Proposed Order 2 - source CodeS AND OBJECT CODES
6 Optiver seeks in proposed Order 2 a series of documents including all versions, whether current, archived, deleted or otherwise, comprising or relating to source codes and object codes for Optiver Computer programs.
7 The expression "Optiver programs" is defined to mean twelve specific programs in a definition to the Proposed Minutes of Order, and "documents" is given the meaning in O 1 r 4 of the Federal Court Rules and this covers a broad range of records. The definition of "functions" in the proposed orders refers to groups of program functions.
8 Under proposed Order 2, documents relating to source and object codes and any historical source code repository are sought, together with source codes for share auto trading. Also included in the proposed order are all source code revision logs relating to the development of computer programs and documents evidencing communications as to the location of programs and the source or origin of computer programs that perform any of the functions, as defined.
9 An affidavit was filed on behalf of Tibra by Mr Andrew King (Mr King), sworn on 17 December 2008. Mr King is the eighth respondent in the proceedings and he is employed by Tibra Global, the fourth respondent. He is an experienced software developer. He points out that the source code repository is an historical facility which stores source codes written by software developers for Tibra Global's computer programs. During his employment with Optiver, Mr King worked on maintaining an automated trading program for twelve months, and was a team leader of the relevant project for approximately six months. He says that in relation to the expression "functions as defined in proposed draft orders," the programs did not perform some of those functions mentioned.
10 In relation to the difficulty, cost and expense of production of the documents sought in Order 2, Mr King says there are approximately 35 software developers world-wide employed at Tibra Global, with many additional staff involved in the production of trading systems, and that there would need to be at least 35 hard disk drives that would need to be duplicated to provide a digital copy for inspection by an independent expert. He says that approximately 30 percent of the hard disk drives are not located in Australia and that it would take at least one to three hours for each hard disk drive to be imaged in relation to the production to the disk drives data storage disks and memory sticks by the fifth to ninth respondents.
11 Mr King notes that Optiver has not limited its requests for production of the documents as to time, and says that compliance would involve a major interruption to work performed by each of the fifth to ninth respondents.
12 Mr King refers to the expression "object code," in respect of which the records are sought. He explained that computer programs are developed in stages and in the first stage a series of documents in a particular programming language are written in a way similar to writing which can be understood by humans. These documents are known as the "source code" and cannot be used by a computer. The second stage involves use of a compiler, which is a computer program that understands the programming language and which converts the source code into an object code. The latter is only created if the compiler can understand the source code, so that the object code is the final working computer program understandable by a computer which is used to run the computer. Where development involves a continual compiling of the source code into an object code this can involve carrying out the above process at least twenty times a day. When a particular code is satisfactorily developed then the change will be committed to the historical source code repository. Object codes are not stored. However, source codes are stored and if an object code is required in order to use a program at a later date the source code is run through the compiler.
13 Mr King's evidence is that to produce an object code for a particular version of the source code it would be necessary to obtain the source code from the historical source repository and run the program. He says that the automated trading application the subject of this proceeding was developed between June 2006 and December 2008, and there have been approximately 10,000 different versions or iterations of the source code developed collectively, and that in order to produce the object code for each of the 10,000 versions held in the repository of Tibra Global it would take in the order of 20,000 hours.
14 The evidence of Mr King was not challenged. In the light of the above, Tibra contend that the discovery sought in Order 2 is excessive.
15 In relation to a number of the paragraphs in Order 2, no time period is specified. The relevant time frame is over about two and a half years. However, in regard to the evidence of various versions of the programs over time and the importance of later versions in determining whether there has been a use of the original programs, I do not consider that the existence of a time limitation is determinative. The Full Court indicated that the current source code used by Tibra could be relevant but various iterations may, for example, bear the hallmark or footprints of copying from Optiver's program and information by and of indirect copying adaptations from intermediate versions. I agree that with regard to the continuing period of development of what is now the current Optiver program there may be a number of iterations of that program which may be relevant and I do not think the objection to specific dates or specific versions is appropriate. That is too restrictive. I do not therefore accept the submission by Tibra that Optiver's entitlement to inspect source codes should be limited to computer programs which performed the functions for example as at 5 September 2006 and 19 June 2007.
16 Having regard to the evidence in relation to the close connection between the object code and the source code as stated by Mr King, I do not agree with the proposition that the relevance of the object code to Optiver's case has not been made evident.
17 In the light of the material provided by Optiver on this application, I am not persuaded that the historical repository of source codes is irrelevant, nor am I persuaded that documents relating to it are unnecessary to enable a proper decision to be made under O 15A r 6 of the Federal Court Rules.
18 Some objections to Proposed Order 2 were made by Tibra on the basis that some of the sub-paragraphs replicate documents that will be produced under other subparagraphs, however I do not think this is a sufficient ground for objecting to those paragraphs.
19 Accordingly, my conclusion is that the documents as sought in Proposed Order 2 should be discovered.