Reportable
Case No 500/2000
In the matter between:
DEXION EUROPE LIMITED Appellant
and
UNIVERSAL STORAGE SYSTEMS
(PROPRIETARY) LIMITED Respondent
Coram: HARMS, SCHUTZ, SCOTT, CAMERON JJA and HEHER AJA
Heard: 26 AUGUST 2002
Delivered: 6 SEPTEMBER 2002
Subject: Copyright infringement; indirect copying of technical drawings.
JUDGMENT
HARMS JA/
HARMS JA:
[1] This judgment is about copyright infringement. The appellant (‘Dexion’) is the owner of the copyright in four technical drawings. They reflect different parts of Dexion’s Speedlock racking system. A racking system consists of uprights or legs bolted to footplates. The uprights are connected to each other by means of horizontal beams. Diagonal bracing elements between beams provide the necessary stability of the racks. Pallets or shelves can then be placed on the beams in order to store goods. The advantage of the Speedlock system is that it is an interlocking system and attaching the beams to the uprights does not require bolting or welding. At the extremities of each beam there are beam end connectors with a number of protruding hooks and these fit into coffin-shaped holes in the uprights during erection of the racks.
[2] The respondent (‘Universal’) produced the Speedlock system in terms of a sub-licence from Dexion’s licencee, to whom I shall refer as ‘Brazier’, until the termination of the sub-licence during 1993. Universal at the time sought a direct licence from Dexion and when it transpired that it would not get one, it decided on a new business strategy of producing a racking system that would be interchangeable and compatible with the Speedlock system. Legal advice was obtained and Universal was informed of the copyright implications and, particularly, the right to design a system by way of reverse engineering. Universal returned all drawings in its possession, instructed its regular toolmaker to manufacture new tools, thereafter returned the Dexion tooling to Brazier and proceeded to market its ‘new’ racking system as the Unirack system.
[3] Copyright protection is accorded to original artistic works (Copyright Act 98 of 1978 s 2(1)(e)). Drawings, including drawings of a technical nature, irrespective of their artistic quality, are considered to be artistic works (s 1 sv ‘artistic work’ and ‘drawing’). Copyright in an artistic work vests in the author the exclusive right of ‘reproducing the work in any manner or form’ (s 7(a)) and this includes the right to convert the drawing into a three-dimensional form (s 1 sv ‘reproduction’ par (b)). An exception exists in relation to this latter right:
‘The copyright in an artistic work of which three-dimensional reproductions were made available, whether inside or outside the Republic, to the public by or with the consent of the copyright owner (hereinafter referred to as authorized reproductions), shall not be infringed if any person without the consent of the owner makes or makes available to the public three-dimensional reproductions or adaptations of the authorized reproductions, provided the authorized reproductions primarily have a utilitarian purpose and are made by an industrial process.’
(S 15(3A)(a). The numbering of the proviso has been omitted for the ease of reading.)
[4] Universal did not dispute Dexion’s copyright in the drawings and the issue was consequently limited to that of copying. Copying is not a term used in the Act but it is understood that ‘reproduction’ requires copying. In order to establish copying, a two-stage inquiry is conducted. It has to be established whether there is the necessary degree of objective similarity between the original work and the alleged infringement; then it must be established that the similarity is causally connected to the original work. The causal connection can either be direct or indirect. See Galago Publishers (Pty) Ltd and Another v Erasmus 1989 (1) SA 276 (A) 280B-D.
[5] The effect of this, in general terms, is that copyright in a technical drawing is not infringed by a three-dimensional version of the drawing, which has no causal connection with the drawing. It is also not infringed if the version is reproduced from an authorised reproduction (reverse engineering). Even if the owner of the copyright made three-dimensional versions (as defined in the proviso) available to the public, third parties are not entitled to make three-dimensional copies by reference to the drawings. In other words, the drawings may not be used, directly or indirectly, to produce copyright-free three-dimensional ‘reproductions’. An example of such indirect copying would be the instance where the third party uses tools that were made from the drawings to make its version. The same would apply if the third party were to produce its own set of tools by copying tools produced from the copyright owner’s drawings. In this event the tools can be likened to negatives of photographs: making a photo from a negative infringes the copyright in the photo.
[6] Objective similarity in not seriously in dispute although, as counsel for Universal pointed out, the similarities are in part due to commonplace or generic elements and that there are some design differences between the two systems.
[7] I then turn to consider the causal connection. Dexion’s principal case as pleaded and presented in the court below in this regard was one of direct infringement: Universal had possession of its drawings and it used them in designing the Unirack system. Dexion ultimately had to concede that the evidence did not establish direct infringement and before us reliance was placed solely upon indirect infringement: Universal, so it was contended, used tooling derived from Dexion’s tooling, which in turn was derived from the Dexion drawings, thereby copying the drawings indirectly. No reliance was placed on subconscious copying, something raised in Francis Day & Hunter Ltd and Another v Bron and Another [1963] Ch 587 (CA). Universal’s principal case as pleaded was a reliance on reverse engineering under s 15(3A)(a) but its reliance on reverse engineering was too highly pitched and in the end it had to be content with a finding in the Court below (per Goldblatt J) that only the footplate is protected by the provision.
[8] It is convenient to deal with the footplate first. The only aspect of the exception in dispute is whether the footplate was reproduced from an authorised reproduction. Dexion mounted a muted attack on the finding of the Court below to the contrary. The undisputed evidence of Mr Jones, an employee of Universal, was that he handed a specimen of the Dexion footplate to a toolmaker with an instruction to manufacture tooling for making similar footplates. That was done. In any event, Dexion’s expert, Dr Nurick, testified that the information on the relevant drawing was not sufficient to enable one to make a footplate; a specimen of the Dexion footplate was required in order to do that. I am satisfied that there is no merit in the attack.
[9] Concerning indirect copying, Goldblatt J held that because it had not been pleaded, Dexion could not rely thereon. Since indirect copying is not a cause of action but simply a method of establishing copying, this finding cannot be upheld. The learned Judge further held that a reproduction of Dexion’s tooling could not have amounted to an infringement of copyright because the tooling would have been a three-dimensional reproduction of the authorised reproduction, ie, the original tooling. This finding is in conflict with the exposition of the law in the earlier part of this judgment. He also held that although tools are a means whereby a three-dimensional reproduction of a drawing can be produced, the tools do not reproduce the drawings. As mentioned, depending on the evidence, tools can be a ‘negative’ of the drawing and an intermediate element in the indirect copying of the drawing. Additionally, the finding fails to take account of the fact that the racking elements, and not the tooling, were made available to the public as required by s 15(3A)(a).
[10] In the light of these problems in the approach of the learned Judge, it is necessary to reconsider the evidence relating to indirect infringement. I will assume for purposes of this judgment, although there is no evidence to that effect, that the Dexion tools were derived from the copyright drawings in issue. Dexion relied on circumstantial evidence to establish copying of its tooling: Universal was in possession of some Dexion tools; when the sub-licence was cancelled, Universal refused to return the tools immediately and undertook to redeliver them once its new tooling was available; and Universal used the same toolmaker who had made the tooling for its Dexion product. These are material and weighty facts and we were referred to cases (Solar Thomson Engineering Co Ltd v Barton [1977] RPC 537 (CA) and Frank M Winstone (Merchants) Ltd v Plix Products Ltd 5 IPR 156, a judgment of the Court of Appeal of New Zealand) where indirect copying was found on ‘weaker’ facts. However, cases have to be decided with reference to their own facts and evidence of similarity, striking or otherwise, with or without evidence of access by the alleged infringer does not dispose of the case in favour of the copyright owner (cf Kambrook Distributing v Haz Products and Others (1986 WLD) per Kriegler J reported in Stranex Judgments on Copyright 243 277-278).
[11] The drawing relating to the bracing shows a standard lip channel with two holes at each end. How such a drawing can be said to be an original work is not entirely clear (cf Jacana Education (Pty) Ltd v Fransden Publishers (Pty) Ltd 1998 (2) SA 965 (SCA) 969C-G). In any event, Universal uses the same type of channel, cuts it into suitable lengths and punches a 10 mm hole at each end. Since the channel is generic, Universal simply purchases it and does not roll it. The only tool required is a 10 mm punch to form the hole. There is no evidence that Universal ever had a Dexion punch or that any tooling was made for producing these holes with reference to a Dexion tool. One would have thought that 10 mm punches are quite common in any engineering workshop.
[12] Another drawing relates to the beam connector. It shows in detail the form and dimensions of the hooks. The hooks made by Universal, although they are intended to and do perform the same function, differ somewhat from those shown in the drawing. In order to make the Dexion hooks, a punch and die are required. Universal uses a punch and pressure pad and no dies. Since there is no explanatory evidence on the matter, it is somewhat difficult to envisage how it can be said that the Universal tooling is a copy of the Dexion tooling. (I will revert to the evidence on how Universal’s replacement tooling was made.)
[13] The fourth drawing depicts the upright. The upright consists of a generic metal channel section, which, in plan view, has a bottle or wide-mouth flask shape. Universal used this section for the manufacture of another racking system prior to the conclusion of the sub-licence. It requested Dexion for drawings of its tools in order to roll this section (an indication that the tooling was not necessarily derived from the product drawings). The drawings supplied by Dexion for this purpose could not be used with Universal’s equipment and they were returned and Universal continued to use its existing tooling. The other aspect of the drawing is the coffin shaped holes. These clearly require a punch to be formed. The holes used by Universal, as noted by Dr Nurick, are somewhat differently shaped and can best be described as having a coffin-shape adapted to accommodate not only the Dexion type of hook but also a button-like hook.
[14] Jones was responsible for the design of the new tools and he gave the necessary instructions to the toolmaker (who has since died). It was not necessary to make new punches for everything because existing punches used for another product were available. He did not provide the toolmaker with any Dexion tooling to copy or use in the manufacture of the new tools. Instead he gave him his own sketches of how he expected the tooling to be made. Certain of the dimensions are well known and standard in the art. He worked backwards from those dimensions and through trial and error increased the size of the hook until it fitted. The tool drawing for punching the upright was one Jones had made personally some years earlier without Dexion’s tooling or drawings.
[15] Jones was not taken to task by the cross-examiner on this evidence because the cross-examiner was still attempting to establish direct copying and to show that reverse engineering had not taken place. The Court below did not make any adverse credibility finding against Jones and it was not argued that his evidence should have been rejected. It was never put or suggested to Jones that the Dexion tooling had been copied. It is manifestly unfair to argue a case on inferences from some facts and ignoring unchallenged direct evidence to the contrary. To accept Dexion’s argument would amount to an implicit rejection of Jones’s evidence even though the submission was not foreshadowed during the trial. The circumstances of this case do not justify a departure from the rule that the witness’s attention has to be drawn to the imputation which will be made during argument, a rule which has generally been adopted by our courts. See President of the Republic of South Africa and Others v South African Rugby Football Union and Others 2000 (1) SA 1 (CC) 36J-38C (par 61-65) and cases there quoted. I consequently agree with Goldblatt J that, since the issue was not properly dealt with in evidence and was not canvassed with Jones, Dexion cannot succeed on this ground.
[16] Dexion relied, in addition to its copyright claim, on a contractual claim. In reciting the facts I intend to continue to ignore the history of the different companies, their name changes, liquidations and compromises and to simplify the matter. Dexion licensed Brazier exclusively during 1971 to manufacture and sell the Speedlock system in South Africa. Brazier was entitled to appoint ‘contract manufacturers, distributors or agents’ subject to Dexion’s prior approval. Any such appointment had to contain for the benefit of Dexion two provisions, namely a right of quality control and the right to effect changes to the design, specifications or standards. On the valid termination of the licence, Brazier was obliged to ‘ensure that all its contract manufacturers distributors and agents’ cease to manufacture the licensed components.
[17] The sub-licence between Brazier and Universal likewise imposed a duty on Universal, on termination, to cease manufacturing the licensed components. Dexion alleges that it is the beneficiary of this obligation and that it is entitled to enforce it. Goldblatt J held that this clause in the sub-licence governed the relationship between Brazier and Universal and was inserted for the benefit of Brazier only. In addition, he held that there was no evidence that, if the agreement were one for the benefit of Dexion, Dexion had accepted the benefit. I agree with these findings. The position as between Dexion and Universal is similar to that between a landlord and a sub-tenant. Without a term (express or tacit) to the contrary, the landlord cannot rely on the term of the sub-lease in order to evict the sub-tenant but has to rely on ownership. Apart from this, I even have some reservations about Brazier’s ability to rely on the provision because, absent any protectable interest, such as copyright or confidential information, the clause may be no more than a bare covenant not to compete (Super Safes (Pty) Ltd and Others v Voulgarides and Others 1975 (2) SA 783 (W) 785D-F).
[18] The appeal is dismissed with costs.
____________________
L T C HARMS
JUDGE OF APPEAL
AGREE:
SCHUTZ JA
SCOTT JA
CAMERON JA
HEHER AJA