Judgment in the case of Navitaire Inc v easyJet Airline Company and Bulletproof Technologies Inc [2004] EWHC 1725 (Ch) was handed down by Mr Justice Pumfrey on
The action was brought by Navitaire Inc for infringement of its copyright in a computer system called OpenRes. OpenRes is an airline reservation system for airlines who employ ‘ticketless’ booking and indeed is the dominant application in that sector. easyJet was one of the licensees of that software, going live with OpenRes in June 1997. Feeling that the software was failing to serve its needs as it grew, easyJet decided to write its own replacement system, eRes, and commissioned Bulletproof Technologies (a Californian software developer) to collaborate in that writing.
The result was an action by Navitaire for infringement of copyright. The decision is interesting for a number of reasons. It concerns, amongst other things, an attempt to bring a claim for ‘non-textual copying’ in relation to computer software – although here the claimant failed. However, not all the judgment will be unwelcome to those involved in software development. Perhaps equally important, although again on the facts the claimant failed, the judgment does confirm that the structure of a database (as distinct from the actual data populating that database) is protectable by copyright.
The claims of infringement
At no stage did easyJet (or Bulletproof) ever have access to the source code of the application. Navitaire alleged (and this was not disputed by the defendants) that easyJet wanted a new system ‘substantially indistinguishable from the OpenRes system, as easyJet used it, in respect of its ‘user interface”. The rationale being that the staff at easyJet were used to a particular functionality and that functionality had to be replicated as much as possible. Put simply, easyJet deliberately intended with eRes to offer to its users an experience similar to the experience offered by OpenRes (in a sense, to ‘copy’ that experience) but of course denied that what was copied was a work protected by copyright.
There were a number of overlapping allegations of infringement in the OpenRes system, mentioned in turn below.
Commands
The interface of OpenRes (and eRes) used by easyJet staff was a command style interface. For example, the command ‘A’ with appropriate values as parameters is used in both systems to check availability of a particular flight; so that, ‘A13JunLtnAms’ would be the command to check availability on 13th June from
Individual and Complex commands
Pumfrey J decided that copyright did not subsist in the individual command names because they did not have the necessary qualities of a literary work (based on the well-known Exxon judgment). He also held that copyright did not subsist in the syntax of the complex commands. This was because the syntax of these complex commands could not be seen as having been ‘recorded’ – which is a fundamental requirement for the subsistence of copyright in literary works (Copyright, Designs and Patents Act 1988, s 3(2)). In OpenRes the commands were recorded only ‘in the sense that it is possible to analyse the code to ascertain that a machine operating according to that code will ‘recognise’ the command A13JUNLTNAMS as requiring the display of available seats on 13 June between
Pumfrey J recognised that this position may not be entirely satisfactory because it is possible to create a parser which would make the commands and syntax recognisable. However, this would depend on the way in which the parser was constructed and Pumfrey J thought that it would be inappropriate to afford protection on this basis.
Instead he approached the problem by deciding that defining a series of commands and their syntax in actual fact defined a computer language. As the relevant law (Software Directive) did not extend to the protection of computer languages, no copyright existed. Pumfrey J went on to state that, although he thought the conclusion was correct, the extent of the exclusion of computer languages from protection was unclear and should be referred to the ECJ.
Compilation of Commands
Pumfrey J came to the same conclusion for the claim regarding the set of commands as a compilation, similarly classifying them as a computer language and thus being incapable of protection. Further, he went on to give a second reason that the set of commands could not, in his view, be properly categorised as a ‘compilation’. The totality of commands were not designed from scratch: all that had happened was that new individual commands were developed and added in perhaps an ad-hoc fashion. It was, the judge said, an accretion of commands which lacked overall design which he thought necessary for the commands to amount to a compilation.
The Screens and Report
There were two types of screens subject to claims in the case: those for a character-based terminal, the VT100, and a Graphical User Interface (GUI). Pumfrey J considered these screens separately because he concluded that they qualified for different types of copyright protection.
VT100
The following is an example of the VT100 screen reproduced in the appendices of the judgment.
This screen was used as a ‘good illustration’ that the VT100 screens should properly be viewed as character based because it is possible to see the layouts in the code. This led Pumfrey J to infer that they could ‘properly be viewed as tables and so literary in character for the purpose of copyright’. His conclusion was that Article 1(2) of the Software Directive (91/250/EEC of
GUI
In relation to the GUI screens, Pumfrey J concluded that the Software Directive was inapplicable. He asserted that the Directive was solely concerned with computer programs which could be viewed as literary works and had no impact on those which had the character of artistic works. Pumfrey J found that the GUI screens satisfied the requirements for artistic copyright, namely that arranging the screens involved the exercise of sufficient skill and labour for the result to amount to an artistic work. Navitaire therefore succeeded on this part of the claim.
Reports
There was also a claim of copyright infringement in certain ‘reports’ which were representations of the data in the underlying database and in various screenshots. The claims in relation to the reports wholly failed: partly on the basis that there was no copying of the substantial parts and partly on the basis that the representations were dictated by the data itself. The data that appeared in the reports was displayed and manipulated in completely different ways in the two systems. The only part that could be seen to be copied was the layouts and these were so trivial that the copying could not be considered substantial in any way.
Non textual copying/ Business logic
The alleged infringements discussed so far illustrate a classic case of copyright infringement. However, of more interest, and what Pumfrey J described as the ‘striking feature’ of this action, was whether there was ‘something else that has been copied’ over and above the limited features of the commands and screens and whether that something could be protected by copyright. The claim was put on the basis that the ‘business logic’ was copied. This is not a term of art and has no precise definition. In addition, use of this term in the context of computer programs is not without difficulty because it can be seen to assume that the programming logic somehow reflects business logic which is not necessarily true.
The main thrust of this claim was Navitaire’s contention that, if a machine in operation is studied, it becomes possible to identify the machine’s response to all possible sequences of inputs, and so construct a new machine that operates to give the same outputs for the same sequences of inputs by writing an appropriate program. It follows axiomatically that any copyright in the source code for the first machine must be infringed in writing the second program. Therefore, according to Navitaire, no access to the OpenRes source code would be necessary to constitute infringement by easyJet or Bulletproof and it was not an issue that the OpenRes and eRes systems were completely different in language, structure and architecture.
This concept is dependent on the fact that the manner in which a machine works under the control of OpenRes is a part of the skill and labour that went into the program. Pumfrey J agreed with this sentiment. Although copying of the business logic meant there was still the need to write the actual source code, it avoided ‘the need to identify the result by any of the normal methods of analysis that either precede or accompany the writing of a substantial piece of business software’.
This raised the issue of whether the structure and functionality of the code had been copied; whether there was non-textual copying of what is sometimes called ‘look and feel’. The main difficulty with such a concept is finding the relevant level of abstraction that represents the skill and labour of the designers and programmers of OpenRes going beyond the limited commands and screens, and which is not merely inherent in the business functions required by a flight booking system (ie check flights – availability – reserve and so on). In turn, this raised the issue of whether what Navitaire was trying to protect was on the wrong side of the idea/expression dichotomy and whether the copying was substantial. In explaining these points Pumfrey J referred to a number of seminal cases.
First, he quoted Lord Hoffman’s well known passage from Designers Guild v Russell Williams Textiles Ltd [2000] 1 WLR 2416 which explained the difference between copying ideas and expression. Indeed the copying need not be of a discrete part of the work but can be a ‘feature or combination of features of the work, abstracted from it’.
Pumfrey J went on to reject the conclusion that this encapsulated the idea of non-textual copying. The judge instead defined non textual copying in the case as ‘copying without access to the thing copied, directly or indirectly’. Therefore simply reproducing the copyright work in a different form was not truly ‘non textual copying’ especially when the copyist did not have access to the original work.
Pumfrey J then discussed the case of John Richardson Computers v Flanders [1993] FSR 497 where the copyist did not have access to the original work but was still found to infringe (albeit on limited grounds). Although there was no deliberate copying, the copyist in that case had such an intimate knowledge of the previous program that he may have made use of it unconsciously. Navitaire argued that the decision in
‘where an idea is sufficiently general, that even if an 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 same applies whether the work is functional or not, and whether visual or literary’.
In Ibcos, Jacob J likened the problem to that of protecting the plot of a novel. Here, Pumfrey J rejected that as an inadequate analogy. He said it was wrong to think of a computer program in that way. Instead a computer program is a series of pre-defined operations intended to achieve the desired result in response to the requests of the customer. A more pertinent analogy was that of a chef who invents a new pudding, the written recipe of which is a literary work. If a competitor, after much culinary labour, manages to emulate the earlier result, and he records his own recipe, that latter recipe cannot be seen as an infringement of the earlier work.
Here, Pumfrey J thought that the passage from Ibcos quoted above did not answer the question because what he was presented with was a problem peculiar to computer programs – ie that two completely different computer programs can produce an identical result: not a result identical at some level of abstraction but identical at any level of abstraction This is so even if the author of one has no access at all to the other but only to its results. Instead he found the answer in Lord Hoffman’s Designers Guild decision, which concluded that the relevant level of abstraction is very important in deciding whether a work should attract copyright protection. When the copied idea is abstract and simple it is less likely to qualify as a substantial part of the work. Instead ‘originality, in the sense of the contribution of the author’s skill and labour, tends to lie in the detail with which the basic idea is presented’. Therefore the copied elements must go beyond the banal to qualify for protection.
Pumfrey J found that, because no code had been copied, what is left when the interface aspects of the case are disregarded is the business function of carrying out the transaction and creating the record of reservations and, whilst it is right that those responsible for devising OpenRes envisaged this as the end result for their program, that does not amount to relevant skill and labour. Therefore the claim for non-textual copying was rejected without regret by Pumfrey J, who went on to state that he believed this approach to be compliant with the policy of the Software Directive. He considered any extension of copyright protection to the matters argued by Navitaire would be inappropriate and unjustified.
Claims in relation to Takeflight
Navitaire also made a claim of ‘non textual’ copying in relation to the Takeflight system, which was the Web interface of OpenRes used by consumers. This included a claim that easyJet had infringed the licence by unauthorised copying of Takeflight. The claims in relation to non-textual copying were also rejected by Pumfrey J. He decided that, even though the systems both used the same flight booking procedure, what was taken was not substantial. However, Navitaire was entitled to relief as regards the part of the claim relating to unauthorised alterations to Takeflight. The licence had been breached by the copying of the code for bug fixes, foreign language versions, adding and removing promotions, adding a new look and feel in the form of a new ‘skin’ and using the program for easyJet Tours.
Claims in relation to the database
A substantial part of the judgment deals with the technically complex area of infringement of the OpenRes databases. Like many complex systems, much development effort went into the design of the underlying database. The structure of the OpenRes database was set out in certain ‘schema’ and the claim was founded in copyright in that structure as described in that schema. The schema did have copyright protection. The claim that the copyright in the schema was infringed was broadly in two parts.
First, there was a claim that easyJet, in creating its new system, infringed rights in the schema. This failed on the basis that easyJet did not have access to the source code of the OpenRes database. Without this access it would have been impossible to extract information as to the logical structure with confidence. In addition, the relationship between the different datasets in the schema remained in the mind of the programmers, therefore it is not possible to draw an entity relationship diagram between all the relevant datasets. The task of copying would require extraneous information and that would thus be far from straightforward. On the facts, therefore, there was no copying.
Secondly, it was claimed that easyJet had breached copyright by making copies of the database schema as part of the migration exercise. As is standard when replacing IT systems, easyJet had to move (migrate) data from its old OpenRes system to its new eRes system. A distinction arises between rights in the database structure (found as just mentioned to subsist and be owned by Navitaire) and rights in the data itself. (The judgment does not address the latter. Any rights in the data would presumably have been easyJet ‘s and likely to be protected by database right.) In order to manage the migration, easyJet copied the database in a number of ways; providing Bulletproof with a tape of the complete database; providing direct links into the OpenRes database; providing screen shots of the database and providing a spreadsheet containing information on the nature of the data and where the data was stored within the database.
Pumfrey J considered the effect of s 50D of the CDPA (which sets out certain acts permitted in relation to databases) and whether that would allow the provision of the database on the tape. This allows ‘a person who has the right to use the database 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 that part of the database’ (emphasis added). He found that the provision of the tape of the database to Bulletproof was an infringement because a neutral file format could have been used, that is, where the database developer tells the user what data to extract and provides a file format of unnamed records separated by separators. This use was not ‘necessary’, albeit it could have been more convenient or desirable. It was also decided that the provision of the direct links was an infringement because this would allow Bulletproof to access information about the database fields even when there was no access to the data. However the provision of the screen shots was not an infringement because this was considered to be a usual method to ascertain which data was held in the database by both experts and relevant textbooks on the issue. He found that the use of screenshots to ascertain which data was stored in the database was within s 50D, if screen shots were not permitted then it would be difficult to ascertain that data held within the database, which easyJet had a right to see. In relation to the spreadsheet there was no infringement because a substantial part of the work was not taken, easyJet were entitled to tell Bulletproof which data items existed in the database.
Comment
Look and Feel
The most important aspect of this case is in relation to the lack of protection for the business logic of the OpenRes application and this is a result that some software developers may view with surprise. The judge applied the recent House of Lords decision in Designer’s Guild v Russell Williams Textiles Ltd [2000] UKHL 58, a decision on copyright in an artistic work, and the analysis there of the idea/expression dichotomy. In applying the test set out by Lord Hoffman in that case, Pumfrey J puts much reliance on the fact that easyJet, unlike the defendants in Ibcos and Designer’s Guild, did not have available to it the work copied (at least in source code form). All that was available to easyJet were basic input screens and output screens (and the result of easyJet ‘s observations). As he says ‘what is left when the interface aspects of the case are disregarded is the business function of carrying out the transaction and creating the record’. That was not relevant skill and labour. Of course, that is arguably so with many software applications. Is the judge then really saying that unless source code, or at least some preparatory document such as a functional or technical specification, were available to the defendant it will be hard, if not impossible, to bring a claim for breach of copyright (when the screens and functionality are simple)? Probably not, and that result does seem a conclusion too far. Inevitably, questions of degree will come up (as in Designer’s Guild).
Command systems as a computer language.
There are many other parts of this case which invite further discussion. The denial of copyright in relation to complex commands on the ground that the commands were not ‘recorded’ is rather interesting. The judge does not adequately seem to explain why recording in this sense is insufficient for protection of the syntax of the commands, but – although not in this case – sufficient for those cases where ‘look and feel‘ is given protection by the courts. In those cases too the look and feel could be said to be recorded only ‘in the sense that it is possible to analyse the code to ascertain that a machine operating according to that code will recognise the command’. He acknowledged in fact that in this sense the syntax has recorded. However, a higher threshold of recording was needed. The syntax needed to be ‘stated’. But again taken further this would then seem to apply to any look and feel case.
The judge found that the command syntaxes were a ‘computer language’. He said that this was exactly the same ‘as to find a language such as BASIC or a simple language to control a calculated programme’. Accordingly, the complex commands were not protectable because he thought the Software Directive expressly excluded computer languages. Recital 14 of the Software Directive states that ‘to the extent that logic, algorithms and programming languages comprise ideas and principles those ideas and principles are not protected under this Directive’. Pumfrey J did acknowledge that there was still some doubt as to the extent to which that is always in fact an exclusion of computer languages from the scope of protection (the words ‘to the extent’ are used within the recital, although he did not focus on that), but of more interest is perhaps the suggestion that the command syntax was a language in the first place.
Most software engineers would understand a computer language to be the means by which other computer programs could be written. It would allow for some flexibility. It is not immediately clear that creating commands with a specific syntax for a specific purpose (albeit of course different commands would be capable of being imputed into the system) is in fact the invention of such a language. All computers need inputs, otherwise they have nothing on which to compute. Sophisticated and complex computer systems often invite interaction with other programs by the publication of interface means (perhaps an application programming interface or API). Any such API would have a syntax and much effort is put towards creating APIs which are clear and easy to use, and that export should surely be sufficient to attract copyright protection for the API (but that is perhaps for another case!). If this is right it is hard to see that there is any meaningful distinction between the command syntax here (the input mechanism for the OpenRes system) and these more sophisticated systems.
Compilation of Commands
The denial of protection to the compilation of commands is perhaps understandable, but the reasoning merits further exploration. The judge found there not to be a ‘compilation’ in the sense used in the CDPA. As this term is not defined in the CDPA, courts tend to look to dictionary definitions: here, to compile is ‘collecting material into a list’ or ‘accumulating’. But this definition does not seem to need the idea of a compilation needing to have an overall design, the grounds on which protection was denied here. Whether there was an overall design only seems relevant to the question of whether the compilation was ‘original’ and therefore entitled to copyright protection. Perhaps then a perceived lack of originality was what drove him to what must be the correct conclusion rather than simply the organic growth of the table.
Screens
The judge’s reasoning in relation to the application of the Software Directive to the copyright existing in the screens is also open to question. It seems that Pumfrey J has not distinguished whether the screens should be protected because copying of the screens infringes the copyright in the underlying computer program or whether the screens should be considered as independent works capable of copyright protection. From the conclusion in relation to the GUI screens, it seems that the judge has considered these issues as one whilst really seeing the screens as independent works. If that interpretation is correct then there seems no real reason why the VT100 screens should not be afforded the same protection as the GUI screens (albeit that they may fall at the hurdle of originality). Pumfrey J describes the VT100 screens as a table, arguably this is distinct from the protection of computer programs afforded in s 3(1) of the CDPA 1988, and therefore the VT100 screens could potentially have copyright protection as a literary work.
Database Copyright
Pumfrey J’s finding in relation to infringement of copyright in the database by the provision of the tape containing a copy, and in particular his application of s 50D, is a little hard to follow. This section allows a person who has a right to use a database to do in the exercise of that right ‘anything which is necessary for the purpose of access to and use of the contents of that database’.
But what was the copyright to which he applied s 50D? Database is defined in s 3A(1) of the CDPA (‘a collection of [data] .’), and most often of course, since the changes implemented as a result of the Database Directive, it will be database right rather than copyright which arises. Nevertheless copyright can still exist in a database pursuant to s 3A(2) CDPA if the selection or arrangement of the contents is such that the database constitute the author’s own ‘intellectual creation’. There are two aspects to the database in this case. First there was the schema (structure) of the database, which was not, he found, a database within s 3A(1) but a separate copyright work (a computer program) and so s 50D can have had no application to that. If it was not the schema, it may then have been the contents of the database itself (the flight details and other data). However, this was the type of data one would have thought would have attracted database right and not copyright (even were that data owned by Navitaire rather than easyJet); and certainly there is no express finding that the requirements of s 3A (as to the author’s ‘intellectual creation’) were met.
Conclusion
By far the most interesting part of this case (and easiest to understand!) is that in relation to look and feel. Navitaire’s copyright gave them insufficient protection to prevent the creation of a competing product, and this leads to the question of whether contractual protection could have been included in the licence itself. A restriction on creating a similar product is unlikely to be acceptable to many licensees, and will often provoke the proviso that it does not apply when there is no licensed copyright being copied – which would of course defeat the object. In any case, if no IP is being used, competition law issues arise.
And whilst on the face of it this is bad news from a software house point of view, in the end as far as look and feel is concerned, Navitaire v easyJet may well be viewed as a result limited very much to its facts and the perhaps rather functional nature of the software alleged to be copied.
The judgment can be read in full at: www.bailii.org/cgi-bin/markup.cgi?doc=/ew/cases/EWHC/Ch/2004/1725.html&query=navitaire&method=all
Renzo Marchini is an IT lawyer at Dechert LLP (renzo.marchini@dechert.com). He was previously a software engineer.
