Goetz Trading Ltd v. Pacific Supernet Ltd
Read the full judgment text of DCCJ 5427/2002 on BabelCite. This District Court judgment was delivered on 11 October 2004.
1. The plaintiff is a trading company in the clothing trade. It uses electronic mail, or e-mail extensively to communicate with the companies to which it sells goods, and the factories in the region from which the goods come. The defendant was its internet service provider (“ISP”) and had since 1997 provided internet services which permitted the plaintiff to communicate by e-mail.
Cited by 1 case
|
DCCJ 5427/2002 IN THE DISTRICT COURT OF THE HONG KONG SPECIAL ADMINISTRATIVE REGION CIVIL ACTION NO. 5427 OF 2002 __________ BETWEEN
__________ Before: His Honour Judge Muttrie in Court Date of Trial: 20 – 22 September 2004 Date of Handing Down of Judgment: 11 October 2004 JUDGMENT 1.The plaintiff is a trading company in the clothing trade. It uses electronic mail, or e-mail extensively to communicate with the companies to which it sells goods, and the factories in the region from which the goods come. The defendant was its internet service provider (“ISP”) and had since 1997 provided internet services which permitted the plaintiff to communicate by e-mail. 2.In May 2001 the services were upgraded to “broadband”, a more rapid communication system and the parties entered into a new contract for the supply of services (the “Agreement”). In July 2001 other customers of the defendant complained of receiving unsolicited advertising e-mails, known as “spam”. The defendant identified the e-mails as coming from the plaintiff’s server, i.e. the main computer of the plaintiff’s internal computer network within its offices through which e-mails from the computers on that network were sent. The defendant blocked the transmission of e-mail messages from the plaintiff’s server from 20 to 26 July 2001. The spam however continued to be transmitted and the defendant continued to receive complaints of it. The plaintiff denied and denies sending it. In fact no one seems to have been able to establish with certainty where the spam was coming from. The plaintiff ceased to pay the defendant’s service charges and sought recompense for its costs in attempting to establish the origin of the spam. Ultimately in December 2001 the plaintiff ceased to use the defendant’s service and changed to another ISP. 3.Now the plaintiff claims that the suspension of service was in breach of its contract with the defendant, and claims damages for such breach. The defendant counterclaims for unpaid service charges for the whole of the contract period of 12 months from 15 May 2001. 4.The Agreement contained the following relevant terms and conditions:
5.In July 2001 the defendant received complaints from various of its customers that they had received spam. From a file of electronic copies of the complaints produced by the defendant’s witness it appears that they commenced about 9 or 10 July. The spam e-mails offered to sell cheap printer cartridges. They bore to come from an e-mail address [email protected] and to solicit replies to [email protected] (the “nn” and “mm” stand for two digits which changed from time to time). Investigations showed that all the e-mails appeared to have come from, or through the plaintiff’s internet server, which was identified by a series of numbers known as an “IP address”; that address was 202.64.56.18. I will deal with the technical matters in more detail below. 6.Spam is the curse of the internet, as is well known to all who use e-mail; at this time that is a large percentage of companies and individuals in Hong Kong. The defendant had procedures in place to deal with spam complaints. Those procedures were in accordance with a Code of Practice published by the Hong Kong Internet Service Providers Association. In accordance with its procedures the defendant between 11 and 19 July 2001 sent to the plaintiff four e-mails. They were headed respectively “First Warning of Spamming”, “Second Warning of Spamming”, “Third Warning of Spamming” and “202.64.56.18 will be blocked by us”. They all read as follows:
7.All these e-mails were sent to the e-mail address [email protected]. In fact in its application for the service and in the agreement itself, the plaintiff had identified as its e-mail address for administrative contact, billing contact and technical contact a different e-mail address namely [email protected]. The plaintiff’s staff did not monitor [email protected]; the result being that the defendant’s e-mails went unanswered; and that is why the defendant suspended the service. What was suspended was in fact the ability to send e-mails; the plaintiff could still receive them. 8.According to the plaintiff’s managing director, Mr Kent, it took some time for the plaintiff to realise that the e-mails it was sending out were not reaching their destinations. Mr Kent was not clear as to when that was realised but said that when it was, the plaintiff’s programmer Raymond Ng was given the task of solving the problem, and then an outside company, Net Solutions Asia Ltd. “(NSA”) was called in. Time sheets show that NSA’s engineer, Ms Psyche Leung, started work on the problem on the afternoon of 23 July. Once it was realised that e-mails were not going out, the plaintiff had to resort to other means of communication including fax messages and telephone calls. 9.It is not in dispute that the defendant restored the plaintiff’s outgoing e-mail service on 26 July 2001. According to the defendant’s senior systems engineer, Mr Poon, someone from the plaintiff telephoned the defendant on that day and said that the problem of transmission of spam had been fixed. 10.Unfortunately the spam problem did not go away. The defendant’s electronic copies of the complaints files show that the spam complaints continued in August. In fact they continued into November 2001 and thereafter seem to have tailed off. Also the spammer latterly went in for advertising other commodities than ink cartridges. However no one, be it the plaintiff, NSA or the defendant seems to have been able to find the origin of the spam. 11.Mr Kent of the plaintiff on 17 August wrote to the defendant complaining about the suspension and seeking assistance. The defendant replied on 24 August suggesting that the e-mail be routed through the defendant’s own server in the hope of checking the origin of the spam; but it appears that this was not desirable to the plaintiff, which told the defendant that it would seek outside assistance (which it had already done) and invoiced the defendant for NSA’s charges of $23,500.00. Correspondence continued, becoming increasingly terse but no resolution was reached. In the meantime the spam continued to go out and the complaints to come in to the defendant, which made further threats to block the plaintiff’s e-mail service. The plaintiff stopped paying the defendant’s monthly service charges of $2,800 from August onwards and the defendant sent payment reminders. 12.In early December 2001 the parties discussed settlement, but they could not agree. In that month the plaintiff ceased to use the defendant’s services, and obtained services from another ISP. In July 2002 the defendant instituted proceedings in the Small Claims Tribunal for the service charges unpaid by the plaintiff during the contract period. The plaintiff commenced proceedings for damages in this court in August 2002. 13.It is as well to set out part of the plaintiff’s pleadings in its Statement of Claim:
14.Particulars of loss, damage and expense follow. 15.So the plaintiff’s claim is for damages for breach of contract in suspending the outgoing e-mail service. The claim that this amounted to repudiation of the contract is quite properly abandoned. The claim for breach of contract and/or duty in terminating the services and/or repudiating the Agreement in December 2001 is not abandoned, so far as I can see, but it cannot be made out; I will return to this point later. 16.There is no doubt that the defendant suspended the outgoing e-mail service. If it did so without justification, that would be breach of contract. Justification might be found in the fact that the plaintiff was in breach of Clause 6.3 by which it warranted and undertook not to use the defendant’s services to copy, reproduce, distribute, publish or otherwise transmit unsolicited advertising or promotional material, i.e. spam. So there are two questions, the first being whether the spam was coming from the plaintiff’s server. If it was not, there is no question of breach of Clause 6.3. If it was then the next question is whether that necessarily means that the plaintiff was in breach of Clause 6.3. If it was not there could be no justification for the suspension. If it was, then the question remains whether the defendant was in suspending the service without notifying the plaintiff that it would do so or had done so. 17.The parties are agreed that spam could have come from the plaintiff’s server in four ways. The first is “open relay”, whereby spam from an outside transmitter is relayed by the server; it can be stopped by the use of an anti-spam programme which will prevent the spam from coming into the server in the first place. The second is hacking whereby an outside spammer obtains control of the server. The parties have agreed that neither of these in fact applied. The third way is wrongful use by someone within the plaintiff’s network of the service, to transmit the spam through the plaintiff’s e-mail server. The fourth way is “IP forgery”; i.e. the sending of messages by a third party which appear to come from a particular IP address when in fact they do not. Much of the trial was taken up with discussion by expert witnesses of whether this latter way is possible; for if it was not possible it would follow, by elimination, that there was wrongful use by the plaintiff or someone in its organisation. 18.So far as the factual evidence is concerned, Mr Kent says that the plaintiff was not sending spam. Its business has nothing to do with printer ink and there is no reason for it to advertise printer ink. Further no one in the plaintiff – which has a small office network of computers – was sending spam either. 19.The e-mail programme used by the plaintiff was Lotus Notes and the server programme is Lotus Domino. According to the plaintiff’s expert, Mr Meyer, spam from the plaintiff would have to go through the Lotus server and would appear in a message log. A message log for 8 August, a date on which spam messages were sent (the complaints invariably include the messages complained of) shows no traffic which could be identified as spam. 20.Ms Psyche Leung’s report, which Mr Meyer adopted shows that the spam was sent by an e-mail programme called Eudora. As I understand it, if Eudora was installed in the plaintiff’s network or server, the spam would not go through the Lotus log so it would go undetected. But, says Mr Meyer, NSA had installed the software when the broadband services started in May 2001. There was no Eudora programme on the plaintiff’s network installed then and none found when investigations were carried out later. The only person who could have installed such a programme would be the plaintiff’s programmer Mr Ng; he was the only person in the plaintiff’s organisation who had the required “administrator privileges”; in other words if anyone else tried to install it, the computer would refuse to accept the installation. Mr Ng has not given evidence but according to Mr Kent, he would not have sent the spam. 21.It follows from the above that, if spam was sent through the plaintiff’s server it must have been sent using another e-mail programme such as Eudora and that programme would have to be installed and then uninstalled after sending to avoid detection. I am not clear, and it was not canvassed whether someone could bring in a portable computer containing Eudora and plug it into the plaintiff’s router (an electronic relay between the plaintiff’s server and the telephone line), in parallel with or in substitution for the plaintiff’s server. This is relevant to the expert evidence below; but it would still mean spamming by someone with access to the plaintiff’s premises. 22.Turning to the expert evidence, there is no dispute that it is possible for anyone to send an e-mail which contains in its header a false e-mail address, e.g. [email protected]; because the sender sets the address, as one who sends a letter by post may write a return address on the envelope. But the argument is whether an e-mail can be sent, the header of which contains a false IP address. The plaintiff’s expert, Mr Meyer of NSA, says that it is possible. The defendant’s expert Mr Chan says that it is not, or practically not. 23.As I understand it, Mr Chan’s reasoning is as follows. A computer can only send an e-mail message to another computer if it has first established a connection. It is possible for another computer, which we may call the false computer to transmit a message, identifying itself by the plaintiff’s IP address, 202.64.56.18 to the intended recipient of the e-mail, be that the ultimate recipient or a server somewhere along the way. The false identification goes out as a synchronising or “SYN” packet (the information transmitted by a computer is said to be sent as “packets”) to the other computer. The recipient computer must send back an acknowledgment or “ACK” packet, which the transmitting computer must receive in order to establish the connection. Such is the architecture of the path of the signal through relays called “routers” that the ACK packet will go to the real owner of the IP address, i.e. the plaintiff’s server, rather than the false computer. The analogy used is that of the return address on the postal envelope; even if the envelope is not sent from that address, that is the address to which the postman will return it. The real owner of the IP address will discard the ACK packet because it has not itself sent a SYN packet. 24.Mr Chan says further that only if the false computer uses an IP address which has been discontinued, or exists but is not yet in use, or if it is connected to the same router as the real owner of the IP address will the ACK packet reach it. The plaintiff’s server connects to the defendant in the first instance through a router in its own office and for the false computer to receive the ACK message and make the connection it would have to be connected directly to that router, i.e. in the plaintiff’s office if not on the plaintiff’s internal network. As I have indicated this would entail the spammer having access to the plaintiff’s office. 25.I find it very difficult to accept that it is impossible for a false computer to pass itself off as the plaintiff’s server and send out spam. The system described has been in operation for many years; the defendant’s expert referred to a paper on it written in 1982. Given that once any technology is invented, someone will set out to subvert it to dishonest purposes it seems that there has been ample time for spammers to find a way to do this. But I have to make a finding on the evidence. Mr Meyer admits that he is not an expert on the IP protocol and although he tells me that a spammer can send spam from a forged IP address, he cannot tell me how it is done. Mr Chan however claims to be an expert on this protocol and he can explain why it cannot be done. I have to accept his opinion as being more probably correct and that an IP address cannot be forged. It would follow by elimination that the spam must be taken as having probably come from the plaintiff’s server or perhaps through another computer attached directly to the plaintiff’s router. 26.This is not, however, the end of the story. One has to go back to the Agreement. The defendant could only rely on Clause 7 to suspend the outgoing e-mail service if the plaintiff was in breach of Clause 6.3, i.e. if the plaintiff had used the services to “copy, reproduce distribute, publish or otherwise transmit” spam.Now Mr Chan, Counsel for the defendant argued his case on the basis that the defendant had a pretty well absolute right of suspension, if spam was coming from the plaintiff’s server; and there would be no need to give notice of suspension or even to inform the plaintiff when the suspension took effect. There was some discussion of amending the pleadings to include an implied term, but ultimately I refused the plaintiff’s application for amendment. So it is now said that if spam is found to have been coming from the plaintiff’s server, the defendant could suspend the service without breaching the contract, so the plaintiff has no claim. 27.On further consideration of Clause 6.3, I do not think that this is right. I do not think that we need to dip into the murky waters of implied terms, although in fact as I have noted above there is some reference, if an inadequate one, in the pleadings to an implied term. I think the answer can be found in the interpretation of the contract as it stands, following the principle restated by Lord Hoffmann in Investors Compensation Scheme Ltd v. West Bromwich Building Society [1998] 1 WLR 896 at 912 that
28.By Clause 6.3 the plaintiff “warranted and undertook” that it would not use the services provided by the defendant to transmit spam. That undertaking would obviously cover the direct use of the services by the plaintiff. It would by extension cover the authorised use of them by a servant of the plaintiff. Further clause 6.4, which I have not reproduced, puts a duty on the plaintiff to control the use of the services by persons within its organisation; if there is more than one permitted user the provisions of the Agreement are to be brought to the attention of all, and the plaintiff must “procure that each permitted user complies with the provisions of this Agreement.” 29.I cannot see how the provisions of Clause 6.3 could be taken as a guarantee by the plaintiff against use by an unauthorised person such as a hacker or someone taking advantage of an “open relay”. Everyone knows, and it must have been known to the parties at the time when they entered into the Agreement that hackers exist and may try to take control of other persons’ computers for their own purposes. The defendant certainly knew of the “open relay” problem; its automatic complaint response message contained the following:
30.Further “open relay” seems to have been the first suggestion made by the defendant to the plaintiff as the possible explanation for the spam, if the plaintiff was not itself transmitting spam. And Mr Kent himself seems to have thought, as he put it in correspondence, that “most 12-year-olds can fake an e-mail header”; at any rate, as he explained, he knew that the sender’s name could be faked, even if he knew nothing about the IP address. 31.The point is that at the time of entering into the contract the parties knew or must have known of the possibility that spam could be transmitted which appeared to come from a particular server when in fact it did not. It would be impossible for the parties, in the light of that knowledge, to have agreed that the plaintiff would accept liability for spam of that nature. It could only accept liability for spam transmitted through its own or an employee’s use of the services. At most it could accept liability for an employee’s wrongful use of the services if the employee had had the requirements of the Agreement, including the prohibition against spam, brought to his notice. 32.Mr Kent of course says that the spam did not come from the plaintiff. I accept this. There would be no reason for the plaintiff to send spam. Nor would there be any reason to authorise an employee to send it. There is no evidence that the plaintiff brought the requirement not to send spam to the notice of all its employees who sent e-mails, or procured them not to send e-mails; but as Mr Meyer has said, it was a small open office with about ten people working in a large room. Certainly after 26 July they would all have known, if anyone had not known before, that spam must not be sent; but the spam went on anyway. So even if the plaintiff never actually directly told its that they must not send spam, it is obvious that that made no difference; if the spammer was an employee he would have sent spam anyway with or without such a warning. If, and no direct accusation is made, the culprit was Mr Ng the programmer he would in any event, by the nature of his work, have known that he must not send spam. 33.It follows that even if the spam was coming from the plaintiff’s server or a computer attached to its router it does not necessarily follow that the plaintiff was in breach of Clause 6.3 because of it. The culprit could have been an unauthorised employee on a frolic of his own. 34.It is also clear that, given that the effect of Clause 6.3 was that the defendant would only have the right either to terminate the Agreement on notice under Clause 9.3 or to suspend the services without notice under Clause 7.1, it could not simply suspend the services as soon as it knew that spam was apparently coming from the plaintiff’s server. It could not assume that the plaintiff was necessarily in breach of Clause 6.3. It would have to investigate first and eliminate the other possibilities. Obviously the defendant recognised that, because it sent out the warning e-mails before suspending. If it had sent them to the correct address, and got no response no doubt it would have been justified in disconnecting. But it sent them to the wrong address, rather than the address which the plaintiff had specified for technical and administrative contact. It is not the plaintiff’s fault that the e-mails went unread. When the defendant got no response it would have been easy enough to pick up the telephone, or send someone to the plaintiff’s premises, but of course no one did; as so often happens, procedures were slavishly followed and no one used any common sense. 35.It follows that I cannot be satisfied, and neither could the defendant, that the plaintiff was in breach of Clause 6.3; and that the defendant was in breach of contract in suspending the service in the way it did and must be liable for the plaintiff’s losses which flow directly from the breach. 36.The plaintiff also claims that in breach of contract and/or duty the defendant terminated the services and/or repudiated the Agreement in December 2001. Now from 26 July up until that time, there was no interference with the services notwithstanding that the spam continued. There were threats to disconnect; once again a procedure was being slavishly followed and e-mails sent out without any consideration of the underlying problem. Given that, according to Mr Chan the defendant’s expert, the plaintiff was paying extra for a fixed IP address, unlike the home user who commonly has a new IP address assigned to him every time he connects to the ISP’s server, one would have thought that it should be easy enough to give the plaintiff a new fixed IP address. If that had been done, and the spam continued, it might have shown definitely that the spam was coming from the plaintiff’s server. But there is no evidence about that. In any event, the services continued without interruption and the plaintiff did not pay for them, from August onwards. In December, according to Mr Poon, the defendant suspended the service. But it is clear from the plaintiff’s evidence that that was done against the background of the plaintiff’s simply taking its business to another ISP because it was not satisfied with the way the defendant was handling the matter, or with the settlement offered. I do not see that the defendant can be said to have breached or repudiated the Agreement by suspending the service in December. So no damages can be claimed under this head. 37.Those losses flowing directly from the breach in July 2001 are, it appears, quite small. They could only relate to the costs of finding out what the problem was, because if the defendant had sent the warning e-mails to the right address the plaintiff would have known what the problem was. It would then have had to take action, through its progammer Mr Ng and if necessary through NSA to try to find the origin of the spam. So the costs of finding out why the e-mails did not go out could be said to flow from the defendant’s breach; but not those of tracing the origin of the spam. Again, no one seems to have used common sense and done what seems to me the obvious thing which would no doubt commend itself to any home computer user who found his e-mails were not going out, i.e. telephone the ISP. But I have heard no evidence as to why this was not done; there may be technical reasons for it. So I think it would be unfair to say, on this basis, that no damages at all should be awarded. 38.A major problem with the question of quantum is that the plaintiff’s evidence, except for what it paid to NSA and what it estimates as the labour costs of Mr. Ng’s work, is vague and not particularised. Mr Kent estimates the losses to the plaintiff at $100,000.00 to $150,000.00 but this, he says is a “ball park” figure. He cannot say how he arrives at it. It covers not simply the losses arising out of the defendant’s failure to get the message through that the service would be suspended, but all the extra time which Mr Kent says that he and his employees have had to spend on the problem between July and December 2001. But neither the time of Mr Kent nor of any employee except Mr Ng is costed. 39.While there is evidence that the plaintiff had to use other means of communication with its suppliers and clients because it could not use e-mail for a period there is no evidence of any loss of profits arising from that. While the employees might have had to spend more time using such other means of communication there is no evidence of overtime payments or anything of that sort. 40.The plaintiff’s solicitor argues that the plaintiff should have damages for distress and inconvenience and relies on a dictum of Bingham LJ in Watts v Morrow [1991] 1 WLR 1421 at 1445 that
41.It is argued that the provision of e-mail provides pleasure, relaxation, peace of mind etc. I do not think that is right. It is simply a business tool. In any event His Lordship made this remark in the context of other remarks that “a contract-breaker is not in general liable for any distress, frustration, anxiety, displeasure, vexation, tension or aggravation which his breach of contract may cause to the innocent party” and that damages in general are “recoverable for physical inconvenience and discomfort caused by the breach and mental suffering directly related to that inconvenience and discomfort.”. 42.Here there is no question of mental suffering. The plaintiff is a limited company and has “neither a body to be kicked nor a soul to be damned”; a more pithy rendering of a view expressed by Lord Chancellor Thurlow in the 18th century. Mr Kent might have suffered mentally, but he is not the plaintiff. However there was no doubt inconvenience to the plaintiff over the six days in that alternative means of communication had to be used and this would be analogous to the physical inconvenience which an individual might have suffered if, on losing his e-mail functions he had to resort to writing letters and posting them, using messengers, telephoning and making notes of his telephone conversations and the like. So I think the plaintiff can have some damages under this head. They would include the actual extra time spent on the problem by a particular employee such as Mr Ng. 43.The plaintiff particularises the time spent by Mr. Ng on the problem as 68 hours. The original calculation that that was worth $24,000 or more does not hold up in the light of the evidence of Mr. Ng’s actual salary. He was paid the equivalent of about $88.64 per hour. The total could only be about $6,000. But this, it appears, was by no means only related to the work done in trying to find out why the service was suspended; it includes, according to Mr Kent other work which was done alongside Ms Leung’s work later. Even if he had to spend all his time on the problem on Friday 20 July 2001 (for he worked Monday to Friday) and on 23 to 26 July (and that is a generous estimate) that would amount to five days at 8.25 hours per day or a total $3,656.40. 44.So far as the extra work of the other employees is concerned I propose to allow $2,000.00 per day for the five days, or $10,000.00. 45.NSA’s charges of $23,500.00 are particularised. It appears Ms Leung spent a total of 16.41 hours on the problem of inability to send e-mail between 23 and 26 July 2001. The charge is $1,000.00 per hour. So of NSA’s total charges the plaintiff could only recover $16,410.00. 46.On the basis of the above I calculate the damages flowing from the defendant’s breach of contract at $3,656.40 + $10,000.00 + $16,410.00 = $30,068.40. There is a claim for an indemnity against future liability to third parties, but I cannot see any basis for that. Any such claim arising from the suspension in July 2001 would have surfaced by now. 47.I turn to the counterclaim. It is not in dispute that the plaintiff did not pay the defendant’s charges from August onwards. As I have indicated the service continued without interruption after 26 July. There was no basis for the plaintiff not to pay the defendant’s fees thereafter. The fact that the defendant was demanding payment and threatening further suspension would not provide such a basis no matter how annoying it was. 48.The contract provided, of course, for a fixed term of 12 months. The evidence is only that the defendant suspended the service in December 2001. If each IP address is globally unique as the witnesses say, the number of them must be finite and decreasing. Why the defendant would not take the address, when the plaintiff had obviously ceased to use it, and sell it to another client, thereby to mitigate its loss has not been canvassed. The simple fact is that the parties contracted for a fixed term of 12 months and, no matter how unfortunate it seems, the plaintiff must pay for that. There is no dispute as to the figure of the outstanding charges which is $31,216.00. 49.There was also a claim for return of the router held by the plaintiff on loan but that has been returned in the course of the trial. 50.The conclusion is that the plaintiff is entitled to, and will have judgment against the defendant for $30,068.40 with interest from the date of the Writ until the date of payment at the judgment rate. The defendant is entitled to and will have judgment against the plaintiff on its counterclaim for $31,216.00. The plaintiff seeks contractual interest at 1.5% per month pursuant to Clause 4.3. This also is contracted for and so the plaintiff will have interest pursuant to Clause 4.3 up to the date of judgment and thereafter at the judgment rate. 51.So far as costs are concerned, each party is awarded a figure which is within the jurisdiction of the Small Claims Tribunal. No costs are awarded in that Tribunal. No doubt this litigation could have been avoided if both sides had taken a more sensible and practical approach than they did in the first place. In the circumstances it seems to me proper that each side should bear its own costs and therefore there will be no order for costs. That includes the reserved costs in respect of the argument about amendment.
Mr. Damien Laracy of Messrs. Laracy Gall for the Plaintiff. Mr. Jeremy Chan instructed by Messrs. Lee Chan Cheng for the Defendant. |
Other judgments that cite this case
Further hearings and rulings under DCCJ 5427/2002