Computer Expert Witness
Graham Dilloway CITP MBCS
Computer Expert Witness

Chartered IT Professional and Member of British Computer Society

Listed in Register of Expert Witnesses


Member of the Academy of Experts


View Graham Dilloway's profile on LinkedIn

Internet Connection Problems

The Claimant, Ms C, bought a computer from the Defendant, Company D.  Ms C subsequently experienced problems when using the computer to connect to the Internet.  The problems included slow connection speed and lengthy periods of inactivity.

My IT expert witness report discusses the symptoms of the problem and the possible causes.  The various attempted remedial actions are also discussed and conclusions drawn from these actions.

Note:   This expert witness report is reproduced exactly as it was when served excepting that company and personal names have been removed.

Personal

1. This report was prepared by Graham Dilloway of 39 Conham Hill Bristol BS15 3AW. I am a Member of the British Computer Society, the chartered professional body for the computer industry in the UK. I am a member of the Academy of Experts and of the Expert Witness Institute. I have worked with computers for 30 years. This work has all involved the implementation and configuration of computers, their operating systems and the core software applications of a computer environment (e.g. word processors and spreadsheets). I have worked with personal computers almost exclusively for more than fifteen years.

Instructions

2. My instructions are contained in an email from Solicitor S dated 24 July 2002:
" Please can you provide your view as computer expert on the following:-
1. On the balance of probabilities, the likely cause of the problems which Mrs C alleged she experienced.
2. If it is very unclear as to the cause of her alleged problems, please give details of the possible causes.
3. If, on the balance of probabilities, the issue involved the computer system, please state whether you consider that, on the balance of probabilities, this indicated that the computer was of unsatisfactory quality."

3. This report is based upon information contained in the bundle of documents sent with the email from Solicitor S dated 24 July 2002. The email describes these documents as:
"1. Extract of communications between Solicitor S and Mrs C,  copy letter from Mrs C to me in response to the Defence.
2. Copy communications between Mrs C and Company D.
3. Copy statements of case."

Background Information Regarding Internet Connections

4. The telephone network that is used by most people to connect to the Internet from home was designed to support voice calls. Modems are used to convert the data generated by computers into audible tones that can be carried by the voice network. The tones can sometimes be heard from the modem at the beginning of a call and are similar to the tones that are sometimes heard from a fax machine.

5. The quality of the voice network is variable. It is, for example, sometimes necessary to ask at the beginning of a voice call for the other person to speak more loudly and to sometimes ask for some piece of speech to be repeated.

6. Audible tones are exchanged by the modems at each end of the connection when an Internet call begins. Each modem listens to the tones that it receives from the modem at the other end of the connection. The modems negotiate the speed of the connection based upon the "quality" of the tones that they receive.

7. A poor connection causes the modems to negotiate a connection that is slower than is possible with a good connection. A slow connection is the modem equivalent of a voice call between two people who need to speak loudly to be heard by each other.

8. There are mechanisms in the hardware and software used in Internet connections that allow portions of the data to be retransmitted when the data becomes lost or garbled. Retransmitting data is the modem equivalent of asking for some piece of speech to be repeated during a voice call.

9. The theoretical maximum speed of the modems currently in common use is "56k". This is 56,000 bits per second where a bit is a single piece of binary data. I cannot recall that I have ever heard of a modem that is able to achieve this theoretical maximum. I would expect a good quality telephone line to achieve a speed of 40,000 bits per second or better and a poor quality telephone line to achieve a speed of less than 40,000 bits per second.

10. My own previous experience serves as an example of the effect of a poor telephone line. About two years ago, the quality of voice calls on my telephone line was poor especially during wet weather. I cannot now recall the speed at which my modem connected but I do recall that I was disappointed that it was so slow. I recall that I was sometimes unable to connect to the Internet and sometimes the connection was the victim of frequent interruptions and delays. I was aware that I was not able to report a fault because voice calls were usually audible and that BT are contracted to provide an acceptable voice service and not a data service for a standard line as described by Ms C in her email of 20 December 2000.

11. Eventually, the difficulties with voice calls became so bad that I asked BT to resolve the problem. I recall that BT engineers visited twice and that the telephone cable between my house and some point in the street was replaced. I recall that the length of cable that was replaced spanned several telegraph poles.

12. Both the quality of voice calls and the reliability of Internet calls improved after the telephone cable had been replaced. A typical connection speed would then be about 40,000 bits per second.

13. The symptoms that I experienced during Internet calls before the replacement of the telephone cable were similar to the symptoms described by Ms C and, as I recall, the connection speed that I experienced during Internet calls before the replacement of the telephone cable was similar to the connection speed described by Ms C.

Ms C’s Internet Connection Problems

14. The exchange of emails between Ms C and Company D describe symptoms of the problem that include:

Date & Time

Sender

Symptom

2 Aug 2000 19.21

Ms C

"I have removed the splitter and it seemed to make a difference in that during that time I was not thrown out but as the problem is intermittent anyway I do not know if that did the trick …"

3 Aug 2000 11.37

Company D

"… the modem appears to stay connected at times …"

14 Sep 2000 14.45

Ms C

"Yes, it does disconnect after one minute. Normally however, the disconnection problem seems to have been solved. Now the big remaining problem is that the modem stops downloading for long periods of time …"

29 Sep 2000 14.51

Ms C

"… the computer connects at 312000 at the moment …" and "… when it runs, it seems to download quite well and fast. It just seems to take these long breaks without any communication going on …"

15 Dec 2000 16.09

Ms C

"… I have also changed my ISP with no improvement …"

18 Dec 2000 11.57

Ms C

"… I did try my computer on a different telephone line and the difference is enormous …" and "… a BT engineer tells me they need to change the card in the exchange and increase the gain, to make it compatible to the Internet …"

15. It is my opinion that the symptoms described above are consistent with a poor telephone connection being the cause of the problem. None of these symptoms indicate that the computer was at fault.

16. Elsewhere, the correspondence includes:

Date & Time

Sender

Symptom

11 Oct 2000 07.08

Ms C

"… I never had a problem with the modem or telephone line with my previous (Company D) computer …"

27 Aug 2001 15.37

Ms C

"… I took my computer and external ADSL modem to my son and tried both on his ADSL line. The same result: 4 minutes good surfing, 2 minutes coffee break when the computer no longer takes in new data …"

17. It is my opinion that, while the specific symptoms described above are not consistent with a fault on the telephone line, nor do they provide positive evidence that the computer was at fault.

Attempts at Rectification

18. In about October 2000, the entire computer was replaced except for the hard disk and the problem persisted. I have no doubt that a fault on the hard drive would have produced symptoms when the computer was used for something other than Internet access. I cannot conceive of any way that a hard drive fault would produce the Internet access problems that Ms C describes. This suggests that the source of the problem is external to the computer.

19. Prior to December 2000, Ms C attempted Internet connections via more than one ISP and the problem persisted. This suggests that the source of the problem was either in the telephone network between the computer and the ISP or in the computer.

20. Prior to August 2001, Ms C began using an external ADSL modem rather than the internal modem supplied with the computer and the problem persisted. This suggests that the source of the problem was not the modem supplied with the computer.

21. Prior to September 2001, Ms C obtained a new telephone number and exchange and the problem persisted. This suggests that the source of the problem was in the telephone cabling local to Ms C’s house or in the computer.

22. Prior to December 2001, Ms C attempted a connection using a second telephone line that was available at her house (as described in the letter to Solicitor S dated 10 December 2001). Initially, a good result was achieved but there was deterioration with one hour. This suggests that the source of the problem was in the telephone cabling local to Ms C’s house or in the computer.

23. Prior to August 2001, Ms C reformatted the hard drive and re-installed the Windows operating system. This suggests that the source of the problem was not the software configuration on the computer.

24. None of the remedies that were attempted eliminates the possibility that the telephone line local to Ms C’s house was at fault. The attempted remedies that involved changes to the computer virtually eliminate the possibility that the fault was with the computer.

Conclusion

25. The majority of the symptoms of the problem and the results of most of the attempts at rectification support a conclusion that the source of the problem was not in the computer and was probably in the telephone network local to Ms C's house.

26. I am unable to draw a firm conclusion because of the conflicting evidence but my opinion is that the source of the problem is most likely to be in the telephone network local to Mrs C's house. The key reasons for this conclusion are:

Summary

29. On the balance of probabilities, the source of the problem with Internet connections is not the computer.

30. The evidence is contradictory and a more definite conclusion might be reached by further work such as examination and testing of the computer.

31. I understand my duty to the Court and I confirm that I have complied with and will continue to comply with that duty.

32. I confirm that insofar as the facts stated in my report are within my own knowledge I have made it clear which they are and I believe them to be true, and that the opinions I have expressed represent my true and complete professional opinion.

 

Graham Dilloway
Expert Witness

27 July 2002

39 Conham Hill

Bristol BS15 3AW