Navitaire Inc v easyJet Airline Company and Bulletproof Technologies Inc  EWHC 1725 (Ch)
Pumfrey J, 30 July 2004
In this judgment of the English High Court, Pumfrey J rejected a variety of copyright infringement claims based on non-textual copying of computer software.
The case concerned a ticketless airline booking system. Navitaire were the owners of the copyright in the source code of various works that made up their ‘OpenRes’ system. In November 1996, they had granted easyJet (the First Defendant) a licence to OpenRes. By 1999 relations between Navitaire and easyJet had deteriorated. easyJet decided to have a replacement system built by Bulletproof Technologies Inc (the Second Defendant), and to migrate the existing data (some 22 million or more records) onto it. The development contract required the new system to “have the same facilities/perform the same functions” as the OpenRes system. The new system (‘eRes’) went live in December 2001.
Navitaire alleged that in designing and building the new system, and in migrating data to it, easyJet had infringed Navitaire’s copyright in various aspects of the OpenRes system. It was common ground that neither Bulletproof nor easyJet had had access to the source code of OpenRes.
The principal aspects of OpenRes alleged to have been infringed were:
User keyboard commands (both individually and as a complete set of commands). The judge found (following Exxon Corp v Exxon Insurance Consultants International Ltd  RPC 69) that no copyright subsisted in simple commands; that the more complex commands were not recorded as such in the OpenRes source code, so were not protected as literary works; and that the command set as a whole had not been systematically designed with an overall structure, so that no compilation copyright subsisted in the set. He held in the alternative that the commands constituted a language, which in the judge’s view was excluded from copyright protection by the Software Directive (although that question would require determination by the European Court of Justice).
Screen layouts (both for character-based terminals and for GUI screens) and icons. The judge found that the templates for the character-based screens were tables, and thus literary works. However as static frameworks for displaying dynamic data they were ideas underlying the program’s interfaces, and therefore excluded from copyright protection by Art 1(2) of the Software Directive. The GUI screens and icons complained of were artistic works and had been infringed.
The ‘business logic’ of the OpenRes system. This was a claim of non-literal copying, since there was no dispute that the languages used, the actual code and (other than in respect of the database structure) the architecture of the systems were quite different. The business logic claim related to aspects of the system such as the relationship between the commands and the screens; the ability to carry out operations with much the same commands and screens with a successful result; and making the same data about all transactions available, and in substantially the same form as in OpenRes. The judge said that all of these had been copied. However, the question was whether any of this was protected by the copyright in the source code. The judge found not. He rejected the claimant’s suggested analogy with taking the plot of a novel.
It is important to note that in this case the claimant relied only on the source code as its copyright work. It effectively had to argue that a higher level of designed structure was implicit in the source code. In this the claimant was hindered by the judge’s findings of fact that there was no formal design process, that the OpenRes system ‘just grew’, and that user interface apart, the higher level features of the system were commonplace in airline reservation systems. This was effectively a finding that in this case there was no originality at that level, and therefore nothing protectable at that level that could have been taken by the defendants.
The OpenRes database structure. This was alleged to have been copied at a variety of different levels, ranging from reproduction of tables, schema defining the tables and table creation scripts, to reproduction at a higher level of the scheme of relationships between entities in the OpenRes database. Complaint was made both of alleged direct copying from the database, and of indirect copying via e.g. maintenance screens containing some information about the database structure.
The judge found on the facts that a few instances of direct copying of low-level aspects of the OpenRes database structure were infringements, but did not influence the design of the eRes database. Indirect reproduction (such as by studying the maintenance screens) was protected by S50D of the Copyright Designs and Patents Act 1988, since it was necessary for the purpose of finding out what data was present in the database. S50D provides that it is not an infringement of copyright in a database for a person who has a right to use it or any part of it to do, in the exercise of that right, anything which is ‘necessary for the purpose of access to and use of the contents of the database or of that part of the database’. The judge held that necessity excluded the merely desirable or convenient, but did not require the merely absurd.
As to the claim for infringement of the overall structure of the database, the judge found that that there was no entity relationship design expressed in the OpenRes database schemas, and that it would have taken considerable detective work to try to deduce one by studying the database. The judge appears to have concluded both that the OpenRes source code did not contain a protectable work at this level, and that as a matter of fact any overall relationships between entities were not copied. These findings were influenced by the fact that, since OpenRes was not built using a relational database, the relationships between entities were hard-coded in the underlying COBOL code rather than being set out in an explicit entity relationship diagram as part of the design process.
That for the purpose of data migration, various infringing interim copies of the OpenRes database structure were made. The judge found these either to be indirect copying for the purpose of migrating data and protected by S50D of the 1988 Act, or to be infringements that caused no loss.
This case was based on copyright in source code. The absence of design documents for OpenRes is significant. When considering the ‘business logic’ claim the judge commented that copying at the non-textual level avoids the need to conduct any systems analysis or to produce functional specifications. In this case there was no functional specification for OpenRes corresponding to any labour saved by copying.
The case therefore leaves open the question whether a case founded on non-literal high level copying could succeed if the claimant were able to produce and rely upon underlying design documents, such as a functional specification. Such documents are literary works and are also given specific copyright protection as computer programs under the Software Directive.
Traditional approaches to indirect copying would suggest that a program arrived at by copying another program at a higher level would be capable of infringing an underlying functional specification, if a substantial part of the skill and labour that went into that design document had been taken.
However, Navitaire suggests that Pumfrey J might be hostile even to that approach. He stated that he did not feel that the courts should be astute to extend software copyright protection into a region where only the functional effects of a program are in issue. He also used an illustrative analogy suggesting he would not regard indirect copying of an underlying work by use of information obtained by studying the program as an infringement of copyright in the underlying work.
The judge’s broader statements about copying at a higher level of abstraction appear to have been influenced both by common software industry practice and by the perceived policy underlying the Software Directive. It will be interesting to see how this subject develops in future cases. Navitaire is unlikely to be the last word on non-literal copying of computer software.
Also published in WIPR.