Delrina Corp. v. Triolet Systems Inc.
Court of Appeal for Ontario
On appeal from judgments of Justice Dennis O’Leary dated February 12, 1993 and July 17, 1998.
 There are two appeals before this court, both commenced by the plaintiff, Delrina Corporation, which carries on business as Carolian Systems (Carolian). In the first appeal, Carolian appeals from the dismissal of its action by O’Leary J. against the defendants Triolet Systems Inc. and Brian Duncombe. In this action Carolian alleged that the defendants’ computer software program, known as Assess, infringed its copyright in its program known as Sysview. In the second appeal, Carolian appeals from a judgment of O’Leary J. assessing the defendants’ damages under Carolian’s undertaking to pay damages given to obtain an interlocutory injunction restraining the defendants from selling, giving away, or marketing Assess. I shall deal first with the appeal concerned with the dismissal of Carolian’s action.
THE COPYRIGHT INFRINGEMENT ACTION : LIABILITY
 Carolian raises four grounds of appeal, all of which allege errors of law on the part of the trial judge. In the circumstances, I do not think it is necessary to set out the facts in detail. They are comprehensively stated in the trial judge’s reasons for judgment, which are reported at 47 C.P.R. (3d) 1. My references to pages in the trial judge’s reasons are to the pages in this report.
 The respondent Duncombe was employed by Carolian for the period from January 1984 to December 1985 to improve Carolian’s performance monitoring program called Sysview. The purpose of this program was to allow a skilled operator of a Hewlett-Packard HP 3000 computer to assess the efficiency of the operation of this computer. Duncombe completely rewrote Sysview for Carolian.
 After leaving Carolian in 1985, Duncombe began to design Assess in January 1986. Assess was designed to be functionally similar to Sysview in order to compete directly with Sysview for the same customers. I shall refer to further facts where they are relevant to a particular ground of appeal.
The Grounds of Appeal
 The four grounds of appeal, as stated in the appellant’s factum, are as follows:
1. The learned trial judge erred in defining “copying” as only including copying which Duncombe committed while having Sysview before him, including its source code, when he created Assess.
2. The learned trial judge erred in excusing the many similarities between Sysview and Assess as being attributable to a collection of factors which are irrelevant in copyright law. These factors included: (a) Duncombe deliberately designing Assess to be similar to Sysview; and (b) Duncombe being the author of both programs.
3. The learned trial judge erred in denying copyrightability to much of Sysview. The learned trial judge made this error, in large part, by applying United States authorities on the copyrightability of computer programs without regard to the significant differences between the copyright statutes and the development of copyright law in the two countries [the United States and Canada]. Based on these authorities, the learned trial judge ignored the fact that Sysview met the standard set under the Copyright Act of being “original”.
4. The learned trial judge erred in: (a) drawing an adverse inference that Carolian did not produce a privileged expert’s report because it would not have been helpful to Carolian; and (b) finding that Fitzpatrick J. might have declined to award an interlocutory injunction had he been made aware of the Aronoff report.
 The respondents raised, as an additional issue in response to the appeal, the question whether the trial judge erred in holding, as a matter of law, that the user-interface of Sysview (specifically the screen displays), was the proper subject matter for copyright protection. In view of my conclusion relating to the trial judge’s factual findings on this issue, it is not necessary to deal with this question of law.
 By reason of the content and organization of the trial judge’s reasons, there is considerable overlap among grounds 1, 2 and 3. Further, if the trial judge is right in his reasoning relating to ground 3 (which the appellant attacks), it necessarily follows that there is no substance in grounds 1 and 2. Because I conclude that the trial judge was right with respect to ground 3, I shall deal relatively briefly with grounds 1 and 2.
1. The trial judge’s use of ‘copying’
 Carolian submits that the trial judge erred in defining “copying” as only including copying which Duncombe committed while having Sysview before him, including its source code, when he created Assess.
 The key passage in the trial judge’s reasons relating to this ground is as follows:
In using the word “copying”, Campbell [a principal of Carolian] uses it in the literal sense. He means that Duncombe had a copy of Sysview before him, including its source code, when he created Assess and that he copied from it when it benefitted his work to do so – which was most of the time. Campbell is not talking about copying from memory that Duncombe might have had for some portions of the source code, the display screens, and other portions of Sysview from having created it from scratch while at Carolian. I also in these reasons use “copying” in the literal sense I have just described (p. 9).
 It is not in issue that “copying” in the law of copyright clearly goes beyond copying from something which is physically before the person who copies. It includes copying from memory, even subconscious memory: Gondos v. Hardy (1982), 64 C.P.R. (2d) 145 (Ont. H.C.) at 159-61, 167; Boudreau v. Lin 1997 CanLII 12369 (ON SC), (1997), 75 C.P.R. (3d) 1 (Ont. Ct. Gen. Div.) at 10; Shewan v. Canada (Attorney General) 1999 CanLII 14930 (ON SC), (1999), 87 C.P.R. (3d) 475 (Ont. S.C.J.) at 496-97; and Bright Tunes Music Corp. v. Harrisongs Music Ltd. (1976), 420 F. Supp. 177 (U.S. Dist. Ct.) at 181.
 The issue with respect to this error relates to the extent to which it affects the trial judge’s ultimate conclusions on copyright infringement. It is a troublesome point but I am satisfied that, when the reasons are read as a whole, it can be seen that the trial judge’s essential findings were not based on the view that copying from memory could not be a basis of copyright infringement. As my reasons will indicate, the main basis for explaining similarities between Assess and Sysview is that features common to both programs are not capable of copyright protection.
 Further, if the trial judge was of the view that copying protectable expression from memory was excusable, he would not have felt obliged to canvass in the detail that he did what Carolian submitted to be intrinsic evidence of copying. He rejected Carolian’s submission.
 Further still, near the end of this canvass, the trial judge said:
Finally, I accept Duncombe’s evidence that he created Assess without copying Sysview in any way (p. 18). (Emphasis added.)
 For these reasons, I would not give effect to this ground of appeal.
2. Excusing copying on the basis of irrelevant factors
 Carolian submits that the trial judge erred in excusing the many similarities between Sysview and Assess as being attributable to a collection of factors which are irrelevant in copyright law. These factors included: (a) Duncombe deliberately designing Assess to be similar to Sysview; and (b) Duncombe being the author of both programs.
 This ground of appeal relates to the reasons of the trial judge which are largely repeated in relation to the question of the copyright protectability of elements of Sysview which I consider, in some detail, in relation to the issues raised under ground 3. Accordingly, I will not deal with it in any detail at this point. I will, however, make one or two observations with respect to the two factors which are expressly mentioned in the statement of the ground of appeal.
 The first factor is “Duncombe deliberately designing Assess to be similar to Sysview.” Stated in these terms, this factor could hardly be a copyright law justification of similarities between Assess and Sysview. In fact, as the appellant points out, it would be evidence which supports the finding of copying from Sysview. However, this is not what the trial judge said. He explained the similarities on the basis that it was Duncombe’s intention to make Assess capable of performing every function that Sysview could perform (p. 13). Functional similarities between these two programs is not necessarily evidence of copying. The trial judge explained this in greater detail later in his reasons when dealing with ground 3, as follows:
As Sieler [the expert witness called on behalf of the defendants] has indicated and as the decided cases have emphasized, common similarities between two competing functional works are to be expected without one being a copy of the other. The inference that copying has taken place which might arise from a pure look at two articles must be tempered by the principle that copying ideas for functional products is not enough. Once anyone has got the idea for a product, the general characteristics of the product are inevitably going to be dictated by functional considerations. Similarities attributable to the nature of the product, the limited ways in which an idea can be expressed, stock devices and common tools of the trade, the use of common sources, knowledge and information, constraints imposed by the nature of the product, are not indicators of copying or of substantial similarity between the copyright and allegedly infringing work (p. 40).
 The second factor, “Duncombe being the author of both programs” is, once again, hardly a justification of copying, if copying took place, but this clearly was not what the trial judge said. His point was that certain similarities between Assess and Sysview could be understood as the result of programming style and experience which Duncombe had, without the similarities being probative of copying.
 I would not give effect to this ground of appeal.
3. Error in denying copryrightability to much of Sysview
 Carolian submits that the trial judge erred in denying copyrightability to much of Sysview. According to this submission, the learned trial judge made this error, in large part, by applying United States authorities on the copyrightability of computer programs without regard to the significant differences between the copyright statutes and the development of copyright law in Canada and the United States. Following U.S. authorities, the learned trial judge ignored the fact that Sysview met the standard set under the Copyright Act of being “original”.
 I should mention, at the outset of this part of my reasons, that the evidence the trial judge relied upon showed, in his view, both that the portions of Sysview which Carolian submitted proved copyright infringement by the defendants were not entitled to copyright protection and, in fact, had not been copied by Duncombe from Sysview. On this the trial judge said:
We have then Burkowski [an expert witness called on behalf the plaintiff] and other witnesses for the plaintiff saying that certain parts of Assess were copied from Sysview but making no attempt at establishing that Carolian had copyright in the parts of Sysview that were copied. We have Sieler [the defendants’ expert] saying that the portions of Sysview, that Burkowski and others referred to, were not copied by Duncombe in creating Assess, and in many cases that the similarities are accounted for by being the logical or only way of accomplishing the particular task at hand, or that the routine came from the public domain or was dictated by the make-up of the HP3000 or other like reasons – all of which in addition to explaining away copying, also deny copyrightability (p. 38).
Sieler lists many reasons for similarities between Sysview and Assess without the suggestion of copying arising, but reasons which likewise establish that those parts of Sysview under [re]view are not copyrightable (p. 41).
 Carolian advanced several specific arguments under this ground of appeal. One fundamental argument was that the trial judge erred in dissecting Sysview into parts and then determining that each of these parts was not entitled to copyright protection. The proper approach, Carolian submits, is, first, to determine whether Sysview as a whole is entitled to copyright and then to determine whether the part reproduced by the defendants is a substantial part of the whole. This latter step involves looking at the quality and quantity of what has been taken. Carolian submitted that the law is “neatly encapsulated” in Ladbroke (Football) Ltd. v. William Hill (Football) Ltd.,  1 All E.R. 465 (H.L.).
 I shall deal with this basic submission now. The House of Lords’ decision in Ladbroke clearly supports the propositions advanced by Carolian, but I think that the major answer to the appeal on this point is that the trial judge, in his decision of the case, did not run afoul of these propositions. In fairness to Carolian, I acknowledge that there are several general statements in the trial judge’s reasons at pp. 27-46 which would tend to support its position. However, in his actual decision on the issue of infringement, the trial judge engaged in what I consider to be a proper analysis relating to whether the defendants reproduced a substantial part of Sysview. I shall elaborate.
 The reasons of Lord Pearce in Ladbroke at p. 481 are in point:
Did the appellants reproduce a substantial part of it? Whether a part is substantial must be decided by its quality rather than its quantity. The reproduction of a part which by itself has no originality will not normally be a substantial part of the copyright and therefore will not be protected. For that which would not attract copyright except by reason of its collocation will, when robbed of that collocation, not be a substantial part of the copyright and therefore the courts will not hold its reproduction to be an infringement. It is this, I think, which is meant by one or two judicial observations that “there is no copyright” in some unoriginal part of a whole that is copyright. They afford no justification, in my view, for holding that one starts the inquiry as to whether copyright exists by dissecting the compilation into component parts instead of starting it by regarding the compilation as a whole and seeing whether the whole has copyright. It is when one is debating whether the part reproduced is substantial that one considers the pirated portion on its own. [Emphasis added]
 This approach has been followed in John Richardson Computers Limited v. Flanders,  F.S.R. 497 [Ch. D.] at 548 and in Hospital Software Inc. v. Hospital Medical Records Institute reflex, (1994), 57 C.P.R. (3d) 129 (B.C.S.C.) at 272-73.
 The trial judge in his reasons at pp. 71-82 carefully examined similarities between Sysview and Assess upon which Carolian relied and concluded that none of them involved parts of Sysview that were entitled to copyright protection or that, in fact, were copied by Duncombe from Sysview. This analytical approach is fully supportable under the emphasized passages in Ladbroke set forth above. While it would have been preferable if the trial judge had avoided using “copyrightable” in considering the elements of Assess alleged to be similar to Sysview (the word applies more accurately to a work as a whole), I think that he properly considered whether those elements were entitled to copyright protection.
 In Ladbroke, the plaintiff alleged that the defendant had infringed its copyright in a football betting coupon which was a “compilation” form of “literary work” under the applicable legislation. The first line of defence was that the plaintiff did not have copyright in the coupon as a whole. In pursuing this position the defendants argued that “the coupons could be dissected and that on analysis no copyright attached to any of their component parts and accordingly no protection is available” (p. 475). This approach was rejected by the House of Lords.
 In the present case, the trial judge at no place in his reasons expressed the view that there was no copyright in Sysview as a whole and, for the purpose of these reasons, I am prepared to assume that Carolian had copyright in the work as a whole. It was sufficient for the trial judge to examine those elements of Assess which Carolian alleged were similar to elements of Sysview, with a view to determining whether Assess was a substantial reproduction of Sysview. If the elements of Assess Carolian alleged were similar to Sysview were not capable of copyright protection, Carolian cannot rely on those similarities to establish copyright infringement.
 I accept that the reproduction of a particular arrangement of elements that are not themselves protectable can constitute copyright infringement if that arrangement is original. However, in this case the trial judge found that all of the alleged similarities between Sysview and Assess, including similarities in the arrangement of elements, were dictated by functional considerations or otherwise not protectable by copyright. Given this finding, it could hardly be concluded that Assess was a substantial reproduction of Sysview, even if the Sysview elements, by reasons of their collocation with other parts of Sysview, were part of a work which was entitled to copyright as a whole.
 The foregoing relates to the particular similarities upon which Carolian relied. The trial judge also looked at the aggregate of what was alleged to be copied. First, with respect to the user interface (which includes the display screens and keyboard commands) he held, contrary to the submission of the defendants, that it was legally capable of being protected by copyright (pp. 27-31) but found that, on the evidence, no substantial part was “copyrightable” (pp. 44 and 46). Second, with respect to the source code part of the program, he was satisfied that portions of it were not copyrightable but, more importantly, that none of it (apart, unwittingly, from a de minimus 40 lines out of a total of 14,000 lines – pp. 29 and 31) was copied by Duncombe (pp. 39 and 46). On this point he made the following significant finding:
After examining the source code for both Sysview and Assess, Sieler stated:
In the clearest possible terms, I want to state: I do not believe that the source for Sysview was altered into the source code for Assess. … Both Sysview and Assess show rigid, but different, coding disciplines.
No independent expert has come forward to say that the Assess source code was copied from Sysview. Accordingly, the evidence of Sieler goes a long way in convincing me that Duncombe, in creating Assess did not copy from the Sysview source code (p. 39).
 The foregoing, in my view, addresses one important aspect of Carolian’s submission on this part of the appeal. I turn now to other aspects of the submission.
 These aspects all relate, in various ways, to what Carolian submits is the trial judge’s incorrect application of the term “original” in copyright law, specifically, that the trial judge ignored the fact that Sysview met the standard under the Copyright Act of being original. This error, according to the submission, stems from the trial judge’s reliance on U.S. authorities as opposed to Canadian and English authorities.
 It is a fundamental feature of the copyright law in all three countries that it protects only original expression. It does not protect the idea underlying the expression. Frequent reference has been made to the following statement of Thorson P.in Moreau v. St. Vincent,  Ex. C.R. 198 at 203.
It is … an elementary principle of copyright law that an author has no copyright in ideas but only in his expression of them. The law of copyright does not give him any monopoly in the use of the ideas with which he deals or any property in them, even if they are original. His copyright is confined to the literary work in which he expressed them. The ideas are public property, the literary work is his own. Everyone may freely adopt and use the ideas but no one may copy his literary work without his consent.
 Although the idea/expression dichotomy is a common feature of copyright law in the three countries, it has been observed that it is applied with greater rigour in the United States, with the effect of enlarging the idea aspect of a work and, correspondingly, reducing the expression aspect. The result is a narrowing of the scope of copyright protection. The submitted wider protection afforded under the English/Canadian approach is based on some recognition of the skill and labour in the creation of the work:
While Anglo-Canadian copyright law also places significant emphasis on the idea/expression distinction, this principle has not been applied with the same rigour. Canadian and British courts have been willing to depart from this principle and grant copyright protection based on the skill and labour used in the creation of a work. This departure has created some doctrinal tension. In choosing to protect skill and labour, British and Canadian courts have accorded a certain degree of protection to ideas.
Michael F. Morgan, “Canadian Copyright and Computer Software: Back to the Future?” (1995), 12 C.I.P.R. 162 at 173-4.
 Accepting that there may be this difference in the law, it has been recognized in Anglo-Canadian law that the non-protection of ideas embraces the view that there is no copyright in any arrangement, system, scheme or method for doing a particular thing or process. Several English and Canadian decisions in support of this statement are set forth in Sookman, Computer, Internet, and Electronic Commerce Law (1991- ) p. 3-151 at footnote 644.117.
 I refer also to the World Trade Organization (“WTO”) Agreement on Trade Related Aspects of Intellectual Property Rights (“TRIPS”) which was incorporated into Canadian law by S.C. 1994, c. 47, s. 8. It provides in Art. 9.2 that “[c]opyright protection shall extend to expressions and not to ideas, procedures, methods of operation or mathematical concepts as such.” (Emphasis added.) I refer to the emphasized words as showing what, authoritatively, falls outside the scope of protectable expression. Cf. Apotex Inc. v. Wellcome Foundation Ltd. 2000 CanLII 16270 (FCA), (2000) 10 C.P.R. (4th) 65 (Fed. C.A.) at 84.
 The foregoing discussion on the scope of protectable expression in “original” works relates to Carolian’s objection to what it submits was the trial judge’s application of the U.S. doctrine of “merger” and the “abstraction – filtration – comparison” method of determining the protectable expression in the course of determining whether there has been a substantial reproduction. I shall return to these matters. I turn now to another objection, of a different order, which Carolian has raised with respect to the meaning the trial judge gave to “original” in copyright law.
 Carolian objected to the following statement in the trial judge’s reasons:
In my view, Sysview, as rewritten and improved by Duncombe, is not the inspired, elusive solution to some unrelenting, intractable problem relating to the performance monitoring of a HP3000 computer (p. 26).
 The trial judge made this statement during his consideration of Carolian’s claim based on breach of a fiduciary obligation by Duncombe, which is no longer in issue in this proceeding, and not in a copyright law part of the case. With respect to the latter, he stated the law correctly:
Perhaps the next most fundamental principle is that to give rise to copyright, the work must be original, that is to say, it must not have been copied by the author from another work, whether that other work was protected by copyright or was in the public domain and free for the taking (p. 32).
Copyright exists in original literary works. There is no copyright in what the author has copied from something already in the public domain or from the work in which another holds the copyright (p. 41).
These statements are in accord with the leading authority, University of London Press Ltd. v. University Tutorial Press Ltd.,  2 Ch. 601 at 608.
 Carolian also raised a procedural issue relating to the originality of Sysview. Carolian submitted that the trial judge misplaced the legal burden of proof relating to the existence of copyright. The Copyright Act, s. 34.1(1)(a) provides that “Copyright shall be presumed, unless the contrary is proved.” There are passages of some generality in the trial judge’s reasons which misstate the law on this issue. For example: “For Carolian to succeed on its copyright infringement claim, it must, of course, first establish it has copyright which subsists in Sysview or some material portion of it” (p. 27).
 It is clear, however, that when the trial judge dealt with the issues on which the case turned, relating to copyright protectability in the parts of Sysview which Carolian alleged were copied, he came to a determinate conclusion that Sysview had no protection:
In my view, Sieler’s evidence in regard to each of the examples of alleged copying I have just reviewed, while relevant to the issue of copying, establishes if accepted (and I accept it) that those portions of Sysview’s interface referred to in those examples were not copyrightable… At the same time, Sieler has convinced me that for many of the reasons that parts of the user interface are not copyrightable, for those same reasons portions of the source code are not copyrightable. Little will be accomplished by identifying those portions of the Sysview source code that are not copyrightable since I have concluded that in any event none of it was copied by Duncombe (p. 46).
This shows that the trial judge’s findings did not flow from any error in the application of the proper burden of proof.
 I return to the substantive application of the requirement of originality. As I have indicated, the appellant submitted that the trial judge, in applying U.S. authorities, erred in applying the doctrine of merger and the “abstraction – filtration – comparison” method of determining whether certain portions of a work are entitled to copyright protection. Both these issues relate to the idea/expression dichotomy which I have earlier discussed .
 The trial judge in his reasons (at pp. 57-65) referred extensively to Computer Associates International Inc. v. Altai Inc., 23 U.S.P.Q. 2d 1241 (2nd Cir. 1992) which discussed and applied the “abstraction – filtration – comparison” method. However, following this, he said at (p. 37) “[w]hether a Canadian court should adopt the abstraction – filtration – comparison method in deciding an action for copyright infringement or some other similar method” it was clear that “some method must be found to weed out or remove from copyright protection those portions which, for the various reasons already mentioned, cannot be protected by copyright”. This indicates that the trial judge did not necessarily apply the abstraction – filtration – comparison method, assuming it, for the moment, to be “wrong.” Further, I see nothing wrong with his general “weeding-out” observation.
 Before referring to the trial judge’s analysis of the evidence relating to the parts of Assess which Carolian alleged were similar to parts of Sysview, I will say something more on “ the law” relating to the abstraction – filtration – comparison method, which is described in some detail in the trial judge’s reasons at pages 33-37. As I have said, it is a method of determining whether a part of a work is entitled to copyright protection in the process of determining whether a defendant has substantially reproduced a plaintiff’s work.
 In John Richardson Computers Ltd. v. Flanders and Another (1993), 26 I.P.R. 367 (Ch. D) Ferris J. generally approved the “abstraction – filtration – comparison method.” He said at p. 392:
But at the stage at which the substantiality of any copying falls to be assessed in an English case the question which has to be answered, in relation to the originality of the plaintiff’s program and the separation of an idea from its expression, is essentially the same question as the United States court was addressing in Computer Associates. In my judgment it would be right to adopt a similar approach in England.
 The method did not find quite the same favour in the subsequent English decision of Ibcos Computers Ltd. v. Barclays Finance Ltd.,  F.S.R. 275 (Ch. D) when Jacob J. said, after referring to John Richardson, at p. 302:
For myself I do not find the route of going via United States case law particularly helpful. As I have said, United Kingdom copyright cannot prevent the copying of a mere general idea but can protect the copying of a detailed “idea.” It is a question of degree where a good guide is the notion of overborrowing of the skill, labour and judgment which went into the copyright work.
 It can be seen from the foregoing that there are differences of view on the method or process to be followed in performing a substantial reproduction analysis. I do not think that a hard-edged question of law is necessarily involved. I shall return to this after I have addressed the question “merger”.
 With respect to merger, the trial judge made the following statement at page 41 in his enumeration of “some general principles applicable to the law of copyright”:
If an idea can be expressed in only one or in a very limited number of ways, then copyright of that expression will be refused for it would give the originator of the idea a virtual monopoly on the idea. In such a case, it is said that the expression merges with the idea and thus is not copyrightable.
 The merger doctrine developed in U.S. case law has been subjected to criticism in England (Ibcos Computer Ltd. v. Barclays Finance Ltd.,  F.S.R. 275 at 290-92 (Ch. D.) and Canada (Apple Computer Inc. v. Mackintosh Computers Ltd. reflex, (1986), 10 C.P.R. (3d) 1 at 21-27 (F.C.T.D.), aff’d (1988), 18 C.P.R. (3d) 128 (F.C.A.); aff’d 1990 CanLII 119 (SCC), (1990), 2 S.C.R. 209.) In Ibcos, Jacob J. said at p. 291:
The true position is that where an “idea’ is sufficiently general, then even if the original work embodies it, the mere taking of that idea will not infringe. But if the “idea’ is detailed, then there may be infringement. It is a question of degree.
 The merger question was raised in rather stark form in Apple Computer. This was not a case of infringement based on an alleged substantial reproduction of the plaintiff’s work but one of an outright copying of the whole work. One of the defendant’s defences was the general submission that “a computer program exhibits a merger of the idea and the expression of the idea and therefore is not copyrightable” (p. 24).
 Reed J. was of the view that there were several difficulties with this argument. One of them was that the assembly code version of the program in question had been conceded to be copyrightable. I do not think that the court’s rejection of the merger argument that computer programs, generally, are not entitled to copyright necessarily forecloses its application to a particular element in a program in the course of a substantial reproduction analysis. As Ibcos said “[i]t is a question of degree.”
 The merger notion is a natural corollary of the idea/expression distinction which, as I have said, is fundamental in copyright law in Canada, England and the United States. Clearly, if there is only one or a very limited number of ways to achieve a particular result in a computer program, to hold that that way or ways are protectable by copyright could give the copyright holder a monopoly on the idea or function itself.
 I have set forth and, to some extent, dealt with Carolian’s submissions relating to the trial judge’s application of U.S. law. I deal now with the trial judge’s specific reasons for concluding that Assess was not a substantial reproduction of Sysview.
 The trial judge addressed, at pp. 41-46, the parts of Sysview which Carolian alleged were copied in Assess. As part of this analysis he was obliged to consider the defendants’ evidence that these parts were not entitled to copyright protection and, in fact, had not been copied. I say at the outset that his rather detailed canvass of the points of alleged similarities between Sysview and Assess shows that the source of similarities was not Sysview but, rather, other sources. Many were dictated by Hewlett Packard’s operating system or reflected common programming practices. They were not the original expression of Sysview.
 The following is an outline of the trial judge’s canvass:
(a) The fact that some of the “names” of items on which the measurement interface will report were the same in Sysview and Assess was not significant because they were “the obvious or logical or customary words to use” (pp. 41-42).
(b) A number of programming problems may only be solved in one or two ways, accounting for similarities between Sysview and Assess. For example, the data respecting how to determine the processor time (CPU) used by a process is kept in only one location and there is only one basic approach to retrieving it (p. 42).
(c) Similar lines of source code in Sysview and Assess are common throughout the HP3000 community (p. 42).
(d) With respect to the method of terminal input/output (use of centralized routines), because all well-written programs have centralized terminal input/output, Assess would be expected to have this (p. 42).
(e) Respecting the technique used to log program information and determine file names, the predominant means of obtaining process information in both Sysview and Assess is similar to that found in SOO, an early contributed library program (pp. 42-43).
(f) With respect to similarities in the technique to gather information about the system, the predominant means of logging information system information in both Sysview and Assess is the Measurement Interface, documented by Hewlett Packard. It is not proprietary to Sysview (p. 43).
(g) Respecting similarities in screen layouts and similar information, many of the displays are similar to SOO and the similarity of the information is accounted for because it comes from the same source, the Measurement Interface. Further, other performance monitors, OPT and Probe, would produce similar information (p. 43).
(h) Small chunks of code in Sysview and Assess are very similar but are also similar to publicly available code (p. 43).
(i) Much of the similar design structures come from the design of the Measurement Interface and some of the design of both programs is similar to the functions of the OPT programs (p. 43).
(j) The common screen handling approach is the same as that used by OPT (pp. 43-44).
(k) Respecting the similarities in the screen displays, much of this is accounted for by the Hewlett Packard manual and the order provided for by the HP Measurement Interface – and the content is similar to that in SOO, OPT or other previously existing monitor tools (p. 44).
(l) Similarities in keyboard commands are accounted for by being generic to users of HP 3000 computers, by sharing an approach used by OPT and many spreadsheets which existed before either Sysview or Assess, by being the logical command to accomplish a particular purpose, and by being the simplest for the user (p. 44).
(m) Other similarities in parts of the user interface are accounted for by both being similar to OPT and SOO or one or the other (pp. 44-46).
 The foregoing is based on the evidence before the trial judge, particularly the expert evidence of Mr. Sieler, whose evidence the trial judge preferred to that of Carolian’s expert witness. I do not think that the analysis shows any denial of copyright protection to ideas reflecting skill and labour on the part of Carolian which would attract protection under what is said to be the less rigorous English application of the idea/expression dichotomy.
 I would not give effect to this ground of appeal.
4. The trial judge’s drawing of an adverse inference
 As indicated above, Carolian submits that the trial judge erred in drawing an adverse inference that Carolian did not produce what it submits to be a privileged expert’s report because it would not have been helpful to it, and in finding that Fitzpatrick J. might have declined to award an interlocutory injunction had he been aware of the report of this expert.
 The trial judge said at p. 17 in his reasons:
Carolian in the fall of 1986, sent copies of both the Sysview and Assess source codes to an expert, Mr. Lowell Arnoff, for his opinion as to whether that of Assess had been copied from Sysview’s. Arnoff reported to Carolian but his report has never been produced and he did not testify at the trial. I assume his evidence would not have helped Carolian.
Of greater concern is the fact that at the time of the successful motion for an interlocutory injunction, on February 3, 1987 before Fitzpatrick J., the existence of a report from Arnoff was not revealed to the Court or the defendants. Fitzpatrick J. might well have refused the injunction unless the report was made available to him.
 Carolian submitted that this passage shows that the trial judge drew an adverse inference against Carolian because of its failure to call Mr. Arnoff as a witness at the trial and that he was wrong in doing so. (Carolian says that the correct name is “Aronoff.” It appears that all of the references in the material at the trial and in the pre-trial proceedings are to “Arnoff” and so I will continue with this version.) The reasons are that the Arnoff report was privileged and, in any event, there was an insufficient evidentiary basis for the drawing of such an inference.
 Following the hearing of this appeal, it was not apparent what the actual evidence was that bore on the Arnoff report and its alleged privileged status. We, accordingly, asked for further submissions in writing on these questions. We have received detailed submissions in response to our request. It is fair to say that the issues each side now raise respecting the Arnoff report go substantially beyond what was put before the trial judge and before this court in the factums and oral argument.
 Carolian now, in its written reply submission, raises the fundamental issue that there was no admissible evidence at all before the trial judge on the subject of the Arnoff report. It also submits, as it did on the hearing of the appeal, that the Arnoff report was privileged and, further, that there was an insufficient evidentiary basis for drawing an adverse inference.
 The defendants contend that
(1) The judge did not draw an adverse inference. The comment was merely an acknowledgement of facts made known to him during the trial.
(2) In any event, the trial judge’s comments had no influence on his conclusions.
(3) No privilege attached to Mr. Arnoff’s report. Rather, Carolian was under an implied, if not an express, obligation to disclose the contents of Mr. Arnoff’s opinion.
(4) Even if the Arnoff report was privileged, it would have been proper to draw an adverse inference from his failure to testify.
 The record is unclear with respect to the “evidence” before the trial judge on the Arnoff report. There is an eleven page passage in the trial transcript which deals with the report, to which both sides have referred, but most of it contains counsel’s statements and not sworn evidence. I think that the trial judge was relying on this part of the “evidence” in making his comment on the failure to call Mr. Arnoff as a witness.
 It appears from the written submissions made by Carolian to the trial judge that it did not take the position that he could not deal at all with the defendants’ adverse inference submission relating to Mr. Arnoff on the basis that there was no admissible evidence relating to it. Rather, it submitted that both O’Brien J. and Master Donkin, in pre-trial orders, had denied the defendants’ motions for production of the “alleged” report and that the defendants could not “now, complain about the plaintiff’s expert after having failed to call two of their own experts and subjecting them to cross-examination.”
 It is not necessary to deal with all of the questions raised on this ground of appeal. They include issues relating to the interpretation of pre-trial proceedings relating to the coming into existence of the Arnoff report, litigation privilege, and issue estoppel within the confines of a single proceeding. I am satisfied that the point can be dealt with on the short ground that this portion of the trial judge’s reasons had no bearing on how he arrived at his ultimate conclusion.
 Contrary to the submissions of the defendants, I think that the trial judge, in the quoted passage from his reasons, did draw an adverse inference. He assumed that Mr. Arnoff’s evidence would not have helped Carolian. While his thinking could have been spelled out at greater length, the wording is a recognized terse way of drawing an inference. Sopinka et al, The Law of Evidence in Canada, 2nd ed. (1999) at p. 297 says: “Such failure [to call a witness who would have knowledge of the facts and would be assumed to assist that party] amounts to an implied admission that the evidence of the absent witness would be contrary to the party’s case, or at least would not support it.” It may be seen that the trial judge drew the milder of the two inferences referred to in this passage. In any event, when the passage is examined in the context of the reasons as a whole, it can be seen that it had no bearing on the trial judge’s ultimate findings.
 The passage was in that part of the reasons concerned with the question of factual copying and, no doubt, was relevant to this issue. The defendants’ expert witness, Mr. Sieler, had testified in some detail in support of his report which stated that “Assess is new and has not been copied from the source code for Sysview” (p. 17). The trial judge accepted his evidence, as well as that of Mr. Duncombe (… “I accept Duncombe’s evidence that he created Assess without copying Sysview in any way” (p. 18)) and referred to portions of the evidence which supported his conclusion of no copying.
 It appears to me that what the trial judge said about the failure to call Mr. Arnoff was purely cumulative to his other, independent, findings respecting copying. I do not think that it was a factor which had anything to do with these findings.
 However, if I am wrong in thinking that this inference was not a factor in the trial judge’s conclusion on actual copying, it is very clear that it formed no part of his reasoning with respect to issue 3, that relating to whether Assess was a substantial reproduction of Sysview. As I have indicated, the trial judge’s reasons on this issue are an independent basis of his ultimate judgment. The reasons on this issue are quite separate from those relating to actual copying, even though there is some overlapping in the findings in the two parts.
 Accordingly, if this was not a proper case in which to draw an adverse inference, I am satisfied that the impugned passage in the reasons is, at most, harmless error.
 In light of the foregoing, it is not necessary to determine whether this was a proper case in which to draw an adverse inference and I shall not do so. I will, however, give some indication of the nature of the terrain that would have to be covered in addressing this issue.
 Carolian submits that no inference should be drawn because the report made by Mr. Arnoff was held to be privileged in a decision of Master Donkin on March 29, 1989 on a motion by the defendants for production of the report. Master Donkin held that “the report was for the sole purpose of this litigation.”
 The defendants argued that because the parties exchanged their source codes, pursuant to two earlier court orders, to have them compared by a court-appointed assessor, the report was not privileged, even though the scheme envisaged by the orders was aborted by reason of the designated assessor not accepting the appointment.
 The defendants’ counsel in that part of the trial covered in the eleven page segment I mentioned earlier said to the trial judge that he “had no intention of now trying to overturn the claim of privilege and an order of Master Donkin.” In light of this, the trial judge would not have been wrong in considering that Mr. Arnoff’s evidence was privileged. However, it appears from his reasons that he did not advert to the effect of Mr. Arnoff’s evidence being privileged.
 On the subject of privilege, the defendants submit that: (a) the trial judge was not bound by Master Donkin’s decision, citing Talbot v. Pan Ocean Oil Corporation (1977), 4 C.P.C. 107 (Alta. App. Div.) and Pocklington Foods Inc. v. Alberta (Provincial Treasurer) reflex, (1995), 28 Alta. L.R. (3d) 96 (C.A.) (they do not deal with their counsel’s concession before the trial judge on this issue); (b) even if the report were privileged, this would not prevent the drawing of an adverse inference, citing Taylor v. The Queen 1977 CanLII 245 (BC SC), (1977), 5 B.C.L.R. 23 (S.C.) and Bates v. Stubbs 1979 CanLII 753 (BC CA), (1979), 15 B.C.L.R. 65 (C.A.) incorporate B cases on p. 28A; and (c) assuming the trial judge was bound by the earlier order, this court is not bound by it, citing Spencer Bower, Turner, and Handley, The Doctrine of Res Judicata, 3rd ed. (1996) at pp. 79-80. Reference may be made to Schiowitz v. I.O.S. Ltd.,  2 O.R. 262 (C.A.) for a decision contrary to this view.
 On the subject of the preclusive effect of an earlier decision, within a given proceeding on later decisions in the proceeding, Watson et al in The Civil Litigation Process – Cases and Materials, 5th ed. (1999) says at page 535 “… in terms of statements of general principles, the case law is in hopeless confusion.”
 As indicated, these issues received little or no attention at trial. For example, if the Arnoff report was protected by lawyers’ privilege, it would seem that the defendants could have obtained disclosure of Mr. Arnoff’s findings, opinions, or conclusions, but not if Carolian had undertaken not to call him as a witness at the trial (rule 31.06(3)(b)). If the matter had been gone into, it is clear that Carolian would have given this undertaking. If it had given it, this would have been a logical explanation which would have prevented the adverse inference from being drawn.
 At this point I shall deal with a further complaint of Carolian on the trial judge’s reference to the Arnoff’s potential evidence. Carolian submits that the trial judge’s view of the matter resulted in his awarding solicitor and client costs against it. I have two observations on this point. The first is that there is no appeal against costs – no motion for leave to appeal costs and, if leave be granted, a request to have the costs order varied (see rule 61.03.1(17)) and the subject is not raised in the appellant’s factum. Second, although the trial judge did consider the failure to deliver the Arnoff report to be relevant to the awarding of solicitor and client costs, it appears from his reasons that he would have awarded solicitor and client costs even in the absence of this consideration. He said in paragraph 174 of the decision on damages: “In any event, if the undertaking to pay damages is to have the effect of compensating the defendants for their losses, it is necessary that the defendants be reimbursed for the cost of the litigation.”
 This case at trial was largely fact-driven. The evidence which was called was crucial. Expert evidence was relevant and very important in assisting the trial judge to appreciate the significance of the facts and to arrive at the right conclusion. The trial judge made clear findings that he accepted the evidence of Mr. Duncombe and that of the defence expert, Mr. Sieler.
 The question of whether a substantial part of Sysview was reproduced by Assess is a matter of fact and degree (Ladbroke at p. 473). It is not for this court to embark on the question of substantial reproduction afresh (Designers Guild Ltd. v. Russell Williams (Textiles) Ltd. (Trading as Washington D.C.),  1 W.L.R. 2416 (H.L.) at 2418, particularly where the onus was on the appellant to prove substantial reproduction. No palpable or overriding error has been shown with respect to the trial judge’s factual conclusions: Hutton v. Canadian Broadcasting Corp.reflex, (1992), 41 C.P.R. (3d) 45 at 49-51 (Alta. C.A.).
 I would dismiss this appeal.
APPEAL FROM THE AWARD OF DAMAGES AGAINST CAROLIAN
 I turn now to Carolian’s appeal from the trial judge’s award of damages against it in favour of the defendants, under Carolian’s undertaking given under rule 44.03 at the time of the granting of an interlocutory injunction given in this proceeding by Fitzpatrick J. on February 3, 1987. The order granting the injunction included the following:
upon the Plaintiff undertaking to abide by any order of this Honourable Court with respect to damages sustained by the Defendants, and on hearing the submissions of counsel for the parties
1. THIS COURT ORDERS that the Plaintiff be and is hereby granted an interlocutory injunction restraining the Defendants, their servants, agents, assigns or any one or more of them, or anyone having knowledge of the within injunction, from selling, giving away or marketing Access/3000.
2. THIS COURT ORDERS that the Plaintiff be and is hereby granted an interlocutory injunction restraining the Defendants, their servants, agents, assigns or any one or more of them, or anyone having knowledge of the within injunction, from using in any way whatsoever or copying, selling or giving away computer programs or parts thereof, being the property of the Plaintiff.
 Fitzpatrick J.’s reasons, in endorsement form, read:
I do not find the explanation for the individual defendant’s having destroyed evidence persuasive. I am of the view that if the defendant is successful at trial after having been restrained, his chances of recovery on a judgment for damages are good, whereas I do not believe the plaintiff’s chances of recovering on a judgment are as good.
An order will go granting the reliefs sought in paragraphs (a) and (b) of the Notice of Motion.
 The trial judge conducted an 18-day hearing into the matter of the damages that the defendants incurred in consequence of the interlocutory injunction. He concluded that the defendant Duncombe’s damages were $130,500 and those of Triolet were $6,762,000. His reasons may be found at (1998), 69 O.T.C. I and  O.J. No. 2967. He awarded prejudgment interest on these amounts.
 Carolian raises three objections to the trial judge’s award:
(1) The amounts awarded were not foreseeable;
(2) They were not caused by the interlocutory injunction; and
(3) Their quantification is not supported by the evidence.
 There is no dispute in this appeal respecting the applicable legal principles. The damages to be awarded must be reasonably foreseeable at the time of the granting of the interlocutory injunction and they must be caused by (“naturally flow from”) the injunction and not by something else: International Pediatric Products Ltd. v. Lambert (1967), 66 D.L.R. (2d) 157 (B.C.C.A.) at 159; Hoffman-La Roche & Co. Attorney General et al. v. Secretary of State for Trade and Industry,  2 All E.R. 1128 (H.L.) at 1150; Douglass v. Butler (1913) 4 O.W.N. 1587 (H.C.) at 1590; Vieweger Construction Co. Ltd. v. Rush and Tompkins Construction Ltd., 1964 CanLII 53 (SCC),  S.C.R. 195 at 207; and 642947 Ontario Limited v. Fleisher 2001 CanLII 8623 (ON CA), (2001), 56 O.R. (3d) 417 (C.A.) at 434-436. The trial judge’s reasons make it clear that he appreciated that these were the applicable principles and that, in arriving at his conclusions, he applied them. What the appeal comes down to, then, is whether these conclusions are reasonably supportable on the evidence.
 Before addressing the grounds of appeal, I shall set forth the trial judge’s basic reasoning which led to his award. He accepted the expert evidence of one Derek Rostant who was called as a witness by the defendants. Mr. Rostant had expertise in accounting, quantification of loss resulting from business interruption, business evaluation, and viability of a business. The trial judge preferred Mr. Rostant’s evidence to that of Mr. Eric G. Rule, the expert called by Carolian. In general terms, the trial judge found that Mr. Rostant’s evidence, in light of the other evidence called during the trial on damages, was more realistic than Mr. Rule’s.
 Mr. Rostant calculated that the defendants’ loss resulting from the injunction of February 3, 1987 was $9,190,000, which was arrived at as follows:
|Lost earnings by Triolet Feb. 1/87 to Apr. 30/97||$3,516,000||Lost earnings by Duncombe Feb. 1/87 to Apr. 30/97||174,000|
|Loss in future value of Triolet at Apr. 30/97||5,500,000|
 Although the trial judge was of the view that Mr. Rostant “ha[d] been conservative in his calculations” (para. 163), he thought that his estimate should be reduced by 25%, $2,297,500, to take into account the uncertainties. The resulting award, before the addition of pre-judgment interest, was $6,892,000.
 I turn now to Carolian’s grounds of appeal. Carolian submits that the damages, apart from those relating to the lost sales of the Assess 3000 classic product, were not caused by the injunction and were not reasonably foreseeable. Because there is some practical overlap between these grounds, I shall deal with them together. Before I do so, I should mention that all of the submissions which Carolian made to us were made to the trial judge, in the form of various objections to the reasoning and conclusions of Mr. Rostant, and each was dealt with in considerable detail in the trial judge’s reasons.
 With respect to causation and foreseeability, Carolian submits that the trial judge should not have awarded damages relating to loss of sales of performance monitors in markets relating to two new generations of Hewlett Packard computers which followed the Hewlett Packard 3000 product – the HP 3000 XL and HP Unix markets. On this issue the trial judge relied upon evidence that Duncombe, in 1986, had a business plan (salient parts of which are quoted in paragraph 85 of his reasons) that included “porting” his current Assess program to the new systems and that the injunction, by shutting Duncombe out of the Assess classic market, also shut him out of the new markets. I quote from parts of the trial judge’s reasons on this issue:
 For a company with the proper expertise, porting the classic monitor to the XL computer could be done. But what all witnesses with knowledge of the market place seem to agree on is that unless a vendor had established itself with a reputable performance monitor in the HP 3000 classic market it could not likely have broken into the XL market. …
 In 1986 (but not in volume until 1990) Hewlett Packard, while not abandoning its XL computer, came out with a new Unix computer; thus competing for the Unix market with others such as I.B.M. Once again a vendor of a monitoring tool to the HP 3000 community wanting to keep its share of the Hewlett Packard software market or indeed to take advantage of the Unix market had to port its XL tool to the Unix operating system. This was less complicated than the adaptation from the classic to the XL had been. It was accomplished by Lund and could just as easily have been accomplished by Duncombe. Once again a presence in the old market with a reliable XL computer monitoring tool was essential. Perhaps even more devastating to Duncombe’s ability to enter either the XL market place or the Unix market place was the fact that he had been unable because of the injunction to stay in the classic market and by marketing first the classic monitor and then the XL to keep such tools up-to-date with the frequent changes made by Hewlett Packard in its classic and XL computers, and to alter the function of the monitoring tools to meet the needs and demands of computer users as their perceived needs changed. Putting it another way, by not being in the market place with performance monitoring tools Duncombe lost touch with the current needs in the market place. Indeed even his performance monitoring skills became rusty.
 One of the important assumptions made by Rostant in making the report is that Duncombe would have but for the injunction, expanded his business into other market areas just as several of the other companies that he had contacted had done. I agree with him. It was a natural progression to expand from the HP 3000 Classic to the HP 3000 XL market and then on to the HP 9000 (HP Unix) market place and then to other Unix market places.
 …. As I have already stated in these reasons if Duncombe had been in the market with a classic monitor he could have migrated his Assess first to XL and later to Unix. The injunction kept him out of the classic market; and so because he had no reputation with a classic monitor he was kept out of the XL and Unix markets.
 But the injunction itself prevented migration of Assess to XL and Unix computers. If Assess/3000 was enjoined because it was copied from Sysview, then to use Assess as the model for an XL or Unix monitor would violate the spirit of, if not the letter of the injunction. Eighty percent of Assess as created by Duncombe could have been used virtually unchanged to create an XL monitoring programme. But surely the XL programme so created would still be a copy of Sysview. A copy of the copy though working on a different computer would still be a copy of Sysview. The same applies to a Unix monitor.
 With respect to the foregoing, the following passage in Canada Metal Co. Ltd. v. Canadian Broadcasting Corporation (No.2) (1974), 4 O.R. (2d) 585 at 603 (H.C.) is relevant:
The following general principles relating to injunctions must be kept in mind:
The respondents were obliged to obey not only the letter, but also the spirit of the injunction: Grand Junction Canal Co. v. Dimes (1849), 17 Sim. 38, 60 E.R. 1041; Halsbury’s Laws of England, ibid., p. 434, para. 919; Attorney-General v. Great Northern R. Co. (1850), 4 De G. & Sm. 75, 64 E.R. 741.
 In support of a submission that the injunction did not keep Duncombe out of the Unix market, Carolian referred to evidence that in 1990 Duncombe became a shareholder in a company, Strategic Systems Inc., to develop performance software for the Unix systems but in 1991 dropped out of Strategic Systems Inc. and offered to sell his shares. This, Carolian submits, shows that it was not the injunction that kept Duncombe out of the Unix market. I think that the trial judge’s finding on this submission, that he accepted that Duncombe did not have the time or the money to continue in the Strategic Systems Group project, is reasonable. He said:
 Duncombe’s inability to continue with the Unix project is no reason for concluding that Duncombe running a thriving utility software business would fail to enter the Unix market. Rather, it is inconceivable that he would not enter it. To fail to stay abreast of changes in computer technology, would be to destroy his business.
In any event, this is just one piece of evidence bearing on the effect of the injunction.
 With respect to the defendants being kept out of the XL market, Carolian referred to an order of McRae J., dated December 11,1989, made on a motion by Carolian to have the defendants held in contempt of an earlier order which extended the injunction orders and for other relief. The order adjourned the contempt motion for trial and, also, provided:
THIS COURT ORDERS that in the meantime, until the trial of the said issue of contempt, to further exemplify the Order of the Honourable Mr. Justice McRae dated March 29, 1989,
(a) that Brian Duncombe not be involved with promotional material including letters, brochures and reader response cards which advertise both Probe 3000 and Probe 3000 XL;
(b) that Brian Duncombe tell any customers or prospective customers who enquire about Probe 3000 that he is not permitted to deal with Probe 3000;
(c) that Brian Duncombe keep records of all enquiries with respect to Probe 3000 and Probe 3000 XL;
(d) that Brian Duncombe keep copies of all correspondence, notes and instructions from Strategic Systems Inc.
 Carolian submits that this shows that the injunction order of Fitzpatrick J., as extended by subsequent orders, did not exclude the defendants from the HL market and that Mr. Duncombe knew this. There are several answers to this, not the least of which is that the wording of the quoted order is not at all clear on the point.
 The major answer is that the Probe 3000 XL product had been designed by a Mr. Apgood. Mr. Duncombe had nothing to do with this. Carolian’s allegation was that Mr. Duncombe had assisted Mr. Apgood to “transform down” the Probe 3000 XL product to create Probe 3000 using his knowledge of the Sysview 3000 product. On this evidence, it is understandable that the December 11, 1989 order would not, if such be the case, enjoin the defendants from entering the XL market.
 In any event, the foregoing does not indicate that the defendants were free to enter the XL market , porting its Assess 3000 Classic product to a new performance monitor.
 Further, no matter how the December 11, 1989 order is interpreted it does not afford an answer to the plain commercial fact, found by the trial judge, that being shut out of the Assess 3000 Classic market prevented entry into the XL and Unix markets.
 With respect to forseeability, there was evidence that both the XL and Unix computers were “on the cards” in 1986 and that this was known to Duncombe and, if it was not known to Carolian, it reasonably should have been. The trial judge said:
 In my view, the undertaking as to damages must relate to those damages that were reasonably foreseeable on March 29, 1989, and not just on February 3, 1987. I add, however, that in my view, Carolian should have foreseen on February 3, 1987 the injunction would keep Duncombe out of the XL and Unix markets as well as the Classic HP 3000 market. I will return to this point later.
 The trial judge made these findings:
 It is not possible that what was known and obvious to Duncombe was unknown to Carolian.
 Carolian then knew of Hewlett Packard’s plan in 1986 for what became known as the HP 3000 XL when the same was released by it in volume in 1988-89. It knew that any vendor of a performance monitor selling to the HP 3000 Classic market would need, in order to keep its share of the Hewlett Packard market, to port its monitor to the new computer. That is exactly what Carolian attempted to do when the XL computer was released. ….
 Porting or migrating a HP 3000 Classic monitor to the XL and Unix computers was not just a matter of choice for a performance monitor vendor. It was something that had to be done to stay in the performance monitor business.
 Carolian knew on February 3, 1987 that Duncombe, if not enjoined, would of necessity, port Assess 3000 to the XL computer and any other computers that Hewlett Packard introduced in numbers. Carolian also knew for the reasons I have already given that the injunction and the effects of the injunction would prevent him from entering the XL and Unix markets with Assess. The injunction not only meant the demise of Assess, it also killed what would have been the natural progression of Assess into the XL and Unix markets and this was clearly foreseeable. Loss from being kept out of the XL and Unix markets was the inevitable result of the injunction and flowed naturally, indeed necessarily from the injunction.
 The third objection to the assessment of damages is to the quantification of the defendants’ loss. This is entirely a matter of the exercise of judgment on the evidence called. There were many aspects of Mr. Rostant’s evidence that were challenged by Carolian and the trial judge gave consideration to each of these challenges (paragraphs 97 to 161). They were along the following lines:
(a) The share of the HP 3000 Classic and XL performance monitoring market and the number of sales of Assess to that market that Rostant projected Duncombe would have captured was unrealistically high.
(b) In calculating the projected revenue, Mr. Rostant assumed too high a selling price for Assess.
(c) Mr. Duncombe did not have the capabilities needed for the kind of business success projected by Mr. Rostant.
(d) Mr. Rostant’s projection as to earnings from maintenance revenues was too high.
(e) Any loss projected by Mr. Rostant because consulting and training fees were not earned should be ignored because the injunction did not prevent such fees being earned.
(f) Mr. Rostant’s projected earnings of Triolet’s pre-tax profit rate of 16% was excessively generous. It should be closer to 4.4%.
(g) It was inappropriate to undertake a future valuation calculation, even if damages were to be awarded for lost XL or Unix profits.
(h) The rate of return relating to the value of Triolet on April 30, 1997 should have been 18% rather than the 8.4% after tax rate of return used by Mr. Rostant.
 As I have said, the trial judge gave full reasons in his consideration of each of these objections. I will not set them forth beyond that part which relates to the second last mentioned objection, (g), respecting the inappropriateness of undertaking a future value calculation. It is:
 … It is likely but for the injunction, that Duncombe would have had on April 30, 1997 a thriving business that would be saleable, or that he could have continued to operate with the reasonable expectation that it was likely to continue to prosper. Duncombe should be compensated because he did not have that business.
 In my opinion, the trial judge’s reasons on Carolian’s objections are unassailable.
 If it be objected that, generally, in relying on the defendants’ evidence, the trial judge reached the limits of reasonableness in the realm of fact, this objection is met by the 25% reduction in the “conservative” calculation of Mr. Rostant.
 The standard of appellate review relating to an appeal from a damage award is set forth in Woelk v. Halvorson, 1980 CanLII 17 (SCC),  2 S.C.R. 430 at 435 as follows:
It is well settled that a Court of Appeal should not alter a damage award made at trial merely because, on its view of the evidence, it would have come to a different conclusion. It is only where a Court of Appeal concludes that there was no evidence upon which a trial judge could have reached this conclusion, or where he proceeded upon a mistaken or wrong principle, or where the result reached at the trial was wholly erroneous, that a Court of Appeal is entitled to intervene.
 In my view, Carolian has not satisfied any of these tests. I would dismiss this appeal.
 For the foregoing reasons, I would dismiss the appeal from the liability judgment and the appeal from the damage award, with costs - as in the trial judgment, on a substantial indemnity basis. Counsel may file written submissions on the amount of the costs.
“J.W. Morden J.A.”
“I agree J.J. Carthy J.A.”
“I agree J.C. MacPherson J.A.”
Released: March 1, 2002
This page has been generated using Mincov Legal Markup Language (“MLML”). While every effort has been made to ensure accuracy and currency of this document, Mincov Law Corporation makes no warranty, representation or undertaking, expressed or implied, nor does it assume any liability, direct or indirect, or responsibility for the accuracy, currency, usefulness or continued availability of this document.The official version of this document is located at http://canlii.ca/t/1dnrr.