S.A.S. ELECTRONIC CO LTD v. LI MAN YUNG t/a MANFORD INDUSTRIAL CO

HCA4885/2001 &
HCA7631/2002

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

ACTION NO.4885 OF 2001

--------------------

--------------------

(HEARD TOGETHER)

----------------------

J U D G M E N T

----------------------

1. These two actions were tried together. The plaintiff in Action No.4885 of 2001 is Gothik Art Ltd, ("Gothik"). The plaintiff in Action No.763 of 2002 is S.A.S. Electronic Co. Ltd ("SAS"). The defendant in both actions is Li Man Yung trading as Manford Industrial Co. ("Mr Li").

2. The actions arose out of the manufacture by Gothik of a soft toy called a "Shakey Bug" for an American buyer. The toy contains an electronic battery-powered module, (the "module") which, when activated by touch or a loud handclap, produces a sound like manic childish laughter, coupled with a shaking motion. The toys produced as exhibits P6A and B produce two repeated cycles (the "cycles") of the same laughing sound, coupled with a continuous shaking motion, lasting in total for about nine seconds.

3. The module consists of a plastic cylinder 70 mm in diameter and 75 mm high which sits vertically in the toy. The base of the module is flat and contains a battery compartment. The top is curved and on it is fixed a sound sensor. This, when activated sets off an integrated circuit ("IC") contained inside the module. The IC is programmed to make an internal loudspeaker produce the laughing sound and at the same time to run an electric motor, which has an eccentric weight fixed to its shaft. The rotation of the weight produces the shaking effect.

4. The IC is a very small component, inside the module. In most of the exhibits it is carried on its own little plastic board less than 1 cm x 2 cm in size, stuck at right angles to a printed circuit board ("PCB") which carries other components such as transistors. The connectors for the IC are soldered to various points on the printed circuit. The IC is covered with a blob of black plastic referred to as "masking". The IC was also referred to in evidence as a "dice form".

5. Another electronic device which appears in this case is the EPROM which is a standard acronym for "Erasable Programmable Read Only Memory". In simple terms it is an electronic recording device which can contain computer data or a programme. The programme which is later fixed in the IC is first written on a computer and copied to an EPROM or series of EPROMs for demonstration before the IC is made. In fact the EPROM was also referred to in correspondence and then in evidence as an "E-Point"; it seems that this came about because of difficulties in pronunciation of the English acronym.

6. For demonstration of the laughing and shaking effect the EPROM may be connected to a module. Another device referred to in evidence is a "demo board"; this is a PCB with an EPROM on it and a battery box and loudspeaker stuck on to it, and a motor attached to it by wires. It demonstrates the operation of the sound cycle or cycles and the effect of turning the motor on and off; but it does not demonstrate the shaking action because there is no weight attached to the motor shaft.

7. I have gone to some length to describe the devices with which this case is concerned because what they are was by no means clear from the pleadings or the witness statements and only came out in the evidence, and indeed when I called for a screwdriver and opened up the modules to see what the witnesses were talking about. While I have no expert knowledge of the internal workings of a module or its electronic components inspection of what is inside them can throw some light on the reliability of the evidence.

8. Gothik's case against Mr Li is that it was agreed that he would first make a mould for the module case and then test samples of modules, conforming to samples provided by Gothik. Those samples produced two laugh cycles. Mr Li made the mould at an agreed price of $15,000.00 and produced some sample modules which produced two laugh cycles. Those were approved and on that basis the parties entered into a production contract dated 26 July 2001 for 100,000 modules at a total price of $380,000.00. The contract required Mr Li to make pre-production sample modules. When he delivered them it was found that they only produced one laugh cycle. Gothik rejected these and Mr Li agreed to adjust the modules to produce two laugh cycles. Gothik agreed to extend the time for delivery of the modules, but Mr Li did not deliver them in time. So Gothik had to engage another manufacturer to make its modules. Because it was a "rush job" it cost Gothik dear; in total $1,285,000.00. So it now claims against Mr Li for that sum, plus the price of Mr Li's mould and minus the original production cost made up of the mould and purchase orders. The final figure claimed is $900,500.00.

9. In Mr Li's pleaded defence he admitted the contract to make modules according to an approved sample. He produced nine sets of pre-production samples which Gothik rejected as unstable. He produced another nine sets which were delivered to Gothik by SAS, and Gothik rejected those as unstable too. A joint meeting was held between himself and representatives of Gothik, SAS and one Novatek Microelectronics Corporation ("Novatek") which had actually manufactured the ICs. At the meeting Gothik's representative refused to produce the original approved sample for comparison. Therefore Gothik is not entitled to claim as alleged or at all. Mr Li further pleads that Gothik refused to produce the approved sample for comparison and refused to accept that the ICs were made to its specification; further or alternatively the modules were produced to the plaintiff's specifications and the plaintiff refused to take delivery of them; and by its conduct Gothik evinced an intention no longer to be bound by the purchase order and repudiated the same, which repudiation Mr Li accepted. Accordingly Mr Li counterclaims for $380,000.00 being the agreed price of the modules which he says conformed to Gothik's specifications.

10. There were some attempts to amend the defence and counterclaim, which were disallowed by another judge prior to the trial. Further attempts were made in the course of the trial, as well as an attempt to put in a fresh witness statement by Mr Li. I will return to this later if necessary.

11. The claim of SAS, which commenced in the District Court, is for $60,329.65 being the price of 105,953 pieces of IC sold and delivered. Mr Li's defence is that the ICs delivered failed to comply with the specification of the IC as approved himself and/or Gothik. Originally he pleaded that specification was for two cycles of laughter and shaking but later the averment of "two cycles" was taken out. In any event, samples were agreed after initial rejection and a contract was made of the supply of ICs conform to sample. Pre-production samples were made and these were put into the modules, which Gothik rejected because they produced one cycle and not two. Despite repeated requests and demands SAS refused to rectify the pre-production samples and repudiated the contract by conduct which repudiation Mr Li accepted; so SAS is not entitled to claim. Further SAS had delivered ICs in excess of those ordered, i.e. 103,000 pieces. Mr Li counterclaims for his loss of profit plus the damages claimed against him by Gothik.

12. The contract between Gothik and Mr Li for the production of a mould for the module case is contained in a document dated 28 June 2001 and signed by the parties. It is for the making of a mould for $15,000.00, with 40% down payment, 30% on test and 30% 30 days after mould completion. It is not in dispute that Mr Li made the mould and that Gothik made the first two payments for it.

13. The contract between Gothik and Mr Li for the modules is contained in a document dated 26 July 2001. Under the description of the goods appear the words, in brackets, "as same as our approval sample". There is a series of "Remarks" of which the fourth reads "All merchandise must be manufactured in compliance with our approval samples". The contract provides for delivery of six pre-production samples on or before 27 August, delivery of 20,000 sets on 30 August and the balance by instalments, every three days until 9 September 2001, to Gothik's factory in China.

14. The contract between Mr Li and SAS is evidenced by the former's purchase order dated 31 July 2001 for 103,000 pieces of MOS NT 59003H-0-H083 for US$7,519.00. Underneath the order is written in Chinese "Laughing sounds and shaking of the IC have to strictly follow the sample". Attached to the order is a document of Novatek dated 1 August 2001 entitled "NT 59003H-0-H083 CHECKLIST" which contains inter alia a "checksum" 00155ABE(H) and some technical data including the entry "Output Pins 2 (pins)". This "checksum" is, as appears from the evidence, the sum of the data in the programme; it will identify the programme in the IC.

15. The order and the checklist bear to be signed and chopped by Mr Li and SAS. The order provides for delivery of 10 pre-production samples but does not give a delivery date. However there in an invoice by SAS to Mr Li dated 27 August 2001 for 105,953 pieces of NT 59003H-0-H083 at a price of US$7,734.57; and it is for the Hong Kong dollar equivalent of this sum that SAS claims.

16. The major factual issue in the case between Gothik and Mr Li is whether the "approval sample" referred to in the contract produced one cycle or two. The major issue between SAS and Mr Li is, again, what was ordered; an IC to make one cycle or two. It is necessary to examine the evidence in some detail because there is a dispute as to what was the sample referred to in the contract between Gothik and Mr Li. Further, it is common ground that neither of the samples on which the contracts were based - if indeed there were two different samples and not just one, which is itself in dispute - is now exhibited before the court.

17. The first dealings between Gothik and Mr Li were conducted by Gothik's managing director Mr Wong Kin Wai, Rugby ("Mr Wong"). He says that Gothik received an order dated 6 February 2001 for 100,000 sets of the "Shakey Bug" toy from a US customer, DS-Max. The toys were intended to reach the US in time for the Christmas market in 2001. (In fact the written order from DS-Max order does not contain a delivery date). He placed his order with Mr Li for the mould in April 2001 and showed him his customer's sample of a module. (Exhibit P1). This exhibit is in the form of a made-up module as described above and it produces two cycles. The price of the mould was agreed in early May.

18. Then on 19 May he delivered to Mr Li a "copy sound masking tape" obtained from his customer. This was produced as Exhibit P5. It is not a tape at all, but a battery box with an attached PCB and loudspeaker and a push button switch. On the PCB is what appears to be an IC, covered with black plastic, and some other components. When the switch is pushed the loudspeaker produces two laugh cycles.

19. Thereafter it appears the matter was handled primarily by Gothik's marketing manager, Ms Chow Ming Chu, Mimi ("Ms Chow"). Mr Wong also gave evidence of the signing of the contract and the steps taken by him in September 2001 to obtain modules from another maker; I will come back to this later.

20. Ms Chow was in fact the major witness for Gothik. She produced the customer sample P1. She said she could distinguish it from other modules later produced by Mr Li, by differences in the appearance of the on-off switch.

21. According to Ms Chow, on about 18 June 2001 Mr Li delivered three test sample modules to her. In her statement she called the "first test samples". These were not suitable because although they produced two laugh cycles, the shaking motion was a tremble rather than the desired circular motion. Ms Chow said that she asked Mr Li to adjust the motion to conform with P1, something which Mr Li said was a simple matter.

22. Ms Chow produced one of these samples as Exhibit P2. It is noted that P2 has inside it a motor, with a smaller eccentric weight than the others exhibited (which is no doubt why the shaking motion is more of a tremble) and loudspeaker, but outside it are a separate battery box, switch, sensor and a PCB measuring 60 mm x 45 mm. The PCB bears the name "Higher Way" and a telephone number. Soldered on to the circuitry of the PCB is a flat socket or holder and into that is plugged a flat 18-pin device which Mr Li calls an EPROM, and which looks like a slightly smaller version of an EPROM exhibited and referred to below, but which one of the witnesses of SAS says is a "packaged dice form"; in any event it is the device which contains the programme. It is not the same as the IC. There is no room inside the module for this PCB.

23. According to Ms Chow's statement, in about mid July 2001 Mr Li produced three more sample modules, i.e. the finished article with the IC inside. She referred to these as the "2nd test samples". Ms Chow was satisfied that the two laugh cycles and the shaking motion conformed to Gothik's customer's samples. The only problem was that the sensor was inside the case. Ms Chow wanted it outside, so that it would be more sensitive. One of these samples was produced as Exhibit P3.

24. From oral evidence, however, it appears that another single sample was handed to her some time in July. This is not referred to in the statement. It is the one referred to in correspondence as the "1st E-Point", i.e. 1st EPROM. She says that after P2 was delivered and before the 2nd test samples, of which P3 is an example were delivered, Mr Li brought her this 1st EPROM which "had all its parts outside". This was in effect a second version of P2; a module case with the shaking motor inside but a circuit board with an EPROM or other memory device outside. It had a good shaking motion; the shaking motion of P2 had been corrected but the sound was the same and had already been approved by the customer. Under cross-examination she said she approved it for the motion only, the sound having been approved earlier. She sent this one to her customer; and she has never retrieved it. The reason she gave was that in her trade this would not be done; she also said that if she asked the customer to give it back, the customer would think there was something wrong with Gothik, and that would affect Gothik's image.

25. About a day after delivering the "2nd test samples" Mr Li delivered three new sample modules with acceptable laugh cycles and shaking motion, and the sensor outside the case. She called them the "3rd test samples". According to Ms Chow, she orally approved these samples for production, subject to what DS-Max or its Hong Kong agent said. It appears from cross-examination that she asked for the shaking motion to be made stronger, after speaking to her customer; but this was not critical, and the 3rd test samples, i.e. the finished modules were the ones she approved. She sent two of the modules with the sensor inside i.e. the "2nd test samples" and one of the latest modules with the sensor outside, i.e. the "3rd test samples" to her customer; and the customer approved the 3rd test sample. She kept one of these 3rd test samples and returned one to Mr Li. According to Ms Chow it was on the basis of this "3rd test sample" that the parties entered into the contract.

26. On about 20 August, Mr Li delivered nine modules as pre-production samples. They were unstable. The laughing and shaking would either stop unexpectedly or keep on going. She complained and Mr Li took them away. Then on about 28 August he delivered another nine pre-production samples. But these only had one laugh cycle. Mr Li said he would adjust it to conform with the approved samples. But he did not.

27. On 30 August 2001, Mr Li asked for and was given one original customer's sample and the remaining 3rd test sample which was in Gothik's possession. There is in evidence a receipt for one "original module" and one "approval sample" but it is not clear who signed it. He said that because there were problems between him and SAS, he wanted the approved sample and the customer's sample; and that is what she gave him. She reminded him of the delivery schedule.

28. On 31 August, Mr Li requested Ms Chow to write a letter of complaint for him to show to his subcontractors. This she did. In her letter she complained that the pre-production samples only produced one cycle which was different from "our approval sample" saying that Gothik could not accept that, reminding him that the contract required full delivery by 9 September and calling for full delivery by 15 September at the latest. Mr Li asked for a letter in stronger terms.

29. On the same day, at Mr Li's request, she wrote another expanded letter on the same theme. In this letter she added the following :

"It's very RIDICULOUS to ask back the one and only EPROM sample which has been approved by our overseas customer already".

30. This, she says, is a reference to the missing "1st EPROM" which was delivered to the customer but never retrieved.

31. Then at Mr Li's request, on 1 September at Gothik's office there was a meeting attended by Ms Chow, Mr Li, Mr Terence Leung of SAS, and Mr Jacky Cheung and Mr Andy Fu, the latter two being employees of Sino Wealth Ltd which is a related company to SAS and Novatek, which actually made the ICs for SAS. At this meeting it became apparent that there was a dispute between Mr Li and SAS. The latter said that the ICs they had produced for him were in order; and the only way to find out who was at fault was to get back the missing EPROM. Mr Li wanted Ms Chow to get it back but she did not agree. According to her this was not necessary because at the meeting she produced one customer's sample (like Exhibit P1, and later produced as Exhibit D2) and one approved or 3rd test sample. Mr Fu said that another full set of ICs, producing two cycles could be made up in one or two weeks. But the question appears to have been, who was to pay for them; a matter which was, according to Madam Wong, between Mr Li and SAS.

32. Thereafter it appears that Mr Li tried to make arrangements to make modules which would produce two cycles. There was correspondence and further argument between the parties. Mr Li produced another module, which made two cycles on 17 September. This was later produced as Exhibit D1 by Mr Li. She wrote to him giving him until 20 September to fulfil the order, but he said he could not produce the modules in time; he said "Sue me, then". He sent her an e-mail on 20 September in effect blaming her for not returning the "1st EPROM" and so causing the delay. He said that if he received it, he could deliver the goods by 20 October. But this was too late; so Gothik went ahead and got another subcontractor to make its modules.

33. According to the statement of Mr Lawrence Gill (Chinese surname Koo) ("Mr Gill"), who was the assistant sales manager of SAS, in early July 2001 Mr Li gave him one sample of a device which was like P2 in appearance. It laughed and shook. Mr Li asked him to quote for the supply of an IC which would produce a similar effect. Mr Gill gave this device to Novatek and it produced a sample for inspection which he took to Mr Li who did not approve it. He got Novatek to make a second sample which again Mr Li did not approve. A third sample was made. This Mr Gill delivered to Mr Li near the Watson's store in Jordan Road, some time in late July. Mr Li was satisfied with it and said he would show it to his customer. A few days later he said the customer had approved it and he placed the order. So Mr Gill took the check list with the specifications on it for Mr Li to sign at his office and this Mr Li did. He told Mr Li that the "checksum" on the check list would be exactly the same as that in the sample which Mr Li had approved; and in fact the ICs were made with this checksum.

34. In oral evidence-in-chief, having adopted his statement, Mr Gill produced a "demo board" which was marked P1 (763). As indicated this is a PCB with batteries and an EPROM connected to a motor; it demonstrates the laughing sound, and the motor switching on and off but it does not shake. It produces only one cycle. He said that this was a sample of what he had given to Mr Li. He also said that what he handed over on each occasion had a module attached like P2, and the final sample handed over at the Watson's store was in the same form. SAS had on 25 August delivered 105,953 pieces of IC. The extra was provided because this was usually done, to make allowances for losses on testing. Usually the customer accepted this and indeed Mr Li did, on this occasion.

35. Mr Gill was not present at the meeting on 1 September. However, he knew that it had been suggested the Mr Li order a new batch of ICs; they could have been produced in about three weeks; but Mr Li would have been charged for them. He knew that the sample he had given Mr Li had been asked for; but Mr Li did not give it back. So SAS could not check if that had complied with the checksum written on the checklist that went with the contract.

36. In answer to my own questions, Mr Gill said that in each case he gave Mr Li a sample with an EPROM containing the programme. The programme would end up in the IC. It was not possible to make a one-cycle IC run twice by modifying the external circuitry; the programme had to be rewritten. The IC itself could not be re-programmed, once made.

37. From cross-examination it appears that Mr Gill was somewhat uncertain about the form of the sample handed over at the Watson's store. He said that it did not have a module connected; then he said that he could not remember if it had one. He said that there was a motor because the testing effect should be similar to the product to be handed to Mr Li. Then he said that there was no eccentric weight attached to the motor. He also said that the original sample (like P2) which he had received from Mr Li had been given to Novatek but later returned to Mr Li. He could not remember if it had a "Higher Way" PCB attached. Nor could he remember if it produced one cycle or two. It was put to him that he had explained to Mr Li that the "two pins" on the checklist meant that the IC would produce two cycles; but this he denied.

38. Mr Fu Man Kwong, Andy ("Mr Fu"), was the sales manager of Sino Wealth Micro-electronics Corporation Ltd which was connected to Novatek. His evidence was that he received a sample from Mr Gill and was asked to simulate the effect. He did not pay attention to the number of laugh cycles. In any event, Novatek did the simulation and produced a demo board like P1 (763) with an EPROM. The sound was modified twice at Mr Li's request through Mr Gill and ultimately the programme was produced with the checksum which appears on the checklist attached to the contract between SAS and Mr Li. In terms of this contract Novatek produced the ICs with the same programme as was in the EPROM on the demo board. In fact he produced, in court, a colleague with a computer who plugged in the EPROM and demonstrated the checksum on the screen. It also appeared from cross-examination that this demo board P1 (763) was exactly that, a demonstration model made up for demonstration in court, with the programme in the EPROM copied from the original in a Novatek computer.

39. Mr Fu did not know if any pre-production IC samples had been given to Mr Li after the signing of the contract, and before the production run, but he was sure that none would be in existence before 1 August when the contract was signed. At that stage everything was done by reference to the EPROM in which the programme was contained; only later was that built into the "dice form" ICs.

40. According to Mr Fu, at the meeting on 1 September he was asking for return of the EPROM that had been given to Mr Li in order to resolve the dispute between SAS and Mr Li. Mr Li in turn said that he had given it to Gothik; and Ms Chow said that she "might not be able to get it back". There was a suggestion that he could manufacture a fresh lot of ICs which would produce two cycles; but no concrete response was given. He could have done this, without having the EPROM back; he could have made up the whole batch of 100,000 in two or three weeks; or his company could, with extra effort, make them up in one or two weeks.

41. Mr Li's evidence was that in April 2001, Mr Wong asked him to make a mould for the module and design the items to be put inside. He said it had to laugh and shake and at that stage it had to have a "3-second IC". At that stage he did not have a sample of the sound; in cross-examination he said that the module which Mr Wong showed him quacked like a duck. Later, Mr Peter Wong, also of Gothik, gave him a large (about 6 inches by 4 inches) demo board with an EPROM, a motor and a sensor and asked him to develop a sample following it. He did not receive any "masking tape" in the form of P5. He took the demo board which he received to a company called ADD who simulated the effect. But it was a 5-second programme and cost too much for Mr Peter Wong. So Mr Li took his ADD demo board to another company, Higher Way. The latter produced three samples in the form of P2, but without the module. He made up P2 in June using a Higher Way board and a module but Gothik wanted it to swing with a stronger force. This was in late June or early July. Gothik returned one sample of P2 and he took it to Higher Way but was told that the IC could not be modified to produce a stronger swing. Indeed, it appears it could not; because what produces the swing is the eccentric weight and that on P2 is visibly smaller, and the swing is different, from that on the other exhibits.

42. Mr Li said that, since he could not get what he wanted from Higher Way, when Mr Gill came to see him he gave Mr Gill a Higher Way sample like P2 and told him to copy it but make the movement faster. Now P2 produces two laugh cycles. I asked Mr Li about the sample which he had given Mr Gill; he said he did not pay attention to whether it produced one cycle or two. I asked him if the sound lasted for 8 seconds or so; and he said that it was only a "3-second IC" and there was a repetition of it. In any event, he said, it was the same effect as P2.

43. Mr Li said that Mr Gill later brought back a sample which effectively did not work at all. He took it away and came back with another one, which started but did not stop. He took that one away for modification. Then in late July Gothik pressed Mr Li for delivery so he pressed SAS. Mr Gill did indeed meet him near the Watson's store at 8 a.m. one morning and gave him a sample; it was like P2, with the sensor and battery box outside the module; and it had attached an SAS baseboard with an EPROM. He did not test this sample with Mr Gill but on the same morning he took it to Ms Chow and she approved it. So he told her to keep it. He also told her to revise the contract which had already been faxed to him on 26 July; he wanted clause 5 of the "remarks" , which provided him to produce an extra 1% of goods free of charge, taken out. This was done and ultimately he signed the contract but after he had signed his contract with SAS for the ICs.

44. Now according to Mr Li, the sample which Mr Gill delivered to him and which Ms Chow approved was not in the form of a demo board like P1 (763); it was like P2, but with an SAS board rather than a Higher Way board, a module which shook and a sensor which activated the programme. So this, according to Mr Li, is the sample on which the contract was based. However in cross-examination he said that he did not know whether it produced one cycle or two.

45. Then, Mr Li says, he had a meeting with Mr Gill and a Mr Terence Leung of SAS at which he signed the checklist. The contract with SAS was signed later. In fact the serial numbers on it had been changed following a telephone call from SAS. At the meeting they had explained to him that the two "pins" referred to on the checklist meant two triggers; in other words each pin would make the module produce one cycle. He felt ambiguous about the explanation so he added the words to the contract, that the laughing and shaking had to follow the sample; and according to him the sample that was meant was the Higher Way sample which he had originally given to SAS.

46. According to Mr Li SAS gave him nine pre-production ICs on 20 August. Then he gave Gothik (at its factory in China) nine pre-production sample modules, containing the SAS dice form ICs. This was after he had signed the contract with Gothik; before that he had delivered four EPROMs in all. He disagreed with Ms Chow's evidence that he had given her Exhibit P3; by reference to its components, he said that was impossible. In particular it had a 20 mm sensor inside, which he did not use and it had a motor with a label on it dated 28 August 2001. Nor had he given her what she called the "3rd test samples". She had indeed returned to him two samples on 30 August but these were a Higher Way sample like P2 and a finished module with a SAS dice form inside and the sensor outside, like P4. She had complained that she wanted two cycles, like the P2 equivalent, but the module with the SAS dice form was not stable; sometimes it did one cycle, sometimes two and sometimes it went on without stopping. Those were what had been signed for, by someone from his company, on the receipt produced by Gothik. He had passed them along to SAS for modification and Mr Leung of SAS had signed a similar receipt for them. But Mr Leung told him that the 2nd "pin" had not been made, on the batch of ICs which SAS had been produced.

47. At the meeting on 1 September, according to Mr Li, Mr Leung had the Higher Way sample like P2 and the one-cycle module which had been passed on to him earlier. Madam Wong produced one of her customer's sample modules, which he produced as Exhibit D2. This was given to Mr Leung so that SAS could produce another EPROM for Novatek to make a new batch of 2-cycle ICs, along with the Higher Way sample like P2; he took them away to see if a new set of ICs could be made.

48. It appears that at this meeting, Madam Wong took the view that it was up to Mr Li and SAS to sort matters out. The SAS and Novatek representatives were prepared to make up a new batch of 2-cycle ICs which they said could be done in two or three weeks; but they wanted to be paid for that work. They wanted the EPROM or demo board back, which had been given to Mr Gill and which according to Mr Li was then passed on to Madam Wong. But she would not get it back from her customer.

49. Later, according to Mr Li, Mr Leung of SAS told him that Mr Gill was in effect blaming Mr Fu of Novatek for getting the number of cycles wrong. Mr Li told Mr Leung just to go ahead and make a new EPROM, pending a decision as to who had got the number of cycles wrong; but Mr Leung said that it was not good, and Novatek were refusing to make up a new batch of ICs. So Mr Leung left the Higher Way sample like P2 and the customer sample D2 in a bag, hanging on Mr Li's office doorknob.

50. Thereafter Mr Li tried to get Ms Chow to retrieve the EPROM which she had approved; but she kept telling him that that was impossible. He tried to get another company to make up a new set of ICs and modules for him. He got one produced by one Cheong Lee Fat, i.e. Exhibit D1, which he took to Ms Chow on 17 September. He says that he could indeed have had the modules made up by 20 October as he said in his letter of 20 September. But, in effect, that was no use to Gothik.

51. There are, it seems to me, all sorts of contradictions and inadequacies in the evidence of the parties. Looking first at the evidence for Gothik, it is in the first place difficult to understand where the customer samples, Exhibits P1 and D2, came from. If Gothik had these in the first place, there would have been no need for Mr Li to be supplied with a different device to demonstrate the sound required; be that Mr Wong's Exhibit P5 or the large PCB referred to by Mr Li.

52. In the second place, although Ms Chow says that Mr Li gave her Exhibit P3 some time in July, this cannot be right. Internal examination of P3 shows that it has a label on the motor with the date 28 August 2001. There is no reason why this should be anything but the date of manufacture. P3 shares with Exhibit D2, which made its first appearance on 17 September 2001, several characteristics namely the date on the motor, the use of a 20 mm internal sensor, the PCB marked CB0102A and the IC on a separate PCB, connected with wires, rather than on a small board stuck on at right angles to the main PCB. I do not believe that this P3 was in existence in July 2001.

53. The third point relates to the missing sample. Whether this is the "3rd test sample" of Ms Chow, i.e. a complete module with the sensor outside, or the "1st EPROM", i.e. a module with an EPROM on an external PCB is important. But there is no sensible explanation from Ms Chow as to why it cannot be produced or what is stopping her from retrieving it from DS-Max. If her evidence of the "3rd test sample" is true, production of it would end the matter. If what she agreed to was the "1st EPROM" and it produced two cycles, that too would probably end the matter. Any embarrassment which Gothik felt in having to approach its customer would be far outweighed by the sum sued for. Further, if the evidence of the "3rd test sample" is true, there should be even less difficulty in laying hands on the "1st EPROM" because it would have been superseded; it had no longer any value as between Gothik and DS-Max so there should have been no problem in getting the latter to hand it over.

54. Then we come to Mr Li. His evidence was rambling. His statements, it appeared, had never been read over to him; so I had him give evidence-in-chief orally. Indeed his case had changed since the case began; though given what I said earlier, about not knowing what the case was until the evidence was heard the exhibits seen, this is perhaps not surprising.

55. Mr Li's evidence included the rather fanciful and indeed somewhat incoherent story about the meaning of the two "pins". It seems that he wanted two pins, not because he knew that Gothik wanted two cycles, but because sometimes a customer would want one and sometimes two. So Mr Gill told him that the two pins would produce two cycles. Then, it appears, only one pin was made. Now I have no technical or expert evidence as to what the pins actually do. Obviously a layman's guess must be that they are for connection of the IC to the rest of the circuitry, rather than something very abstruse and technical. But whatever they were the story of them did not surface until Mr Li, late in the day, attempted to amend his pleadings, put in new statements and so on. If one pin was missing, and Mr Li thought that was why the ICs would produce only one cycle and not two, that would have been known before the meeting on 1 September, from the ICs already delivered. Mr Li would surely have been protesting at that meeting that SAS had failed to give him the two pins he had contracted for. But there is no evidence from anyone that this question was raised.

56. There are various other points which may be made. For instance, he insisted that on any PCB which his company made up there would only be three transistors (which he identified), whereas P3 had four. But so has P4, the pre-production sample made by his company. So this is clearly wrong.

57. Then we have Mr Gill. He was unclear about whether what he was given as a sample to copy made one cycle or two. So, indeed, was Mr Li and even Mr Fu. It seems that what he got laughed and shook; it was like the Higher Way sample, P2. But he did not know what he had given back to Mr Li; he was inconsistent about whether it shook or not. Then he said that the motor had no eccentric weight; but there would be no reason for him not to return the module he was given albeit with a different EPROM attached containing the new programme. That is what he said in oral evidence-in-chief and it is also the sequence of events which Mr Fu describes. Generally I had the impression that Mr Gill was trying to tell the truth as he remembered it, but on some points he was confused.

58. Mr Fu however I considered to be a straightforward and credible witness. What he said made sense. Further, he was one step back from the parties; he worked for a related company, rather than directly for SAS. He seemed to me to be a technician who was trying to do his job properly and to tell the truth as he remembered it.

59. It is of course now argued that Mr Li is bound by his original pleadings, no matter what evidence may have come out. In particular it is argued that Mr Li cannot now go back on a general admission of the first paragraphs of the Statement of Claim which apparently include an admission that the various samples produced by him before the contract was signed in fact produced two laugh cycles. As I have indicated, neither the pleadings nor the witness statements were clear. They were all to do with "samples" of one kind or another. It was not until the evidence came out that I had some idea of what the parties were really talking about in factual terms. I took the view that the pleadings as they stood were of little assistance; they did not do what pleadings are supposed to do, as set out by Bokhary JA in Aktieselskabet Dansk Skibsfinansiering v. Wheelock Marden & Co. Ltd [1994] 2 HKC 264 at 269-270. In the end, however, the new version of Mr Li's case was properly put before me and identified, so that Gothik knew what the his case was, and what it had to meet. Further Gothik had a full opportunity to deal with this new case. It is therefore open to me to make findings based on this new case. See Poon Hau Kei v. Hsin Chong Construction Co. Ltd, CACV 167 of 2002.

60. What is clear is that the ICs which Novatek made for SAS produced only one cycle. I accept the evidence of Mr Gill and Mr Fu, for it makes sense, that what was finally produced and given to Mr Gill, was an EPROM containing a programme identified by a specific checksum, and that is the same programme which is in the ICs; i.e. a programme which produced one cycle. I also accept, because it makes sense, that Novatek would have produced no ICs before 1 August 2001; for there would be no point in making ICs until the contract between Mr Li and SAS was finalised. Before that, any sample passing between Novatek, SAS and Mr Li would have been in the form of an EPROM. (I do not think the demo board, Exhibit P1 (763) takes us anywhere; it was made up after the event, for the court proceedings, using the original programme).

61. So far as Mr Fu was concerned the object of Novatek's work was to simulate as far as possible the effect produced by the device passed to it by Mr Gill. No one seems to have paid much attention to whether that device made one cycle or two. It probably did, because on Mr Li's evidence, which Mr Gill does not contradict, for he cannot remember, it included a Higher Way PCB like P2; and P2 produces two cycles. Clearly somewhere along the line, between Mr Li, Mr Gill and Novatek, what has come to be produced is the EPROM containing the programme which only produces one cycle. That is the programme with "checksum" shown on the checklist appended to the contract between Mr Li and SAS.

62. That the difference between one and two cycles should have been blurred is perhaps not so unlikely as it first sounds, given that it is possible for a module to work inconsistently, in particular depending on how it is handled. For instance Exhibit D2 when first produced to Ms Chow in cross-examination showed a tendency to run without stopping even though, if put on a carpet and held down, it would only run for two cycles. Further it has to be noted that there is nothing in the order from DS-Max to Gothik, the contract between Gothik and Mr Li, or the contract between Mr Li and SAS specifying in writing the number of cycles. Nor was there any evidence that anyone instructed orally that the module, or the IC be made to produce a specific number of cycles. One wonders whether anyone was really paying attention to the number of cycles, before the pre-production sample modules were delivered to Gothik.

63. As between Mr Li and SAS it is clear that Mr Li contracted for a batch of ICs containing a particular checksum. He says that the reference to a sample was a reference to the Higher Way original. Mr Gill says that he told Mr Li that what would be produced would have the same programme as the checksum on the documents. It seems to me that by the time the contract was written, any sample referred to would necessarily be the latest one produced by Novatek for SAS; for the original Higher Way sample had been imitated and the imitation modified, in order to get to that stage. As to the matter of the two pins, I have dealt with this above; I do not believe Mr Li's assertions that he was told that the two pins would produce two cycles.

64. It seems, therefore that SAS produced exactly what Mr Li contracted for and he cannot complain now that the ICs do not produce two cycles.

65. As between Gothik and Mr Li, the question is whether the sample on which the contract is based is Ms Chow's "third approved sample", i.e. a working module with nothing outside but the sensor, stuck on the top, or the "1st EPROM" which according to Mr Li was what he received from Mr Gill at the Watson's store in Jordan Road.

66. It is difficult to see how the sample could have been Ms Chow's "third approved sample". There is no room inside a module to take either of the external PCBs exhibited, be it that forming part of P2 which has something very like an EPROM on it, or that forming part of P1 (763) which has an EPROM on it. If Ms Chow's "third approved sample" existed it must have contained an IC, and not an EPROM. But before the contract was signed between Mr Li and SAS, there were no SAS or Novatek ICs. So if the "third approved sample" existed it must have contained an IC made by someone else. It is argued that Mr Li could have got one from some other supplier; he had been dealing, before this, with Higher Way and ADD. That may be so but it would not make sense to make a contract based on a sample with someone else's IC in it, when, as we know, there is no way to transfer the programme from one IC to another; it has to be imitated by starting from scratch and writing a new programme. On the other hand, Ms Chow agrees that she was given a "1st EPROM", which she says did everything she required of it. The difference is that according to her it was given to her earlier in the sequence of events.

67. I am dealing here with witnesses of whose reliability I am not satisfied. I am particularly unsatisfied with the failure by Gothik to produce, in court or to the other parties, the either the "1st EPROM" or the sample on which she claims to rely, as well as her production of P3 which cannot be what she claims. For the reasons I have given I do not find Mr Li reliable either. I can only decide who is the least unreliable on any given point by looking at the evidence against the background of the sequence of events and the inherent probabilities in the surrounding facts and circumstances. Looking at the evidence against the background, I cannot be satisfied that Ms Chow's "third test samples" existed. It seems to me much more likely that what she accepted as a sample on which to base the contract was the missing "1st EPROM" which Mr Gill passed to Mr Li and Mr Li passed on to her. It is more probable, it seems to me, that that EPROM contained the programme which was ultimately put into the ICs, i.e. a programme which produced one cycle.

68. It follows that Gothik did not have the right to reject the pre-production samples produced by Mr Li on 28 August. They were what it had contracted for by reference to the sample on which the contract was based, namely the "1st EPROM".

69. Gothik by its refusal repudiated the contract in advance of performance. But the matter does not end there, because Mr Li did not simply accept the repudiation; presumably because he did not know whether the EPROM produced one cycle or two. Certainly as between himself and SAS, he later claimed through solicitors that the EPROM had produced two cycles. In any event he tried, or appeared to try to give Gothik what it wanted. It must be borne in mind that by this stage Gothik had extended the final delivery date to 15 September 2001. Mr Li kept the contract alive. According to him, he had by this time got all the module components together ready for assembly. He sought to get a new batch of ICs made up by SAS but this plan foundered because Novatek refused to make them. So he went off and tried to get others to make them, ultimately finding Cheong Lee Fat who made up Exhibit D2.

70. I have difficulty in understanding why, if Mr Li was prepared to get another set of ICs made up, he did not simply have SAS or Novatek do it. It was a matter of money, at that stage; the dispute between Mr Li and SAS could have been sorted out later. Mr Fu said in cross-examination that Mr Li simply did not contact him; though, from his evidence, he could have made up a new batch, without getting the EPROM back, and in something like two weeks; and there would be no difference in price between a one-cycle IC and a two-cycle IC. As I have indicated I found Mr Fu a credible witness and where there is a conflict between him and Mr Li I prefer the evidence of Mr Fu.

71. It is argued that notwithstanding his failure to accept the repudiation, Mr Li was absolved from his obligation to perform the contract in accordance with its terms, despite his efforts to comply with it. Miss Yeung who appears for Mr Li relies on Braithwaite v. Foreign Hardwood Co. [1904] 2 KB 543. She also relies on the following passage in Chitty on Contracts, 28th Edition, 25-011 :

"Finally, where the repudiating party stipulates for a mode of performance which is at variance with the terms of the contract and the innocent party attempts to comply with the new stipulation, the repudiating party cannot rely on a failure by the innocent party to perform his original obligations where that failure is attributable to his attempt to comply with the fresh stipulation"

72. Mr Wong, counsel for Gothik, argues that there is no halfway house. A repudiation is either accepted or it is not. He relies on the following dictum of Lord Ackner in Fercometal v. Mediterranean Shipping Co. [1989] 1 AC 788 at 805 :

"When A wrongfully repudiates his contractual obligations in anticipation of the time for their performance, he presents the innocent party B with two choices. He may either affirm the contract by treating it as still in force or he may treat it as finally and conclusively discharged. There is no third choice, as a sort of via media, to affirm the contract and yet to be absolved from tendering further performance unless and until A gives reasonable notice that he is once again able and willing to perform."

73. It seems to me that Mr Li could have accepted the repudiation. But he did not. He kept the contract alive on the basis that, whether the EPROM contained a programme for one cycle or two, he would deliver to Gothik 100,000 modules which produced two cycles. He either accepted that the contract was for two-cycle modules, or he accepted a fresh stipulation for two cycles. He could, it appears, have got new ICs in a couple of weeks and produced the modules. No doubt it would have cost him the price of a new set of ICs but that could have remained as a matter for litigation between himself and SAS. But he did not. I do not see that it can be said that his failure to comply with the fresh stipulation, if there was a fresh stipulation caused him to be unable to deliver the modules by the extended time limit especially since that was extended again later until 20 September 2001. He is therefore liable for his failure to deliver.

74. Counsel attacks the quantum of Gothik's damages which are admittedly large; something like four times what Mr Li agreed to charge. According to Mr Wong, he had to pay this high price to one Wah Fat Wai Ming to make a new set of modules, because it was a "rush job". This was necessary to get the Shakey Bugs shipped to the foreign buyer in time for the pre-Christmas sales. It is argued that the loss should be viewed with suspicion. There is no precise delivery date in the DS-Max order, nor any evidence that Gothik tried to get its customer to accept later delivery. There are no other quotations; we do not know if the price was reasonable.

75. The problem is, of course, that these Shakey Bug modules were not something which could be bought "off the shelf". Gothik had to get them made up, starting from scratch. There is no evidence to show that in doing what it did, Gothik did not act reasonably. Nor is there any evidence that it could have got the modules made up more cheaply. There is in effect no evidence on which I could make any alternative finding.

76. It follows that in Action No. 4885 of 2001 Gothik's claim for $900,500.00 succeeds and Mr Li's counterclaim fails. There will be judgment for the plaintiff for $900,500.00 with interest thereon from the date of the writ to the date of judgment at prime rate plus 1% and from the date of judgment until payment at the judgment rate and costs to be taxed if not agreed. The defendant's counterclaim is dismissed with costs to be taxed if not agreed.

77. In Action No. 763 of 2002 the claim of the plaintiff SAS likewise succeeds and the counterclaim of the defendant Mr Li fails. There will be judgment the plaintiff for $60,329.65.00 with interest thereon from the date of the writ to the date of judgment at prime rate plus 1% and from the date of judgment until payment at the judgment rate and costs (on the District Court scale) to be taxed if not agreed. The defendant's counterclaim is dismissed with costs (on the High Court scale) to be taxed if not agreed.

78. Since the judgment is to be handed down the costs orders are nisi.