The Facts
24 The most important witnesses at the trial were Mr Petty, the first respondent and the second respondent. The trial judge did not make express general findings about the credit of each of these witnesses. He appears to have accepted most of Mr Petty's evidence and the evidence of the second respondent. He accepted parts of the first respondent's evidence and rejected other parts. Perhaps because of the way he understood the appellant's case, the trial judge concentrated on the first respondent's state of mind between February and May 2008 and his findings with respect to the second respondent's state of mind are not entirely clear. However, as I will explain, I have been able to reach conclusions about that matter.
25 The other general point to note is that the appellant does not challenge any of the trial judge's findings of fact, save for his finding that Simple Card did not represent a business opportunity. Whether that is a finding of fact may be debatable.
26 The trial judge found that the appellant launched various products during the period from November 2002 to April 2006. In November 2002, the appellant launched a product called 'DeckLink', a capture card, which was, at the time, one of the first 10-bit uncompressed capture cards to work seamlessly with the Macintosh operating system. Some time between November 2002 and April 2004, the appellant launched its 'DeckLink HD' (high definition) product. In April 2004, the appellant launched its 'DeckLink HD Pro' product which Mr Petty said was the world's first true 10-bit and 12-bit dual link 4:4:4 and 4:2:2 high-end video capture card suitable for both Macintosh and Windows systems. The appellant also launched what Mr Petty described as its 'HDLink HD Monitoring Solution', which allowed customers to use a 24-inch computer monitor for the highest quality HD TV video monitoring. In April 2005, the appellant introduced a new type of video editing product called 'Multibridge Extreme' which solved the problem of insufficient space on standard computer ports to add extra audio channels by moving the electronics to an external box. In April 2006, the appellant released its 'DeckLink HD Extreme' product, which featured both SDI and analogue input/output connections that switched between high definition and standard definition video.
27 After the first respondent had taken up the position Director of Hardware Engineering, he built up and managed a large hardware engineering team. The trial judge considered that an idea of the first respondent's responsibilities could be gained from a job specification written by the appellant for the first respondent's replacement:
Responsibilities
• To lead hardware development for Blackmagic Design
• Oversee new product development and design implementation in conjunction with the CEO and Product Managers
• Define product specifications
• Manage product timescales and time to market imperatives
• Manage product testing specifications
• Manage new part selection and pricing negotiations
• Manage a bug system and assignment of tasks to engineers as part of a planned product release
• Joint responsibility, along with the Director of Software, for ensuring interoperability between hardware and software development
• Authority to Hire and Fire
28 The second respondent was employed by the appellant in April 2006. He executed a written employment contract on 23 November 2006. The written contract contained the following description of his position:
Schedule B: Job Description and Hours of work
• Business Development Manager - Developing the sales channel for Blackmagic Design, and finding new opportunities for relationships with other manufacturers and industry partners.
As the trial judge put it, the second respondent was, in effect, the head of sales for the appellant.
29 In late 2006 and early 2007, the first and second respondents showed an interest in making electronic products in the audio industry. It is not necessary to set out all the evidence of this. It included the first respondent building a pair of loud speakers at his home and the creation of a spreadsheet named 'industry.xls' on 31 May 2007 and modifications to it up to 27 November 2007. It also included the incorporation of the third respondent on 12 August 2007 and the drafting of patent claims to do with loud speakers by the first respondent. In addition, at about this time the first and second respondents were liaising with a design or branding firm about the 'Atomos concept'.
30 On 28 November 2007 the fourth respondent, at the request of the third respondent, opened an account for Atomos at a collaboration web site called 'Grouphub'.
31 The trial judge referred to events in late 2007 which he said provided 'a kind of context' for the creation by the first respondent of the contentious worksheets in 'ycalc'. In describing the events, the trial judge referred to the evidence of Mr Petty and the evidence of the first respondent without stating whether he preferred one over the other. I think he must have taken the view that the broad effect of the evidence was the same and where it differed the differences were not material to the issues before him.
32 In late 2007, Mr Petty had an idea for the replacement of all of the appellant's then existing models with a new one-lane PCI express card, based on the product called 'Intensity Pro'. The replacement product was to encompass all of the features offered by the other cards in the appellant's range. The first respondent was opposed to this idea. He considered that it would leave too great an opportunity for a competitor to come up with a much simpler and cheaper card. He was also concerned that Mr Petty's idea meant that there would be no model which did not have a high definition feature because the inclusion of high definition added considerable cost to the product. Mr Petty did not share the first respondent's concerns. He considered that his idea would result in a card that was so attractive that people would not want a cheaper card and, in any event, if the appellant needed to come up with a cheaper card it could do so quite quickly.
33 The first respondent set about costing the kind of one-lane card that, on his understanding, Mr Petty had specified. A fellow employee, Mr Hunter, helped in this task. They reached the conclusion that the card would only be about $20 cheaper to build than a new version of the DeckLink HD Extreme which was in development at the time, the DeckLink HD Extreme 2.
34 The trial judge said that the development of the DeckLink HD Extreme 2 was itself the cause of a difference in view between Mr Petty and the first respondent. Mr Petty was of the view that the existing DeckLink HD Extreme would be discontinued simultaneously with the commencement of production of the new product. The first respondent considered this course of action to be a 'highly risky strategy'.
35 On 10 January 2008, the first respondent created a spreadsheet on his private computer called 'ycalc.xls'. The following features of the spreadsheets should be noted. First, the spreadsheet consisted of different pages or 'worksheets' as they were called. Secondly, the worksheets fell into two different categories, viz, ideas and calculations for audio products about which there was no complaint by the appellant and three worksheets not created before 7 February 2008 which contained details about video products, including products of the appellant.
36 The trial judge found that the first respondent examined an audio product made by a company called 'Motu'. A short time later, he examined Motu's V3 HD video product and thought that Motu could quite easily launch a range of SDI video PCI cards which matched the appellant's range at every level where the appellant had a product.
37 The first respondent then took action in relation to the 'ycalc' spreadsheet. He said that he did that because he was concerned about potential competition from Motu and because of his concerns that Mr Petty was specifying a more expensive product than was necessary. He prepared a worksheet in the 'ycalc' spreadsheet called 'SDI'.
38 The first respondent did two things in the SDI worksheet:
1. He wrote down a list of five video interface cards which he called 'simplest option', 'killer card', 'ultracheap SD card', '4 lane megamotha' and the 'ultracheap HDMISDI interface'. He set out the components of those products and the costs of those components and he calculated the profits which might be made by those products.
2. He wrote down the appellant's products and calculated the costs and profits which might be derived from those products.
39 The first respondent concluded that a hypothetical competitor could well produce products equivalent to the appellant's products but with a lower cost base.
40 The first respondent raised his concerns about what Motu might do in competition with the appellant with Mr Petty in early February 2008. In a vigorous exchange he was rebuffed by Mr Petty who said that the appellant should not and would not concern itself with what a competitor might do.
41 The trial judge found that the first respondent's concerns and Mr Petty's reaction prompted the first respondent to undertake further development of the 'ycalc' spreadsheet. He prepared two further worksheets, one entitled 'costs' and the other entitled 'allcosts'. The trial judge said it was not clear on the evidence when, and in what sequence, the first respondent developed the SDI worksheet and the two further worksheets entitled costs and allcosts.
42 The trial judge described the contents of the allcosts worksheet as follows (at [26]):
There was a column with the names of products, the first four of which corresponded with the first four mentioned on the "SDI" worksheet (see para 23 above) and were, respectively, "Simple card", "Killer card", "SD card" and "4 lane". They represented products of the hypothetical competitor. The remainder were products of the applicant, or which the applicant was contemplating introducing. The production cost of each of the applicant's then products was set out. The major component of that cost, and one which was stated accurately to the cent in the worksheet, was the amount which the applicant would pay to obtain the components that, when assembled, made up the product. Mr Overliese said that these costs figures were "based only on information which I used daily for my work" with the applicant. Mr Overliese included in the worksheet the usage of the parts in question per month from August 2007, details which he also obtained from the applicant's records to which he had access. The manufacturer's margin, 65%, was, he said, a "guess" which he applied across the board. However, he said that he knew "roughly" what the applicant's margins were, because Mr Petty would discuss with him how much profit he would make whenever they talked about the total cost of his ideas for a product.
He described the contents of the costs worksheet as follows (at [27]):
Mr Overliese then created a further worksheet, called "costs". Into that worksheet, Mr Overliese copied the contents of the "allcosts" worksheet. He added two columns, headed "Bang for buck" and "percent profit". At the time (ie in early February 2008) he entered data only with respect to the applicant's products. The cells in the new columns corresponding to the four hypothetical competitors' products were left empty. The figures in the "bang for buck" column presented the annual unit profit for each of the applicant's products, and the column headed "percent profit" represented the calculated percentage contribution to overall profits made by the profit on each product.
43 The trial judge said that it was not possible to put a precise date on when the 'ycalc' spreadsheet reached this stage of development. Nor was it possible to say when the first and second respondent 'came to thinking' that the third respondent might produce video, as well as audio, products. He found that by 2 or 3 February 2008 the first and second respondents at least had a provisional intention that the third respondent should produce video products. He found that prior to 3 February 2008 the first respondent did not envisage that the third respondent would produce video capture cards. He found that the second respondent appreciated that if the third respondent offered the service of an uncompressed card it would 'cross over' with the capacities of the appellant's cards.
44 In early February 2008, Mr Petty raised again his idea of introducing a new one-lane card that might take the place of some of the appellant's existing products. On 4 February 2008, he sent an email to the first respondent which was in the following terms:
What are your thoughts about reducing some of the features on the one lane version of Decklink Extreme PCIe 2.
We could then enable them if we need to as we want to later.
I was thinking about:
No 24 fps frame rates, just 1080p/50 and 1080p/5994
No 3D lookup table?
I am looking for a few things that we can disable, but then instantly enable if we need to in the future. Any ideas?
45 The first respondent responded seven minutes later as follows:
We can easily do both those, which would mean the differences would be:
* Only 8 bit HD (forever)
* No RGB HD in premiere (forever)
* No 3GSDI (ever) - i.e. no 2K or RGBHD
* no 24 frame rates (can be enabled at any time)
* No 3D Luts (can be enabled at any time)
* no HD desktop (could possibly add a RGBYUV s/w conversion?)
* every 3rd frame of HD output is watermarked with a Blackmagic Design logo (could be disabled any time)
46 Mr Petty took the matter further in an email sent to members within the appellant's organisation on 5 February 2008. Mr Petty's idea at that point was for a PCI express one-lane version of the DeckLink HD Extreme 2 card to replace the DeckLink SP PCIe and the DeckLink Extreme PCIe cards. The new card would not have a desktop feature, and would be only 8-bit HD, but 10 and 8-bit SD. It would have all the features of DeckLink Extreme and 'plug into the cheaper machines'. Mr Petty's idea was that the new card would be sold at a retail price of $695 and produce a profit of $226 per unit. An employee of the appellant, Mr Kristian Lam, responded to Mr Petty's email, expressing quite obvious reservations about the proposal. The first respondent responded to those reservations, coming to Mr Petty's defence albeit in a very limited way. The second respondent responded by expressing a preference for SD only at the start and saying it was very important to keep HD Extreme going. Mr Petty terminated this round of exchanges by saying the matter could wait until after the exhibition conducted in April each year by the National Association of Broadcasters in the United States, at which the appellant was a significant exhibitor.
47 There was a telephone call between the first and second respondents in early February 2008. The second respondent was in Japan at the time. The first respondent raised the possibility of the third respondent making products which would compete with the appellant. The second respondent rejected the idea saying to the first respondent, 'You are crazy'.
48 The trial judge found that there was a good deal of activity involving Grouphub on 7 February 2008. The second respondent uploaded a document entitled 'Objectives to Outcomes' which the trial judge described as quite 'benign' and I need not mention it any further.
49 The first respondent uploaded onto Grouphub a version of 'ycalc' and a message which I will call the first message. It had the subject line 'ycalc update' and was as follows:
BOM sheet has 3 cost/price scenarios for a 12ch audio box.
As it stands, it is about right. Whether we do this product or not is of course another matter. "Costs" sheet is interesting: it is BMD's last 6 months and cost/sale price comparison across the whole product range.
Password is "atomos2924" and is not compatible with Numbers or any other mickeymouse apps :) (squeeek!)
Let me know if this is a problem.
I have some dramatically profitable ideas for SDI products, but have not detailed them yet.
50 The first respondent uploaded a word processor document called 'marketing plan'. It is not easy to reconcile the trial judge's findings with respect to this document. At one point in his reasons he states (at [43]):
The "marketing plan" document uploaded by Mr Overliese on to Grouphub on 7 February 2008 contained a series of bulletpointed observations, grouped under the headings "Objectives", "The Message", "Target Customer Demographics" and "USPs" (the latter of which represents "Unique Selling Points"). The next (and the largest) section of the document was headed "Competitors", and named three other companies, including Motu but not including the applicant. Under each company was a listing of the characteristics thereof, including its strengths and weaknesses. The document related only to audio products and the companies trading in them.
Later he states (at [53]):
I refer next to the "marketing plan" document which I have mentioned briefly at para 43 above. It related to the video industry, and was headed "Atomos Video Marketing Plans". It included details relating to two "competitors", the first of which was the applicant. The way Mr Overliese (who created this document) set out the details which related to the applicant and those which related to the other potential competitor is instructive. For example:
• Size
The applicant: "$25m turnover"
The other:
Revenue estimate TBD. Guess around $50M?
Would say around 15 hardware & 10 software engineers, sales offices?
• Weaknesses
The applicant:
- Profitability too low, a 20% slump in sales will result in cutbacks.
- Quality of manufacturing poor, few capabilities and high return rate
- Marketing & advertising strategy nonexistent
- Reputation for "cheap, so you get what you pay for"
- Poor customer support
- Software very weak and not much chance of improving
The other:
- Slow and conservative engineering, easy to beat features
- Actual signal quality does not live up to the reputation - though mostly in spec.
- Marketing extremely techie, relies almost 100% on sales channel for explanation.
The evidence is that this document was last saved by Mr Overliese on 30 March 2008. To an extent at least, Mr Overliese was using his knowledge of facts about the applicant that would have come to him only in the course of his employment, to develop a video marketing plan for Atomos.
51 The documents before this Court suggest that what his Honour says in the later part of paragraph 43 is not accurate.
52 The first respondent uploaded yet another version of the 'ycalc' spreadsheet, together with a message which I will call the second message. It had the subject line 'SDI products added to ycalc - exciting news' and was as follows:
I have put down 3 SDI cards we could do, that all start from Intensity and cost/features blow away everything BMD has.
Check the "Killer" card specs: I think it would be a product that really hits the sweet spot. The cheap SD card is $65 cheaper to make than decklink extreme and would be hard to beat on price. Not that I think a price war is a good idea, but that is another story.
53 The trial judge drew certain inferences from the course of events which neither party challenged on appeal. First, he inferred from the first message that the first respondent had in mind details of the appellant's products. Secondly, he inferred from the second message that as the first respondent had included details of three SDI cards including the 'killer' card and a cheap SD card that he was thinking 'however theoretically and however unrealistically' that the third respondent might provide SDI cards in competition with the appellant. It was at about this time that the first respondent contacted the second respondent in Japan and had the telephone conversation referred to in [47] above.
54 The second respondent could not view the new spreadsheet on Grouphub on 7 February 2008. On 9 February 2008, the first respondent uploaded a version of the spreadsheet entitled 'ycalcNOPW.xls' and that spreadsheet was downloaded by the second respondent.
55 The trial judge said the following of the 'ycalcNOPW.xls' spreadsheet uploaded by the first respondent onto Grouphub on 9 February 2008 (at [41]):
It has formed the basis of my description of the various worksheets as set out above. In the "costs" worksheets, it is revealed that Mr Overliese summed the monthly and annual profits with respect to the applicant's products, and marked that "BMD", and he performed the like calculation with respect to the products of the hypothetical competitor, and marked that "Atomos". Mr Overliese sought to defend his actions by pointing out that the worksheet was organised from the applicant's perspective in the sense that Atomos was the hypothetical competitor whose activities had been the subject of his concerns raised with Mr Petty. I consider, however, that Mr Overliese's messages on Grouphub on 7 February 2008 give the lie to that self-serving explanation. I accept that, initially, Mr Overliese had no intention of competing with the applicant. However, something happened in the first week of February 2008 (or thereabouts) to cause him (and Mr Young, it seems) at least to give consideration to the notion that Atomos might trade in video products. The most likely inference is that Mr Young's email of 3 February, together with Mr Petty's peremptory rejection of any consideration of the potential cost structures of a competitor, caused Mr Overliese to switch his analysis to one in which Atomos would now be in competition with the applicant, but for whatever reason, that is unambiguously what happened on 7 February 2008 or thereabouts.
56 There is a particular topic which I must address at this point. It will be recalled that in an email dated 4 February 2008 (see [44] above) the first respondent said:
*only 8 bit HD (forever)'
57 In the 'SDI' worksheet in the 'ycalcNOPW.xls' spreadsheet there is an entry (which the trial judge found must have been made by the first respondent before 9 February 2008) which read as follows:
10 bit HD should be doable - if the DMA style is changed.
That entry is in the form of text in the 'simplest option' section of the worksheet, associated with a side heading reading 'software features'.
58 DMA is an acronym for direct memory access. The trial judge set out the first respondent's evidence as to what he meant by this entry. He accepted the first respondent's evidence that as at 9 February 2008 (and indeed as at the end of his employment with the appellant) he had not actually worked out, even with a change in DMA, how to achieve 10-bit HD over one lane.
59 The trial judge then turned to consider whether the first respondent had a continuing intention to compete with the appellant. In other words, did he briefly entertain the idea in early February 2008 or did he have an intention to compete with the appellant throughout February, March and April 2008? The trial judge said (at [45]) that it was 'nigh impossible to arrive at an objective resolution of this point of evidentiary conflict between [the first respondent] and [the appellant]'.
60 The first respondent's case was that the idea left him as quickly as it had come to him over the course of a few days in the first half of February 2008. There were variations to the 'ycalc' spreadsheets on 28 February, 23 March and 7 May 2008 but some of the changes were automatic and the reference to the third respondent was not removed, more because of inertia than anything else. The trial judge said that the question was made more difficult by the fact there was no way of knowing when over a three-month period detailed changes to the spreadsheet were made and by the fact that the first and second respondents continued in their broad proposal to deal in video, as well as audio, products whenever the third respondent should commence business.
61 The trial judge considered subsequent events in an attempt to resolve the issue.
62 The fourth respondent carried out work on 13 February 2008 but as the trial judge made no relevant findings with respect to it I need not mention it any further.
63 On 19 February 2008, the first respondent saved a spreadsheet with a new name, 'cashflow_iano.xls'.
64 The trial judge made the following finding (at [48]):
A worksheet called "sales" set out the prices, ratios, costs and sales (on "high", "med" and "low" assumptions) for eight products. They were: "Killer external no SDI", "Killer external (with SDI)", "Ultra cheap SD only", "Basher", "Flasher", "Killer 1HD extreme killer 1 x lane PCIE", "Killer 2 - high end killer" and "Killer 3 - SD only".
His Honour went on to say (at [48]):
Mr Overliese acknowledged that none of these was an audio device, and that all of them (save "Basher" and "Flasher") were video devices. However, he said that none of them was a "real product" - they were all no more than "placeholders". He continued:
This was, you know, conjecture. It was just if we did - you know, if you had a company that had these kind of product line up and, you know, I fiddle around with the cost of goods and the retail price and see what kind of cash flow you'd get for products. … You [ie counsel for the applicant, crossexamining] keep saying we're making plans all the time, but there's a difference between making a plan to do something and just playing with an idea.
65 On 4 March 2008, the first and second respondents met. The first respondent subsequently prepared notes of the meeting which were as follows:
Agenda
• Go over the cashflow_iano spreadsheet and insert realistic sales targets and analyse results
Outcomes
• Growth goals are realistic and achievable!
• It is very important, for the strength of the company, to quickly produce as many products as possible in the early stages. This is for company growth and protection from competition.
• Do video products as early as possible, related to the above point, and the fact that these are strong capabilities we have
• Sales have three main mechanisms for growth:
1. Volume (maximization). Sales and marketing techniques to ensure maximum sales of the product by considering and targeting the end users/purchasers
2. Growth into new markets
3. Producing products which answer the needs and requirements of end users/purchasers
Actions
Ian
• Get the windows DDK as quickly as possible for software engineer(s) to start work on drivers
• Book appointment for lawyers to go over Jeromy's contract and advise on competitive actions in video market
Jeromy
• Put sales growth mechanisms into sales plan document
66 The trial judge rejected the first respondent's evidence that his focus was not then on the sale of video capture cards and that the video products he envisaged getting out more quickly were not capture cards saying that it was 'hard to take [his] denial seriously' (at [49]).
67 On 18 March 2008, the first respondent created a word processor document headed 'business Plan Jeromy Ian Both'. The document was an outline for a business plan and it had bullet-pointed notes under the headings 'Sales Channel', 'Competitive Analysis', 'Product Plan' and 'Marketing Plan one for Audio, one for Video'. The trial judge found that the notes under 'Product Plan' made it clear that both audio and video products were contemplated. Three scenarios were referred to in the document:
1. All video: 5 video products first year, 5 second year more video
2. Video start, adding video: 3 video capture, then 2 other video + 2 audio, second year 2 & 2
3. Audio start, adding video: 3 audio products then 2 video, second year 3 video + 1 audio
Details were provided of the first scenario only:
Start with 3 video devices:
1. Killer external SDI
2. Killer external analog
3. Killer internal 1 lane
The document also stated:
Next 2 devices:
1. Killer Pro x 4 + codec accel
2. Killer converter SDIDVI single?
3. Killer audio embedder?
4. CF - raid
68 The trial judge said that this document was concerned with video capture cards and he said that he was unconvinced by the first respondent's evidence that he and the second respondent never intended that the third respondent would produce such cards and that the document was another example of purely conceptual work on his part.
69 On 21 March 2008, the first respondent created a word processor document which was, he said, a series of biographical notes about him which might form the basis of a later document that could be used to impress investors in the third respondent. The trial judge drew two conclusions from the document. First, he concluded that the first respondent was still, on 21 March 2008, intending that the third respondent should operate in the video, as well as in the audio market, and secondly, that at the time he created this document the first respondent had in mind his assessment of the potential profitability of video products, derived from his knowledge of the appellant's cost and profit figures.
70 I have already referred to a finding the trial judge made in relation to the marketing plan document which was last saved by the first respondent on 30 March 2008 (see [50] above).
71 The trial judge inferred from a word processor document created by the first respondent on 1 April 2008 and saved as 'Plan Overview' that participation in the market for video products including uncompressed products (which was the appellant's business) was a significant element of the first respondent's intentions.
72 In mid-April 2008, the trade show of the National Association of Broadcasters was held in Las Vegas. Mr Petty, the first and second respondents and other senior personnel of the appellant attended. The trial judge said (at [55]):
At the show, a Canadian competitor of the applicant called "Matrox" announced that it would introduce a new 1lane card with 10bit HD capacity for use in conjunction with Macintosh computers. This was, it seems, an announcement of some importance, and the matter was discussed as between Mr Petty and Mr Overliese. Mr Overliese said (to Mr Petty) that it was not possible for the applicant to do 10bit with its DMA, "but that since Matrox claimed it, we could investigate it by changing the DMA". Mr Petty did not deny Mr Overliese's account of this conversation, but said that the environment at the show was not conducive to holding confidential discussions about future product ideas, so the issue was not progressed at that stage.
73 The trial judge dealt in some detail with the circumstances in which each of the first and second respondents left the employ of the appellant. I do not need to set out the details. However, two points should be noted. As far as the first respondent is concerned the trial judge found that at the time of his resignation the first respondent had in mind going into business with the second respondent with the third respondent as their corporate vehicle in the production and marketing of audio and video products. As far as the second respondent is concerned, the trial judge found that on 28 April 2008 (which happened to be the date of the first respondent's letter of resignation) he had no intention of resigning.
74 It is convenient at this point to summarise the trial judge's factual findings on the important issues. In order to do so it is necessary to have regard not only to matters set out above but also to his Honour's discussion of the appellant's claims and also his reasons for judgment as to costs.
75 I think his Honour found that during their employment with the appellant each of the first and second respondents undertook detailed preparations to establish a forthcoming business venture of which the third respondent would be the corporate vehicle.
76 In the second half of 2007 the business venture was to involve the making of electronic products in the audio industry. There is no complaint about the respondents' proposed involvement in the audio industry.
77 However, at some point, probably in early February 2008, both the first and second respondents formed an intention that the business would also involve making electronic products in the video industry ([46]). The first respondent used confidential information of the appellant in his preparations for the new business. The second respondent received but did not use the first respondent's workings based in part at least on the appellant's confidential information.
78 The first respondent formed an intention in early February 2008 that the third respondent would not only make electronic products in the video industry but products which would compete with the appellant's products and continued to hold that intention in the months which followed. The trial judge rejected the first respondent's evidence to the contrary ([49], [50], [52], [54]).
79 The trial judge's findings in relation to the second respondent's intentions are less clear. I think he found that the second respondent rejected the idea of competing with the appellant's product on 9 February 2008 ([35]) and despite some indications in his reasons to the contrary (see [32], [41], [50]), I do not think that the trial judge found that the second respondent ever formed an intention thereafter that the third respondent would compete with the appellant ([81]).