Anton Haupt (appellant) and Christopher Brewer (second respondent) were brothers-in-law who worked as marketing director and managing director respectively of Brewer's Almanac (Pty) Ltd. In 1998, Brewer's Almanac commissioned Byron Coetzee (third respondent) to write a computer program called Project AMPS to interrogate AMPS (All Media Products Survey) data. Coetzee created the program using Delphi, and Haupt filled database tables with questions from AMPS questionnaires. By 21 June 1998 the program was functional but produced sometimes correct, sometimes incorrect results. Haupt and Brewer parted ways on 31 July 1998. Thereafter, Coetzee continued developing the program exclusively for Haupt for 20% of gross sales. Major developments after July 1998 included: tree preparer program, converter program, search function, and graphics server incorporation. The program was renamed Data Explorer. In 2001, Coetzee secretly negotiated with Brewer to provide compiled data and assistance for developing a competing program (Brewer's AMPS program) through first respondent (Brewers Marketing Intelligence). Coetzee supplied source code, databases, and tree.txt files to the first respondent. When Haupt discovered the competing program, he applied for an Anton Piller order and sought an interdict for copyright infringement.
The appeal was upheld with costs including costs of two counsel. The High Court order was set aside and replaced with: (1) An interdict preventing respondents from infringing Haupt's copyright in the Data Explorer program; (2) An order requiring respondents to deliver up all infringing copies within 7 days; (3) Respondents ordered jointly and severally to pay 50% of costs of suit (reduced from full costs due to all parties' responsibility for inefficient trial process), including qualifying fees, Anton Piller application costs, and costs of the application on 11 September 2002. The appeal against costs orders of two interlocutory applications (11 September and 11 November 2002) was dismissed as the court found no irregularity or misdirection in those discretionary costs orders.
The binding legal principles established are: (1) A computer program under the Copyright Act 98 of 1978 is defined as instructions that direct computer operations to bring about 'a result' - correctness of the result is not required for copyright eligibility. (2) Improvements or refinements to existing copyrighted works attract independent copyright protection if they are original and involve substantial (not superficial) material alteration or embellishment. (3) Database structures consisting of tables with columns are literary works, not computer programs, as they do not constitute 'instructions' directing computer operations. (4) A work is 'computer-generated' only when created by a computer in circumstances where there is no human author; if a human author uses a computer as a tool (computer-assisted work), it is not computer-generated. (5) 'Original' in South African copyright law does not require creativity; a work is original if: (a) it has not been copied from an existing source, and (b) its production required a substantial (not trivial) degree of skill, judgment or labour. (6) 'Control' over making of a computer program (determining authorship) has a wider meaning than control via employment contract - it includes situations where a person exercises restraint or direction over the free action of the creator, including through detailed instructions, approval processes, and authority to direct or terminate development, even in independent contractor relationships. (7) Infringement of copyright by reproduction of a 'substantial part' depends on the qualitative importance of what was taken, not merely quantitative measurement; copying of difficult, valuable components constitutes substantial reproduction even if numerically small. (8) A person who assists copyright infringement by providing essential materials (such as source code) and cooperates with infringers for commercial benefit (royalties) themselves infringes copyright by making common cause with the primary infringers.
Streicher JA made important observations about procedural matters: The court strongly criticized the manner in which the application was referred to trial, noting that Uniform Rule 6(5)(g) requires appropriate directions as to pleadings or definition of issues when referring matters to trial. The failure to define issues properly resulted in a grossly inefficient 33-day trial addressing irrelevant matters including: shareholding in Brewer's Almanac, whether Haupt and Brewer's relationship was similar to a partnership, and who first conceived the idea for the program. The court observed that when a matter is referred to trial on affidavits (which in this case comprised 330 pages with poorly defined issues), a pre-trial conference is essential, and suggested the appropriate order would have been to direct that the notice of motion stand as a simple summons, require a declaration within fixed time, and apply Uniform Rules on pleadings and trial conduct. The court noted that although the referral order was made by consent, the trial judge should have insisted on proper issue definition when the matter was allocated to him. Regarding costs, the court exercised discretion to deprive Haupt of 50% of his costs despite his ultimate success, because all parties bore responsibility for the inefficient trial process resulting from failure to define issues. The court also noted (without deciding the point) the question of whether South Africa should follow the Canadian approach rejecting 'sweat of the brow' as sufficient for originality, observing that Canadian law differs from UK, South African and Australian approaches which accept that application of skill, judgment or labour (even without creativity) satisfies originality requirements.
This case is a leading South African authority on copyright protection for computer software and databases. It established important principles regarding: (1) the low threshold for what constitutes a functional 'computer program' (need not work correctly); (2) the distinction between computer programs and database structures/literary works; (3) the meaning of 'computer-generated' works (requiring absence of human author); (4) the rejection of creativity as a requirement for originality in South African copyright law (following UK approach rather than US 'minimal creativity' standard from Feist); (5) clarification that originality requires only substantial skill, judgment or labour, not creativity; (6) interpretation of 'control' over making of computer programs as broader than employment relationships, encompassing situations where one party exercises authority and direction over an independent contractor; (7) the principle that substantiality of reproduction depends on qualitative rather than quantitative assessment. The judgment also serves as a cautionary tale about proper case management when referring applications to trial, emphasizing the necessity of defining issues clearly to avoid protracted and costly proceedings.
Explore 1 related case • Click to navigate