Dynamic's case on infringement
101 Dynamic disputed Mr Rendell's claim that Tonnex 2008 had been independently created in the way described by him. It submitted that, when Tonnex 2008 was compared with the March 2008 CSV file, the Court should conclude that Tonnex 2008 had been indirectly copied from the March 2008 CSV file, by Mr Rendell and Mr Furtado cutting and pasting entries obtained from searching the compatibility information on Dynamic's website, provided by means of the March 2008 CSV file.
102 Dynamic pointed to the fact that five of the nine columns of information in the March 2008 CSV file had been repeated, in the same sequence, in Tonnex 2008 and that each of the 1,683 product line entries in Tonnex 2008 was matched by an equivalent entry in the March 2008 CSV file (representing 60% of the product line entries in that file).
103 In respect of these entries, Dynamic pointed to the following matters:
(a) The information in the OEM Code column in Tonnex 2008 was expressed in the same letters and numbers in the equivalent column in the March 2008 CSV file (including idiosyncratic hyphens and errors), in all but 58 cases.
(b) The wording of the product description in the Description column in Tonnex 2008 matched the wording of the product description in the equivalent column in the March 2008 CSV file, except for what appear to have been systematic changes to certain terms which were likely to have been brought about by "global" corrections or changes having been made. In its written closing submissions, Dynamic produced a schedule describing the nature of these changes. It exemplified these changes in a separate document. I will return to the question of "global" changes.
(c) The information in the Yield column in Tonnex 2008 was expressed in identical terms in the corresponding column in the March 2008 CSV file, except in one case.
(d) The words and numerals appearing in the Printer Models column in Tonnex 2008 (which gave the compatibility information) were identical to those appearing in the corresponding column of the March 2008 CSV file, except in four cases.
104 Tonnex did not dispute the details of this analysis. Its submissions were directed to other considerations to which I will refer.
105 One answer to these submissions could be that the information in Tonnex, being factual in nature, was quite capable of being obtained from a variety of sources in the way that Mr Rendell described in his evidence. However, Dynamic provided detailed particulars (56 pages) of 471 instances of what it described as indicia of copying (the copying indicia). It submitted that the copying indicia represented idiosyncratic features of the March 2008 CSV file that had found their way into Tonnex 2008. Dynamic submitted that the presence of these idiosyncrasies in Tonnex 2008 was only consistent with wholesale copying by Mr Rendell and Mr Furtado of the March 2008 CSV file indirectly from Dynamic's website.
106 A small number of examples will serve to illustrate Dynamic's submission:
(a) OEM Code: In the March 2008 CSV file Dynamic used variously a hyphen between letters and numbers in the code. For example, in relation to Brother products, the code in the March 2008 CSV file was sometimes expressed using a hyphen (such as TN-2025 or TN-2130 or TN-2150) and, in other cases, without a hyphen (such as TN100 or TN200). Also, in relation to some of the OEM codes for Epson products, the letter "O" had been incorrectly used for the numeral "0". The same variances in expression are shown in Tonnex 2008 with respect to the same product line entries.
(b) Description: In cases where hyphens were used in the OEM code listed in the March 2008 CSV file, it did not follow that hyphens were used in the corresponding product description. To use the examples previously given, for OEM code TN-2025 the product description was Brother TN2025 Toner; for OEM code TN-2130 the product description was Brother TN2130 Toner, and for OEM code TN-2150 the product description was Brother TN2150 Toner. In other cases, where an OEM code included a hyphen, the hyphen was carried over into the product description. For example, for OEM code TN-12BK the description was Brother TN-12 Black Toner. Similarly, sometimes the letter "O" had been used incorrectly for the numeral "0" in the product description. Where this error was present, it was not necessarily reflected correspondingly in the OEM code. The same variances in expression are shown in Tonnex 2008 with respect to the same product line entries.
(c) Yield: In cases where yield information was missing in relation to a particular product in the March 2008 CSV file, the same information was correspondingly missing in Tonnex 2008.
(d) Printer Models: This information yielded an extremely large number of different examples of idiosyncratic expression in relation to compatibility information. Despite Dynamic's attempt to systematically order the compatibility information in the product reference field in the Navision database, it is clear on examining the corresponding compatibility information in the printer models column of the March 2008 CSV file that this information was not always entered in an alphabetical and numerical sequence in the Navision database. Also, the information was not always expressed in a consistent form. Capitalisation was used variously (for example FAX/Fax; STYLUS PHOTO/Stylus Photo; PictureMate/Picturemate; E-Studio / e-studio / e-Studio / E Studio / Estudio). Sometimes a product descriptor was used before each compatible model number and sometimes the descriptor was used only at the beginning of a sequence of product model numbers (for example, FAX 2850, FAX 8070P, and so on, compared with Fax 1010, 1020, 1025, 1030, and so on). There were variances in spelling (for example, Colour/Color). There were variances in abbreviation (for example, Clr/Col for colour; BK/Bk/Blk for Black; Mag/Magent for Magenta; Yel/Yell for Yellow). There were variances in the use of complete words and abbreviations. There were variances in description (for example, Laserjet/LaserJet/Laser Jet/LJ/HP Laser Jet; Deskjet/DeskJet/deskjet; Officejet/OfficeJet/OJet/Oj). Quite apart from these matters, the compatibility information in the relevant column of the March 2008 CSV file contained misspellings, inconsistent spacing, erroneous product references (including non-existent products) and obviously erroneous or redundant punctuation. Also within this column of information in the March 2008 CSV file there were numerous instances of multiple idiosyncratic expression. The same variances in expression are shown in Tonnex 2008 with respect to the same product line entries.
107 When one works through all instances of the copying indicia (as I have done), the following matters become abundantly clear.
108 First, in relation to the March 2008 CSV file there are a great many examples of the use of idiosyncratic expressions, including multiple idiosyncratic expressions present within and across columns in relation to the one product line entry. Given the evidence that data was entered into the Navision database manually, rather than by a "cut and paste" operation, and given that the particularised examples of these indicia are so numerous, I am satisfied that these idiosyncrasies arise from either data entry errors or decisions deliberately but inconsistently made by Dynamic's employees as to how the data should be entered in the Navision database.
109 Secondly, one cannot help but be struck by the sheer number of instances of precisely the same expression used within and across columns in relation to the very same products itemised in Tonnex 2008. It is inconceivable, in my view, that the same idiosyncrasies, in the same number and combination, and in relation to the same products, found their way into Tonnex 2008 by chance.
110 Mr Rendell said that the sources of information he used to compile Tonnex 2008 included Tonnex's own database of information collected by Tonnex's employees. This information was said to have been used in Tonnex's price lists (which, as I have noted, also include compatibility information). However, it was persuasively demonstrated in Mr Rendell's cross-examination that there were clear differences between the compatibility information in the price lists and the compatibility information in Tonnex 2008. I am satisfied that the Tonnex database was not the source for the expression of compatibility information that found its way into Tonnex 2008. Indeed, in his oral evidence, Mr Rendell referred to the fact that compiling Tonnex 2008 was a separate exercise to compiling the other compatibility information in Tonnex's price lists.
111 As I have noted, Mr Rendell denied that he had used a competitor's website (and Dynamic's website in particular) to "cut and paste" information into the file that was used to create Tonnex 2008. He said that the website information he used came from "e-commerce" websites. However, he could not remember what sites he had used; nor could he provide any working papers which showed the investigatory trail that either he or Mr Furtado followed at the time. Indeed, he could only describe in broad terms the general process that he undertook (and the process he believed that Mr Furtado undertook) in compiling Tonnex 2008.
112 Mr Rendell sought to illustrate this general process by undertaking an exercise in August 2010 in which he selected examples of the expression of information from the copying indicia, as well as some other entries in Tonnex 2008. He made a general web search against these examples and then charted the results. The conclusion which Tonnex sought to demonstrate from this exercise was that, for each selection, the expression of the information was either an exact match for the expression of the same information in Tonnex 2008 and in the corresponding column of the March 2008 CSV file or, if not an exact match, an illustration that the same information could be obtained from numerous sources by undertaking the general process he had described.
113 This exercise did in fact show what Mr Rendell had sought to illustrate. But there were a number of deficiencies in his methodology which show that Tonnex's defence is not significantly advanced by the results of the exercise.
114 First, the exercise did not purport to represent what Mr Rendell or Mr Furtado actually did at the time they compiled Tonnex 2008. Specifically, the exercise did not purport to identify the websites from which Mr Rendell or Mr Furtado actually obtained information and carried out the "cut and paste" operations to which Mr Rendell referred.
115 Secondly, Mr Rendell's exercise in August 2010 involved searching against a known form of expression (that is, the information in a particular column with respect to a particular product item in Tonnex 2008) rather than searching for separate items of information that, collectively, might go to make up the expression adopted by Tonnex in Tonnex 2008.
116 Thirdly, in carrying out his exercise, Mr Rendell looked at particular alleged instances of idiosyncratic expression by reference to a discrete column of information in Tonnex 2008. He did not look at multiple columns in Tonnex 2008 with respect to a particular product line entry. However, as I have noted, the idiosyncrasies identified by Dynamic included, in many cases, multiple idiosyncrasies across columns in relation to the one product line entry. Mr Rendell's exercise did not address this feature of the comparison between Tonnex 2008 and the March 2008 CSV file.
117 Fourthly, the exercise carried out by Mr Rendell in August 2010 was undertaken at a time far removed from the time when Tonnex 2008 was compiled. It is simply not known on the evidence whether the information extracted from the websites as part of Mr Rendell's exercise in 2010 was information that was displayed on those websites at the time that Mr Rendell and Mr Furtado compiled Tonnex 2008 or, indeed, whether the particular websites themselves were even in existence. For example, in the course of Mr Rendell's cross-examination Dynamic showed that at least one of the results obtained by Mr Rendell as part of this exercise related to a product that was not even part of the particular supplier's catalogue until 13 April 2009. Plainly enough, the website of this supplier could not have been a source of information for Mr Rendell or Mr Furtado in relation to that product at the time that Tonnex 2008 was compiled. This example left open the question of how many similar instances might exist in Mr Rendell's exercise.
118 Fifthly, the source of the e-commerce suppliers' information, as shown on the websites visited by Mr Rendell as part of this exercise, is not known. It may be, for example, that the expression of this information was itself taken directly or indirectly from Dynamic's own earlier published compatibility charts, such as the March 2008 CSV file.
119 What the exercise did not show was that, given the multiple idiosyncrasies to which I have referred, those same idiosyncrasies existed in relation to the same products in the same number and in the same combinations in a source of information that did not originate with Dynamic but which was nevertheless accessible to Tonnex for copying, in that form, at a time prior to 7 March 2008, into an Excel file by "cut and paste" operations.
120 There are, however, other features of Tonnex 2008 to which reference should be made.
121 It was part of Tonnex's case that it only supplied OEM products as opposed to generic or "white box" products that were compatible with or substitutable for OEM products. This was made clear in Mr Rendell's evidence. It was also made clear by Mr Kozman, one of Tonnex's directors.
122 Dynamic supplies generic products as well as OEM products. Both generic and OEM products were listed in the March 2008 CSV file. The entries for generic products were cross-referenced against the OEM codes for the products with which they are compatible. Thus a search made on Dynamic's website (after the March 2008 CSV file had been uploaded) against the OEM code for a particular product would throw up both the OEM product and corresponding compatible generic product that Dynamic supplied at that time. The product description in the line entry for the generic product used the word "compatible" (or some form of it) to indicate the compatibility of the generic product with the printer models of the OEM.
123 There are instances in Tonnex 2008 where references to certain products listed under a particular OEM code have been duplicated. These instances relate to Canon and Kyocera products. In the case of the duplicated entries, the product description is the same as that appearing in the corresponding column in the March 2008 CSV file with respect to the generic product supplied by Dynamic, including use of the word "compatible" (or some form of it). There are a number of examples of this.
124 Mr Rendell was unable to explain this apparent duplication in Tonnex 2008 and the fact that the word "compatible" (or some form of it) was used in the product description in the duplicate entry. Mr Kozman disputed that there was duplication for some Canon products listed in Tonnex 2008. His explanation was that Canon had different code numbers for different regions and that the apparently duplicated entry must have been referring to the same Canon product from a different region. Mr Kozman said that, in those instances, it was appropriate to use the word "compatible" as part of the product description. I do not accept this explanation. It defies a commonsense reading of the relevant entries in Tonnex 2008 and ignores the fact that the obviously duplicated entries in Tonnex 2008 did not use different OEM codes for different regions, but the same OEM code.
125 From this I conclude that: (a) in Tonnex 2008, Tonnex duplicated certain entries it had made; (b) these entries were in fact redundant because, by the product description, they plainly relate to products of a kind that Tonnex said it does not supply (that is, generic compatible products); and (c) these redundant entries correspond with actual entries in the March 2008 CSV file which were separately listed by Dynamic because they were separate product items in Dynamic's catalogue which it, unlike Tonnex, does supply as generic products that are compatible with OEM products. I am satisfied that these entries in Tonnex 2008 have been copied from Dynamic's website and thus indirectly copied from the March 2008 CSV file.
126 Another feature relates to the question of "global" changes. As I have noted, Dynamic submitted that the wording of the product descriptions in Tonnex 2008 matched the wording of the product descriptions in the equivalent column of the March 2008 CSV file, except for what appear to have been systematic "global" changes. Many instances of this were noted. I will not describe all of them in these reasons. It will suffice to refer to some examples as illustrations. Examples include, in a number of instances:
(a) references to "Cart" after the word "Ink" being replaced with "Cartridge";
(b) references to "Cart" or "Cartridge" after the word "Toner" being deleted;
(c) references to "Mag" being replaced with "Magenta"; and
(d) references to "Yel" being replaced with "Yellow".
127 Dynamic submitted that the making of these "global" changes by Tonnex was evident from some anomalous product descriptions in Tonnex 2008. It submitted that these anomalies could be explained by Tonnex having copied product descriptions from the March 2008 CSV file (including demonstrable examples of inconsistencies in Dynamic's own form of expression) and then having applied "global" changes to them when, in some cases, it was inappropriate to do so.
128 For example:
(a) In the March 2008 CSV file the description of the Lexmark product identified by OEM code 10S0063, was "Lexm 10S0063 Toner Cartridge" (the abbreviation "Cart" for the word "Cartridge" not being used by Dynamic in this particular instance). Dynamic submitted that the corresponding description in Tonnex 2008 was corrupted when, in making a global change, Tonnex deleted the word "Cart" after the word "Toner", so that "Toner Cartridge" in the March 2008 CSV file became "Tonerridge" in relation to the corresponding entry in Tonnex 2008.
(b) In the March 2008 CSV file the description of the Xerox product identified by OEM code 106R01145 was "Xerox Phaser 6350 Magent Toner" (the abbreviation "Magent" rather than "Mag" being used by Dynamic in this particular instance). Dynamic submitted that the corresponding description in Tonnex 2008 was corrupted when, in making a global change, Tonnex replaced "Mag" with "Magenta", so that "Magent" in the March 2008 CSV file became "Magentaent" in relation to the corresponding entry in Tonnex 2008. Similarly, when Dynamic used "Yell Toner" rather than "Yel Toner" the resulting global replacement by Tonnex of "Yel" with "Yellow" resulted in "Yell" in the March 2008 CSV file becoming "Yellowl" in the corresponding entry in Tonnex 2008.
129 Dynamic submitted that systematic "global" changes were also evident in the compatibility information in Tonnex 2008. In some instances Tonnex 2008 refers to "Deskjet PortaBlacke*" as a compatible product. The description "Deskjet PortaBlacke*" is plainly anomalous. There is no product so described. The corresponding compatibility information in the March 2008 CSV file refers to a "Deskjet Portable*". The anomalous expression in Tonnex 2008 is consistent with global editing by the substitution of the word "Black" for the abbreviation "bl", so that "Portable*" in the March 2008 CSV file became "PortaBlacke*" in the corresponding entry in Tonnex 2008. The presence of the asterisk in "Deskjet PortaBlacke*" in Tonnex 2008 is also significant in light of the description "Deskjet Portable*" in the March 2008 CSV file.
130 I am satisfied that these examples show that, in compiling Tonnex 2008, Tonnex not only made global changes but applied these changes to extracts from the March 2008 CSV file that had been copied by Tonnex.
131 It is clear on the evidence that a copy of the March 2008 CSV file was published on Dynamic's website before Tonnex 2008 was created. Mr Rendell accepted that he kept an eye on Dynamic's website in 2006, 2007 and 2008 and that, in 2008, Dynamic had moved from what he called a "typical website" to one that had a search function. He accepted that it was possible that he could have seen this new search function in March 2008 and that he "saw how it operated and how it worked and how it derived data".
132 On the objective evidence, the similarities between Tonnex 2008 and the March 2008 CSV file, as to layout and format and the expression of information and its arrangement, are so numerous that I am satisfied, on the balance of probabilities, that large parts of Tonnex 2008 have been copied indirectly from the March 2008 CSV file. This copying occurred by a process which involved "cutting and pasting" from the compatibility chart on Dynamic's website. This took place relatively shortly after the March 2008 CSV file had been uploaded to Dynamic's website on 7 March 2008.
133 Although Mr Rendell gave his evidence in a careful and considered manner, and responded directly to questions put to him, his evidence in relation to the process that was undertaken in compiling Tonnex 2008 simply cannot stand with the overall and compelling picture presented by the detailed objective evidence adduced by Dynamic on the issue of copying. In the circumstances, I prefer the objective evidence. It follows from this conclusion that I reject Mr Rendell's explanation that Tonnex 2008 was compiled only from information that was sourced from Tonnex's own database and computer system or otherwise from websites that did not include competitors' websites (particularly Dynamic's website).
134 Tonnex submitted that, in light of all the evidence, Dynamic has not shown, on the balance of probabilities, that Tonnex had reproduced a substantial part of the March 2008 CSV file in Tonnex 2008. It submitted that the question of infringement overlapped the question of copyright subsistence and that it was necessary to show that Tonnex had appropriated the expression of the author's (in this case, Mr Campbell's) creative and intellectual effort, not merely individual items of information that were in the public domain. Tonnex stated the question for resolution as follows: Has, therefore, the respondent appropriated the selection and ordering of items which originated with the author?
135 Having considered and reflected on the objective evidence, I am satisfied that that question must be answered affirmatively. For the reasons I have given, I am satisfied that copying of the March 2008 CSV file has taken place. I am satisfied that, considered quantitatively and qualitatively, a substantial part of the March 2008 CSV file has been reproduced in Tonnex 2008.
136 In that regard, I am satisfied that the indicia of copying reflect, in a broad way, the minimum extent of the copying that has taken place. I conclude that the indicia of copying show that, more likely than not, Tonnex copied the OEM code, product description, yield information and printer models (compatibility) information with respect to each product line entry in the March 2008 CSV file to which the indicia relate. However, I infer that the extent of copying was far greater than that and, more likely than not, included substantial parts of the March 2008 CSV file not marked out by the use of idiosyncratic expression. It is not possible for me to be any more precise as to the extent of copying.
137 Also, although not all of the March 2008 CSV file was copied in Tonnex 2008, qualitatively important elements of it were copied. I refer in particular to the combination of the OEM code and the corresponding, particularly expressed, product descriptions and compatibility information that Mr Campbell had selected for the March 2008 CSV file. It was the selection of these elements that gave the March 2008 CSV file its essential character as a meaningful compatibility chart in relation to the products to which it referred. Equally, these elements gave Tonnex 2008 its essential character as a meaningful compatibility chart in relation to the products to which it referred. I do not ignore the yield information that I have found was also likely to have been copied. That element of the March 2008 CSV file does not, however, have the same qualitative significance for a compatibility chart as the other elements to which I have referred.
138 Elements not taken from the March 2008 CSV file were either obviously surplus to Tonnex's requirements (such as Dynamic's own product code or the separate identification of the manufacturer (given that Tonnex 2008 is arranged according to manufacturer)) or unnecessary, albeit possibly helpful, information, such as the information in the separate columns in the March 2008 CSV file giving manufacturer and colour information. The fact that this possibly helpful information (which itself contributed to the originality of the March 2008 CSV file as a literary work) was not taken does not mean that Tonnex 2008 is not a reproduction of a substantial part of the March 2008 CSV file.
139 Finally, the fact that Tonnex has reproduced a substantial part of the March 2008 CSV file is also evidenced by the fact that five of the nine columns of information in the March 2008 CSV file have been repeated, in the same sequence, in Tonnex 2008.
140 In light of Mr Rendell's evidence that the compatibility charts comprised in Tonnex 2008 are included in the Tonnex International February 2009 Price List and the Tonnex 2010 Price List, the same conclusion must follow with respect to those publications. Tonnex did not seek to contend otherwise.
141 I am therefore satisfied, on the balance of probabilities, that Tonnex has infringed Dynamic's copyright in the March 2008 CSV file.