Sas Institute Inc. v World Programming Ltd

Introduction

The underlying issue in this case is the extent to which the developer of a computer program may lawfully replicate the functions of an existing computer program; and the materials that he may lawfully use for that purpose. The rival programs in our case are both sophisticated pieces of business software.

The SAS System is an integrated set of analytical software programs which enables users to carry out a wide range of data processing and analysis tasks, and in particular statistical analysis, developed by SAS Institute. SAS Institute publishes numerous detailed technical manuals for the SAS System. The core component of the SAS System is Base SAS, which enables users to write and run application programs to manipulate data. These applications are written in a language called the SAS Language. The functionality of Base SAS may be extended by the use of additional components. In addition, SAS Institute produces at lower cost a cut-down version of the SAS System software called the SAS Learning Edition. As its name suggests it is intended to educate users in the functions of the full system.

WPL is a competitor of SAS Institute. It perceived that there would be a market demand for alternative software which would be able to execute application programs written in the SAS Language. The idea was that its alternative software would produce the same outputs as the SAS Components in response to the same inputs. WPL therefore created a product called World Programming System ("WPS") to do this. In developing WPS, WPL sought to emulate much of the functionality of the SAS Components as closely as possible. This was so as to ensure that WPL's customers' application programs executed in the same manner when run on WPS as on the SAS Components. SAS Institute contends that WPL has both committed a series of infringements of copyright and acted in breach of contract in creating WPS and its accompanying documentation.

In the court below SAS Institute alleged that in creating WPS:

i) WPL had used the SAS Manuals as a technical specification for WPS and copied a substantial part of those manuals in creating WPS itself, thereby infringing copyright in the SAS Manuals (the "Manual to Program Claim");

ii) WPL had indirectly infringed copyright in the SAS Components in creating WPS (the "Program to Program Claim");

iii) WPL had infringed the copyright in the SAS Manuals by reproducing a substantial part of them in WPL's own WPS Manual and WPS Guide (the "Manual to Manual Claim"); and

iv) WPL had repeatedly used the SAS Learning Edition outside the scope of the applicable licence to obtain additional information about the SAS System, and to check that the operations of WPS precisely replicated those of the SAS Components; it had thereby infringed SAS Institute's copyright in the SAS Learning Edition and acted in breach of contract (the "Learning Edition Claim").

Arnold J tried the action in June 2010, and gave judgment on 23 July 2010. His judgment is at [2010] EWHC 1829 (Ch) [2011] RPC 1 where the facts are rehearsed in very great detail. Despite a detailed and exhaustive discussion of the law, the judge was unable finally to dispose of the claims without guidance from the Court of Justice of the European Union; and referred a series of questions to that court. His second judgment dealt with the form of the questions, and need not concern us. Following an opinion from Advocate-General Bot the court answered those questions (or at least its paraphrase of them) on 2 May 2012. Its judgment is at Case C-406/10 [2012] RPC 31, where the Advocate-General's opinion is also reproduced. The case then returned to Arnold J and he gave a third judgment on 25 January 2013. That judgment is at [2013] EWHC 69 (Ch) [2013] RPC 17. Unfortunately the parties could not agree what the CJEU had actually decided. The language in which the court expressed its judgment was, at times, disappointingly compressed, if not obscure. Moreover, although the judge had referred specific and detailed questions to the CJEU, the CJEU refrained from answering them, but instead answered its own paraphrase. This led to a disagreement about whether the court had actually given answers to all the questions posed. It would, perhaps, be more helpful if in response to a national court asking for help the CJEU, in the performance of its duty of sincere co-operation, answered the questions it was asked unless there are cogent reasons not to.

Following his third judgment the judge dismissed the claims of SAS Institute, except that he found limited breaches of copyright in relation to the Manual to Manual Claim. The judge gave permission to appeal on the Learning Edition Claim, and I gave permission to appeal on Manual to Program Claim, and the remainder of the Manual to Manual Claim. There is no appeal against the dismissal of the Program to Program Claim.

Mr Geoffrey Hobbs QC, Mr Michael Hicks and Mr Guy Hollingworth presented SAS Institute's appeal. Mr Martin Howe QC and Mr Robert Onslow presented WPL's response. I would like at the outset to express my appreciation of the excellence of the arguments, both written and oral, which made the task of hearing the appeal a real pleasure.

Although I disagree with some of the judge's reasoning, those disagreements do not affect the ultimate result. Since the appeal is an appeal against the judge's order rather than against his reasons, I would dismiss the appeal. My reasons follow.

Procedural matters

Some of Mr Hobbs' criticisms of the judge tie in with the way in which the case was managed, prepared for and presented at trial. It is therefore necessary to refer to the case management order which I made when the case came before me, sitting at first instance, in April 2010. The order was made in the context of a tight timetable to trial which was thought necessary in view of the commercial imperatives. By paragraph 3 (1) of that order I limited the number of similarities between the SAS Manuals and the WPS Manual which SAS Institute were permitted to put forward at trial. But paragraph 3 (2) of the order stated:

"such similarities shall be taken to be a representative sample of the similarities present between the SAS Manuals and the WPL Manual as a whole."

It is clear from the transcript of the hearing that the underlying purpose of these orders was that whatever proportion of copying was found in the representative samples could be extrapolated across the manuals as a whole.

I made a similar order in relation to the number of examples of similarities between the WPS software and the SAS Manuals and SAS Components; and again the order provided that:

"such specific similarities shall be taken to be a representative sample of the similarities present between the SAS Manuals and the SAS Software Components and WPS as a whole."

Paragraph 8 of the order stated that:

"The case shall proceed on the basis that insofar as [WPL] has used SAS Manuals for the purposes of developing WPS, it shall be treated as having worked from versions of the manuals in respect of which copyright has been admitted to subsist."

The SAS System and its creation

There is no dispute that each of the SAS Components is an original computer program in which copyright subsists. Nor is there any dispute that the creation of each of the SAS Components involved very great intellectual effort on the part of SAS Institute's employees. Nor is there any dispute that the intellectual effort involved in creating each of the SAS Components included both (i) intellectual effort in determining the requirements for the software and (ii) intellectual effort in designing and writing source code to implement those requirements. The first of these will have involved making choices about what the SAS components would do. Equally, however, it is common ground that WPL did not copy the program directly, because it had no access to the source code or the object code. As the judge found, WPS is not written in the SAS Language. It was first written in the Java programming language, and has subsequently been translated into C++.

The judge made the following additional findings about the SAS System:

i) Many of SAS Institute's developers have expertise in statistics as well as in computer software.

ii) Although many of the statistical analyses and computations performed by the SAS System are based on methods published by others in the academic statistics literature, some have been devised by SAS Institute's own employees.

iii) The SAS System consists of DATA steps and PROC steps. A PROC step invokes a PROC (of which there are more than 3,000) which processes and analyses data in SAS data sets to produce statistics, tables, reports, charts, plots and so on.

iv) Each PROC step represents a significant body of work in its own right, including research into the best approach to statistical analysis. Choices need to be made as to what features to introduce, what syntax to use, what statements and options are appropriate for a particular procedure, and what output should...