Mr Adrian Gilbert, a director of the applicant, which contained documents said to have been taken by Mr Cugura during the period he worked with the applicant and said to have been returned by him after he ceased working for the applicant.  Mr Golvan criticises the applicant’s evidence on the basis that it does not support the proposition that the respondents must have had the Admar source code in order to develop a particular aspect of the Ezy source code.  However, the overall thrust of Professor Willis’ evidence is that in many respects he concludes there has been a use of the Admar source code.  In my view Mr Golvan’s criticisms go to issues of weight rather than relevance and admissibility.  Accordingly I rule that subject to particular objections as to form and specific admissibility the evidence of Professor Willis and Dr Yeh in relation to their pseudocode investigation and analysis is admissible in relation to what I will call loosely the breach of confidence and mis‑use of confidential information claims. 

Associate:

Dated:              29 August 1997