1 Tuesday, 25 June 2024 2 (9.45 am) 3 MR BEER: Good morning, sir, can you see and hear us? 4 SIR WYN WILLIAMS: Yes, thank you very much. 5 MR BEER: Thank you. May I call Gareth Jenkins, please? 6 SIR WYN WILLIAMS: Yes. 7 GARETH IDRIS JENKINS (sworn) 8 Questioned by MR BEER 9 MR BEER: Good morning, Mr Jenkins. My name is Jason Beer 10 and I ask questions on behalf of the Inquiry. Can you 11 give us your full name, please? 12 A. My name is Gareth Idris Jenkins. 13 SIR WYN WILLIAMS: Mr Beer, before you go any further, is it 14 appropriate for me to direct Mr Jenkins about 15 self-incrimination? 16 MR BEER: Yes, sir. 17 SIR WYN WILLIAMS: Mr Jenkins, I am sure you've been advised 18 but, under our law, a witness at a public inquiry has 19 the right to decline to answer a question put to him by 20 any of the lawyers involved in the Inquiry, or by me, if 21 there is that the answer to that question would 22 incriminate you. This legal principle is known in 23 shorthand form as the privilege against 24 self-incrimination. I've decided that fairness demands 25 that I remind you of that principle before you give your 1 1 evidence. I must stress to you, however, that it is for 2 you to make it clear to me that, in respect of any 3 question put to you, it is your wish to rely upon the 4 privilege. If, therefore, any questions are put to you 5 by anyone who is asking you questions which you do not 6 wish to answer, on the ground that to answer such 7 a question might incriminate you, you must tell me 8 immediately after such question is put. At that point, 9 I will consider your objection to answering the question 10 and thereafter rule upon whether your objection should 11 be upheld. 12 I know that you are legally represented. No doubt, 13 if the issue relating to self-incrimination arises, your 14 lawyers will assist you and, if at any stage during the 15 questioning, you wish to consult your lawyers about the 16 privilege, you must tell me so that I can consider 17 whether that is appropriate. Do you understand all 18 that, Mr Jenkins? 19 A. Yes, I understand, sir. 20 SIR WYN WILLIAMS: Thank you. 21 Yes, Mr Beer. 22 MR BEER: Thank you, sir. 23 Mr Jenkins, you've made five witness statements to 24 the Inquiry. Thank you very much for making those 25 witness statements and for attending today and over the 2 1 next three days to give evidence. Can we start, please, 2 with your witness statements. 3 A. Yes. 4 Q. If we can start with the fifth witness statement, 5 please, which was, in fact, a witness statement you made 6 only yesterday. It's WITN00460500. 7 A. Yes. 8 Q. Have you got that witness statement in front of you? 9 A. I have indeed. 10 Q. It's three pages long and it consists of three pages of 11 corrections to your previous witness statements; is that 12 right? 13 A. That's correct, sir. 14 Q. Is that your signature on the third page? 15 A. It is. 16 Q. Are the contents of that, your fifth witness, true to 17 the best of your knowledge and belief? 18 A. They are. 19 Q. That. Can we go to your first witness statement, then, 20 please, which is WITN00460100. It's dated 6 February 21 2023 and is 15 pages long. 22 A. Yes. 23 Q. Have you got that? 24 A. I have got that. 25 Q. Is your signature on page 15? 3 1 A. It is. 2 Q. Are the contents of that witness statement true to the 3 best of your knowledge and belief? 4 A. They are. 5 Q. Thank you. Can we go to your second witness statement, 6 please, WITN00460200, dated 1 June 2023, which is 7 67 pages long; is that your signature on page 67? 8 A. Yes, it's actually slightly longer because it's got 9 an appendix at the end but that's my signature on 10 page 67, yes, sir. 11 Q. I'm only dealing with the body of the statement rather 12 than the additions at the end, which are exhibit sheets, 13 essentially? 14 A. Right, yes. 15 Q. With the corrections you made in statement 5 brought 16 into account, are the contents of that, your second 17 witness statement, true to the best of your knowledge 18 and belief? 19 A. Yes, they are. 20 Q. Thank you. Can we deal with your third witness 21 statement, please, 21 March 2024, WITN00460300, 234 22 pages long. 23 A. Yes. 24 Q. Is that your signature on page 234? 25 A. It is. 4 1 Q. With the corrections in statement 5 brought into 2 account, are the contents of that witness statement true 3 to the best of your knowledge and belief? 4 A. Yes, they are. 5 Q. Thank you. Then, lastly, statement 4, 29 April 2024, 6 WITN00460400, 83 pages long. Is that your signature on 7 page 83 of the witness statement? 8 A. It is. 9 Q. With the corrections in statement 5 brought into 10 account, are the contents of that, your fourth witness 11 statement, true to the best of your knowledge and 12 belief? 13 A. They are. 14 Q. Thank you very much. 15 They can be put to one side now. Any further 16 references I give you will come up on the screen. 17 A. Okay. 18 Q. Those witness statements, all five of them, will be 19 uploaded to the Inquiry's website. They are 20 a substantial body of evidence and, therefore, I'm not 21 going to refer to all parts of them when asking you 22 questions. Additionally, you've exhibited to your 23 witness statement a very large number of documents and 24 you have been provided by the Inquiry with still further 25 documents over the last year and a half, and you and 5 1 your legal representatives have had access to the 2 Inquiry's Relativity database. I think all of that has 3 resulted in you seeing a very large volume of material; 4 is that right? 5 A. Indeed. 6 Q. If, at any time, you want me to stop or to slow down 7 when considering a document, then please do say so. 8 A. Thank you. 9 Q. Can I start with your background, please. I think you 10 graduated in 1973 from Cambridge University with 11 a degree in mathematics? 12 A. That's correct. 13 Q. Upon graduation, you immediately worked for 14 International Computers Limited, ICL; is that right? 15 A. That's correct. 16 Q. ICL became Fujitsu Services Limited in 2002? 17 A. Around then. I'm not sure exactly when that happened 18 but, yes, it changed its name in the early part of the 19 century. 20 Q. I was taking that from your witness statement. 21 A. Yes, I wasn't 100 per cent sure of the date but that's 22 the best of my recollection. 23 Q. You remained there for the whole of your professional 24 career, is that right -- 25 A. That's correct. 6 1 Q. -- until you retired in 2015? 2 A. Correct. 3 Q. Dealing briefly at the moment with the stages of your 4 career at ICL Fujitsu, so far as concerns us, from about 5 1996 to about 2000 did you work in the agent team? 6 A. That's correct, yes. 7 Q. You were one of a team of 10 to 20 architects working on 8 the Horizon system; is that right? 9 A. Something like that, yes. 10 Q. From about 2000 until retirement in 2015, you worked in 11 the Architecture and Design Team; is that right? 12 A. That's right, yes. 13 Q. That was part of the Development Team; is that right? 14 A. It varied. At some points it was, at other points it 15 was a separate team, but the work was similar in those 16 respects. I -- 17 Q. It was part of the Development Team, I think, until it 18 merged with what was called the Requirements Department; 19 is that right? 20 A. Yes. 21 Q. How many people were in the Architecture and Design 22 Team? 23 A. I can't remember the exact numbers but, again, it varied 24 a lot. It started off with about 20 or 30, I think I've 25 got the figure in my witness statement, and, by the time 7 1 I retired, it was down to something like about half 2 a dozen, that sort of number. 3 Q. Were they all the same level or pay grade as you or was 4 there seniority within the department? 5 A. Different people were at different grades but that 6 wasn't necessarily reflected in terms of a management 7 structure as such. 8 Q. Throughout your time, I think your role was a technical 9 or operational one, is that right -- 10 A. Correct, yes. 11 Q. -- not a managerial or supervisory role; is that right? 12 A. No, I never had any man-management responsibilities 13 during that time. 14 Q. You tell us in your witness statement that from 2002 to 15 2008 you "took on the role of defining the technical 16 changes required by Project IMPACT and worked on the 17 implementation of the Bureau de Change and the 18 acceptance of credit cards"; is that right? 19 A. Among other things. So those were the main 20 developments. So I worked on IMPACT until that went 21 live in 2005, I think it was, then I worked on Bureau de 22 Change and credit card changes and a few other minor 23 things before I moved on to Horizon Online Requirements 24 Capture. 25 Q. Did you move on to Horizon Online from about 2008? 8 1 A. It was around then, yes. I can't remember the exact 2 time. 3 Q. Did you work on Horizon Online from 2008 to about 2015? 4 A. Yes. 5 Q. You tell us in your witness statement that you "worked 6 with Post Office analysts in defining the technical 7 requirements for Horizon Online"; is that right? 8 A. Yes, that's true. 9 Q. And that you "supported Post Office's requirements team 10 in specifying use cases for the counter functionality"? 11 A. That's correct, yes. 12 Q. What does that mean? 13 A. The contractual definition of how Horizon Online was to 14 work was that the counters should look -- should operate 15 in the same way as it did on Legacy Horizon and, 16 therefore, it was decided that we needed to put in place 17 something that actually defined exactly what that was, 18 and the technique that was decided was that Post Office 19 would put together the use cases to define the various 20 activities that went on in a Post Office to define the 21 steps that people went through for doing -- from how do 22 you sell a stamp to how do you balance at the end of the 23 week or whatever. So -- 24 Q. Sorry to interrupt you, that's a use case, is it? 25 A. Each one of those would be a separate use case, yes. 9 1 Sorry. 2 Q. Did you then develop the implementation of those use 3 cases and then test them? 4 A. Not personally. There was a team who were developing 5 the counter, and I was acting as a guide to that team in 6 terms of interpreting what the use cases meant and 7 looking at how the code that they developed came out, 8 and I was responding to when queries came out of the 9 testing process as to "Should it work this way or should 10 it work that way", "Is this a problem or is that how 11 it's supposed to work", and reacting to things like 12 that. 13 Q. You tell us that, during your time at Fujitsu, you were 14 a group of anywhere between 20 and 100 software 15 designers and developers who would provide fourth line 16 support for both Legacy Horizon and Horizon Online? 17 A. Yes, so the way fourth line support worked was that the 18 people who were doing the development were expected to 19 spend a small part of their time in actually supporting 20 the live service and the system under test, and so 21 that's what fourth line support was all about, and I did 22 that just like everyone else in the Development Team. 23 Q. That was an adjunct to your main role? 24 A. Correct. 25 Q. Although I'm sure the fraction varied, what proportion 10 1 of your time was given over to fourth line support? 2 A. It was supposed to average out at about 10 per cent and 3 I think that's roughly what it ended up as. 4 Q. Thank you. 5 In your first witness statement, you refer to ICL 6 introducing what's called a Distinguished Engineer 7 Scheme, correct? 8 A. Yes. 9 Q. You tell us that you became a distinguished engineer in 10 the mid-1990s? 11 A. Yes, I can't remember exactly when. 12 Q. You were one of about 100 other ICL employees who were 13 described as Distinguished Engineer; is that right? 14 A. That's correct, yes. 15 Q. Is it right that was an honorific title conferred upon 16 you by ICL and Fujitsu? 17 A. Yes, it was. It didn't make any difference to what 18 I got paid, or anything like that, it was purely seen as 19 an honour and just recognition of that I had a fairly 20 senior design role and was respected within the company. 21 Q. What was the process that led to you becoming 22 a Distinguished Engineer? 23 A. I think I was nominated by my manager who filled in 24 various complicated forms to justify why I should become 25 one. 11 1 Q. Was there any further training or qualification or 2 a particular professional experience that was required 3 before you could would be required before you could 4 become a Distinguished Engineer? 5 A. Not, as such, no. 6 Q. So it was simply a title given to you by your own 7 company? 8 A. Yes, correct. 9 Q. Did you acquire any level of management responsibility 10 as a result of the appointment? 11 A. No, people did try and persuade me that I ought to 12 manage some people and I resisted it and successfully. 13 Q. I think you've told us that it didn't affect your pay? 14 A. Correct. 15 Q. You tell us in your first witness statement that, from 16 about 2015 until August 2022, you were on a retainer 17 with Fujitsu as a consultant; is that right? 18 A. That's correct. 19 Q. What was the nature of your consultancy with them? What 20 were you consulting about? 21 A. It was various things. I -- the expectation was that it 22 would amount to something like about half a dozen or 23 a dozen days a year and, for most of the period, it was 24 at that level. I don't think -- although I was on the 25 retainer until August 2022, I think the last time 12 1 I actually did any paid consultancy was just before 2 Covid lockdown in 2020, which is why they decided that, 3 since they hadn't been using me for a couple of years, 4 that they didn't need to retain me any further. The one 5 exception to that was the time when the Group Litigation 6 was happening in 2018/2019, when I was working a lot 7 more than that, probably an average of about one to two 8 days a week, for a period of about five or six months. 9 Q. So did your retainer essentially come to an end at the 10 time of the publication of the Horizon Issues judgment, 11 which was December '19? 12 A. No, the retainer came to an end in 2022 but I didn't 13 actually do much retained work after -- 14 Q. Okay. 15 A. The last few things were actually talking to the lawyers 16 about the consequences of the Horizon Issues judgment. 17 Q. Other than the provision of information, assistance and 18 evidence in connection with the litigation concerning 19 the Horizon system, what work did you perform in this 20 consultancy? 21 A. I was called to go to a few meetings. A lot of it was 22 to do with the hangouts from prosecution. I wasn't 23 involved in any particular prosecutions as such but 24 there was quite a lot of work going on in the background 25 which -- talking to various lawyers. I got pulled in in 13 1 2016 for a bug that had been found that was quite 2 serious, the Dalmellington bug, that I think you may 3 well be coming back to later on, and it was felt that 4 I could maybe help explain what the consequences of that 5 was to Post Office. 6 I spent -- there was a problem that Post Office 7 had -- again, I can't remember exact date now it's in my 8 witness statement -- when the -- there was a mismatch in 9 Post Office's back end accounts. They tried to do 10 a reconciliation against data they had from Fujitsu, and 11 I realised what the problem was. It was to do with the 12 fact that they were taking the accounts at different 13 times. One was taking what the cash positions were at 14 7.00 in the evening and the other was taking the cash 15 position at midnight and, therefore, this accounted for 16 the mismatch that they had in the accounts. 17 So it was various ad hoc, fairly self-contained 18 tasks, each of which lasted maybe two or three days. 19 Q. There seems to be a fly -- 20 A. Yes, there is, it's a bit of a pain. 21 Q. -- buzzing around. I don't want you to be distracted. 22 A. No, I'm trying to ignore it. 23 Q. If it becomes too much, we'll break and get rid of it. 24 A. Okay. 25 Q. Would it be fair to say that the majority of this time, 14 1 certainly between 2015 and the Horizon Issues judgment 2 in December 2019, was given over to the provision of 3 Litigation Support? 4 A. Certainly in 2018 and 2019. I'm not sure it was direct 5 litigation, it was probably about half and half. As 6 I say, I can't remember the gory details of exactly what 7 I was doing when but there were a number of notes that 8 I produced which I wasn't quite sure what the background 9 was but, looking back at it now, I realise was actually 10 maybe in preparation for what eventually became the GLO. 11 Q. Have you read the Horizon Issues judgment? 12 A. I skimmed through it at the time and I've looked at 13 various sections but I don't claim to have read every 14 word, I'm afraid. 15 Q. In high level summary, would you agree that Mr Justice 16 Fraser concluded, firstly, that Legacy Horizon was not 17 remotely robust? 18 A. I accept that that's what he said. 19 Q. Do you accept that Horizon Online was susceptible to 20 accounting flaws? 21 A. Um -- 22 Q. That's what he -- 23 A. Again, that is what he said. I think he is maybe 24 putting it do strongly. I felt that Horizon Online was 25 and is -- because, after all, it is still the system 15 1 that's operating today -- doing a good job in terms of 2 the accounting, particularly with the monitoring that 3 was going on in terms of being able to detect things 4 when they'd occurred. 5 Q. He concluded that subpostmasters were not informed of, 6 and were therefore oblivious to, a range of bugs, errors 7 and defects. Do you agree that that is a summary of 8 what he found? 9 A. That is certainly what he said, yes. 10 Q. And, lastly, that those bugs, errors and defects could 11 result in, ie cause, discrepancies or shortfalls in the 12 branch accounts of subpostmasters, thereby corrupting 13 transactions, disrupting data processing and recording 14 transactions? 15 A. That was a possibility but I'm not sure that I, even 16 today, I understand what bugs actually did cause the 17 problems that people are -- that people have suffered 18 from. 19 Q. Do I take it that you accept his first finding that 20 Legacy Horizon was not remotely robust? 21 A. I don't accept that finding. 22 Q. You don't accept his finding that Horizon Online was 23 susceptible to accounting flaws? 24 A. There were some discrete bugs that caused problems to 25 the accounts but they were very discrete and I believe 16 1 they were all well controlled and managed at the time. 2 Q. You don't accept his findings that bugs, errors and 3 defects could result in, ie cause, discrepancies or 4 shortfalls in branch accounts? 5 A. They could cause discrepancies in branch accounts but 6 not at the sort of levels that are being talked about 7 and, in general, the systems, I believe, were operating 8 as they should. 9 Q. Robustly? 10 A. It depends exactly what you mean by "robust" but as long 11 as you're not saying "infallibly" then, yes, because 12 I think "robust" meant that there were mechanisms in 13 place that would monitor what was going on, detect 14 problems, and that they were then investigated and 15 resolved correctly. 16 Q. Horizon, both Legacy and Online, were working well in 17 your view? 18 A. Most of the time, there were clearly problems during the 19 pilots in both cases and there were clearly individual 20 problems that affected individual branches, and I'm sure 21 we'll come on to those at some time but, in general, 22 then I felt that the systems were working well. 23 Q. The judge got it wrong? 24 A. I wouldn't like to say that but I think there's 25 a difference in emphasis between -- there were clearly 17 1 problems and he identified a number of problems and 2 I won't dispute those problems happened but, on the 3 whole, I felt that the systems were working well. 4 Q. Can I turn in more detail to your role in the 5 development and rollout of the Horizon system, and in -- 6 no need to turn it up -- paragraphs 18 to 23 of your 7 first witness statement, you describe your roles and 8 responsibilities at ICL and then Fujitsu. So between 9 1996 and 2002, you tell us that your initial role with 10 Legacy Horizon was "to integrate Riposte and the Oracle 11 based software", yes? 12 A. That's correct, yes. 13 Q. And that you helped develop an agent layer that would 14 allow those two types of software to communicate with 15 each other? 16 A. Correct. 17 Q. Firstly, can you describe briefly what Riposte was? 18 A. Riposte was a -- well, let's step back one. There were 19 two aspects to Riposte and this was software that we had 20 bought in at Post Office's request from a company called 21 Escher in the United States. The bit of it that I was 22 primarily involved in was what was called the message 23 store, which was a way that all the data about what had 24 happen in a branch was being record. 25 So it's a sophisticated database, in crude terms, 18 1 and that was where the data, not only of transactions 2 but working data during the day, what had gone on when 3 people logged on, all sorts of things, was all recorded 4 in this message store. 5 Then the other part of Riposte was a development 6 environment that allowed applications to be built on 7 that and the application environment that caused the 8 human interface that the postmasters saw the branches, 9 and that bit of Riposte I had not very little 10 involvement with in the early days. I got more involved 11 in that when I moved on to Project IMPACT in 2003 or so 12 but, in the early days, my main interaction with Riposte 13 was to do with the message store, and -- 14 Q. Thank you. Then, the Oracle-based software. Can you 15 describe again briefly, as you've just done, what the 16 Oracle-based software was? 17 A. This was really an Oracle database and this was used to 18 communicate with Post Office's back-end systems where 19 they did their back-end accounting. So that was really 20 the boundary between what Fujitsu was doing, or ICL was 21 doing, and what Post Office was doing. So what was -- 22 that was basically an overnight batch processing system, 23 so it would process through all the transactions, it 24 would produce files of transactions that had happened 25 during that day, which were then transmitted to Post 19 1 Office's back-end systems. 2 Q. You say that you developed an agent layer; what was 3 an agent layer? 4 A. What an agent layer was doing was it was reading data 5 from the message store, and we had copies of the message 6 store in the data centre, and extracting the relevant 7 transactions and other information that was of interest 8 to Post Office's back-end systems, and writing records 9 to the Oracle databases to reflect those transactions. 10 Then there was also a flow in the other direction, 11 for example, reference data was being generated by Post 12 Office and, by reference data, the simplest example is 13 what's the price of a stamp today because, obviously, 14 the price of stamps changes every now and again. 15 So, therefore, things like prices weren't actually 16 built into the code; we just knew that you needed to 17 sell a stamp and then, behind that, there would be a bit 18 of reference data that says, well, if you sell a stamp, 19 it's going to cost you whatever a stamp costs today, 20 £1.15 -- back in the day it was about 30p or something. 21 Q. Given that the work started in 1996, what challenges 22 existed in ensuring that the communication by the agent 23 layer was enabled? 24 A. I'm not quite sure I understand the question. 25 Q. Yes. You tell us in your witness statement that your 20 1 role lasted between 1996 and 2002, which is 2 post-rollout? 3 A. Yes. So -- 4 Q. What took six years? 5 A. Right. The work changed a lot over time because, at the 6 beginning, back in 1996, the first application that was 7 rolled out was actually a Benefits Agency application 8 which, by 1999 -- '98/'99, had actually got dropped 9 because the Benefits Agency pulled out of the contract. 10 So a lot of -- 11 Q. Just stopping there, I wonder whether you mind slowing 12 down a little bit. 13 A. Sorry. 14 Q. Although the shorthand writer hasn't said anything so 15 far, I can imagine -- 16 A. Sorry. 17 Q. -- that she wants to. 18 A. Yes. 19 Q. So if you can just so down a little bit, please. 20 What did you know as to the reasons why the Benefits 21 Agency pulled out of the contract? 22 A. I'm not sure that I was really aware of the detail. 23 That was something that was taking place at levels way 24 above where I was working. 25 Q. What were you told? 21 1 A. I can't remember now what I was told. I think they 2 decided they were going to do things differently. What 3 they eventually did was they paid people's benefits 4 directly into their bank accounts, rather than doing it 5 through special application with the Post Office. 6 Q. Were you told anything to the effect that one of the 7 reasons why the Benefits Agency withdrew was a lack of 8 confidence in ICL -- 9 A. I'm not sure that I was aware of that at the time. 10 Q. -- and doubts over the integrity of the data that the 11 nascent system was producing? 12 A. I don't remember hearing that at the time. 13 Q. Anyway, you continued in the role until 2002. Does that 14 mean that the agent layer required a continual level of 15 support and maintenance, even after rollout? 16 A. No, because there were few -- there was new 17 functionality. So the main thing I was doing in the 18 early 2000s was -- network banking was introduced into 19 the system in 2003 and, a few months later, the support 20 of debit cards was introduced. So the main work I was 21 doing with the agents then was supporting the changes 22 that were required because one of the things the agents 23 did was they interfaced between the data centres and the 24 banking systems, and that was clearly a fairly 25 significant development. 22 1 Q. You tell us that you moved away from the agent team and 2 took on the role of defining changes required by Project 3 Impact. Did the work of the agent team continue after 4 you left? 5 A. Oh, yes, there was a team there. Someone else took over 6 the role of Chief Designer within the agent team. 7 I think he actually took over that role about 2000, 8 something like that, and I was gradually moving on to 9 other things on the side. So it wasn't a sudden "Drop 10 this and do that"; it was more of an evolution than 11 a revolution. 12 Q. In your role between 1996 and 2002, were you in any way 13 responsible for communicating with the Post Office team 14 as to their requirements? 15 A. Not so much with their requirements. That really 16 started with Project Impact but I certainly was working 17 with the Post Office team in terms of defining the 18 interfaces between how Fujitsu's agents were operating 19 and how they interacted. Initially, there was a box in 20 between us and the banks called the Network Banking 21 Engine which was provided by IBM so, initially, we 22 provided the interface into that. Then, later on, that 23 was replaced, after I was no longer involved, by 24 a direct interface to the banks. 25 Q. Did you come into contact with Jeremy Folkes, the 23 1 Infrastructure Assurance Team Leader from the Post 2 Office? 3 A. Not at that time but I did come across him a few years 4 later when he moved roles and he actually joint Escher 5 as one to their consultants. So I did come across 6 Jeremy in his role with Escher but I don't remember 7 coming across him when he was working with Post Office. 8 Q. He told the Inquiry, it was back on 2 November 2022, 9 that "we", and by that he meant the Benefits Agency and 10 the Post Office Counters' team "had been denied 11 viability of the application design". 12 A. I'm not quite sure what he meant by that. 13 Q. He continued: 14 "When I say we had some visibility of risk areas, 15 what I mean is that the areas where we had raised formal 16 risks to the service provider at the start, such as 17 Riposte, in those cases, we did get more information 18 but, as far as the application design, in particular 19 EPOSS, we had been denied visibility." 20 A. I think that was something to do with the way the 21 contract had been negotiated between -- or ICL 22 management, as it was then, and Post Office, and DSS. 23 I wasn't involved in any of that but, as I understand 24 it, it was a PFI contract and part of that meant that 25 the customer that to treat the whole things as a black 24 1 box but, as I say, this is hearsay, if you like, rather 2 than something I was directly involved in. 3 Q. Where did the hearsay come from? 4 A. Just chat around the office. As I say, I wasn't 5 directly involved in all that but, clearly, people were 6 talking about -- I think I was aware that Post Office 7 did not have the right to see various design documents 8 that were being produced. 9 Q. What do you mean that they were buying a black box? 10 A. That it was up to Fujitsu exactly how they designed the 11 application and the details of how the internal design 12 was done was something for Fujitsu or ICL to worry 13 about, and it wasn't something that Post Office would 14 get involved in. 15 Q. Would a consequence of that, in relation, for example, 16 to EPOSS, mean that it would be difficult, if not 17 impossible, for the Post Office to ensure that there was 18 data integrity? 19 A. I don't think it went that far. There were certainly 20 joint test teams. There was a lot of testing going on. 21 There were Post Office testers involved in testing the 22 system to make sure it came up with the right output. 23 So I -- yeah. 24 Q. He suggested that the fact that the Post Office had 25 purchased a black box and had no rights to see how the 25 1 black box worked meant that it was difficult, if not 2 impossible to ensure data integrity within the Horizon 3 system; you disagree with that? 4 A. I would disagree with that, yes. 5 Q. You say that the answer to that is that they were 6 entitled to participate in testing? 7 A. Yes, as I say, I wasn't really involved at that sort of 8 level at that time. I was just looking after making 9 sure that the agents worked, rather than worrying about 10 how the counter worked and things like that. 11 Q. Were you conscious, and were those with whom you worked 12 conscious, that the nature of the contract was a PFI 13 contract and, therefore, the client had no rights, no 14 visibility, over of the way the system was built? 15 A. I think I was aware of that, yes. 16 Q. What effect did it have on your work? 17 A. Nothing really, other than -- at that time I wasn't 18 having any real contact with the customer anyway. 19 I only started getting involved with the customer later 20 and it was really a case of -- and the main work I was 21 doing with the customer in the early 2000s was with 22 agreeing these banking specs and there I was working 23 very openly with the customer because they were their 24 specs that were defining the interfaces as to how 25 Horizon was going to interface with, initially, the 26 1 banking engine and then also with Streamline for the 2 debit card payments. So that was something that wasn't 3 part of the black box. 4 Q. Why was the black box being mentioned; why was it being 5 talked about? 6 A. Well, I don't think it was being talked about as such, 7 it was really -- I probably picked that up from Jeremy 8 Folkes' stuff. As I say, I was aware that there was 9 a PFI contract and that it meant that we didn't 10 necessarily need to share detailed design documents with 11 the customer because it was Fujitsu's intellectual 12 property. 13 Q. Was that the usual way of working or an unusual way of 14 working? 15 A. I'd not worked on the contracts like that before. 16 Before, I'd been in an internal development role 17 actually producing products that were sold rather than 18 working as a client for a specific customer like that. 19 Q. So you wouldn't know one way or another? 20 A. Correct. 21 Q. In relation to your fourth line support role, you tell 22 us in your witness statement, it's the first witness 23 statement, at paragraph 25: 24 "I would only have been aware of bugs, errors and 25 defects specifically allocated to me by third line 27 1 support or where I was asked to route the issue to the 2 correct person in fourth line support, or where I was 3 asked for specific input on the problem because of my 4 particular expertise." 5 Correct? 6 A. Yes, yes. 7 Q. Within your roles in Legacy Horizon, did you have any 8 oversight of all bugs, errors and defects within the 9 system at any given time? 10 A. No. 11 Q. So you couldn't, at any particular time, speak to the 12 existence or non-existence of bugs, errors and defects 13 in the system? 14 A. Not from my personal knowledge, no. 15 Q. If you did not have knowledge of, or oversight of, all 16 bugs, errors and defects within the system at any given 17 time, who did? 18 A. I would have thought the SSC would and -- there was 19 an area called Customer Services, SSC was part of that. 20 There were also problem managers who would look at 21 specific incidents. Now, incidents could be to do with 22 bugs, errors and defects or, more commonly, they would 23 be operational issues, where a bit of the network went 24 down or connections to banks went down for a few hours 25 which clearly was -- caused fairly catastrophic effects 28 1 at the time, and so their job was to manage these sort 2 of things. So I'd have thought that, if there was 3 an overall knowledge of how well the system was 4 behaving, it would be part of their role. 5 Q. We've heard from a lot of people from the SSC, and 6 they've told us that there were between 20 and 30 people 7 at any one time working in the SSC -- sometimes working 8 shifts, some working at home, some working in the 9 office -- that they only knew about the bugs that they 10 were allocated on the stack and that it was correct that 11 there might be informal sharing of information by people 12 who were sitting next to each other or talking about it 13 at the water cooler but there was no formalised system 14 for sharing knowledge within the SSC; did you know that? 15 A. I didn't know that. They would obviously be able to 16 speak better about that sort of thing than I was. 17 I wasn't part of the SSC. 18 Q. But when I asked you who may have oversight of all bugs, 19 errors and defects within the Horizon system, you 20 pointed to the SSC. 21 A. As part of that but there was also -- there was -- the 22 whole point of the Customer Services Directorate was to 23 actually manage the interface with Post Office, 24 including knowledge of what had been going on the 25 system, what sort of problems there were and things like 29 1 that. 2 Q. That's a day-to-day function. What I'm looking for, it 3 might be described as an all-seeing eye, but somebody 4 who took a step back and would be able to say, "We've 5 had these problems in the past with Horizon, these bugs, 6 errors and defects, and, at the moment, we're servicing 7 the existence of these possible bugs, errors and 8 defects"; was there any such person? 9 A. I can't immediately think of anyone who would have that 10 specific role. 11 Q. When you were in fourth line support, how would you 12 generally become aware of such bugs, errors and defects 13 as you did become aware? 14 A. There was a system called PEAK and it had a predecessor 15 called PinICL, and the way that operated, that was 16 effectively a database of the various defects, and that 17 included those found during testing as well as those 18 during live, and what happened was that PEAKs or PinICLs 19 would be allocated to a particular person, and I think 20 you'd normally get an email saying, "You've had this 21 PEAK that's allocated to you", you could then log in to 22 the PEAK system, have a look at the PEAK and see what it 23 said and then decide what you needed to do with it next. 24 Q. Again, same questions: within your various roles within 25 Horizon Online, did you have oversight of all bugs, 30 1 errors and defects? 2 A. Not as such but I did have a -- probably a better view 3 of them, particularly during the pilot phase of Horizon 4 Online because, during the pilot phase, there was 5 a project team in place that was managing the pilot and 6 taking very close monitoring of any issues that were 7 going on during the pilot, and I would often get called 8 in by that team to investigate particular problems 9 particularly ones -- those that affected the counter. 10 Q. Are the pilot phase of Horizon Online was over, did you 11 have oversight of all bugs, errors and defects within 12 the Horizon Online system? 13 A. No. 14 Q. Same questions: if not you, did anyone? 15 A. Again, I think it was something that Customer Services 16 were supposed to be monitoring and managing. Now, 17 whether they did -- had that sort of stand back 18 oversight role that you've described earlier, I can't 19 think of any one person that did actually have that sort 20 of role but I hadn't thought of it that way at the time. 21 Q. Again, same questions: how would you generally become 22 aware of such bugs, errors and defects as you became 23 aware of in Horizon Online? 24 A. The PEAK system was working. It was the same PEAK 25 system that operated for both Legacy Horizon and Horizon 31 1 Online. 2 Q. So for both Legacy Horizon and Horizon Online, you 3 became aware of things that might be bugs, errors and 4 defects through the PinICL and then the PEAK system? 5 A. There was that but there was also the problem management 6 system. So then when something was identified as being 7 a serious problem -- and I think we're probably going to 8 go on and talk about the receipts and payments mismatch 9 problem at some stage -- then I was often called in to 10 help -- well, firstly, analyse the problem and explain 11 that problem so that Fujitsu management could understand 12 what it was all about, and I often got involved in 13 explaining it to the equivalent people in Post Office 14 about what the impact of that was, and what could be 15 done to remedy it, both in terms of fixing the code but 16 also fixing whatever the business impact of the problem 17 was, which wouldn't necessarily be fixed directly just 18 because you'd fixed the code. 19 Q. So, for you personally, there were two ways in which you 20 became aware of bugs, errors and defects in Horizon -- 21 A. Yes. 22 Q. -- firstly, through your role -- the '10 per cent of 23 your time' role -- 24 A. Yeah. 25 Q. -- with fourth line support, and that would largely be 32 1 happenstance, would it, ie which PEAKs were allocated to 2 you? 3 A. It would -- the 10 per cent wasn't just for the PEAKs. 4 So any time that I spent talking about serious incidents 5 and things like that, that was all part of what I saw as 6 the 10 per cent of the time. 7 Q. In any event -- 8 A. Yes. 9 Q. -- in your role in fourth line support -- 10 A. Yes -- 11 Q. -- it would be largely happenstance whether it was you 12 or somebody else, one of the other people performing 13 fourth line support, who had a particular bug, error or 14 defect allocated to them? 15 A. Yes, it was based really on the skills and what area the 16 problem had happened in. 17 Q. If it was largely happenstance that you became aware of 18 a bug, error or defect, in both Legacy Horizon and 19 Horizon Online, how could you subsequently give full, 20 complete and accurate evidence about the existence or 21 non-existence of bugs, errors and defects in Horizon? 22 A. I was -- what I was aware of was the fact that bugs that 23 actually impacted the accounts were very rare, there was 24 good monitoring in place to detect them and they got 25 fixed shortly afterwards. So, in terms of what was 33 1 actually there in the live system at any one time, it 2 was very rare for there to be bugs there that would 3 cause problems and, therefore, I was confident in the 4 way that the system was operating, that it was operating 5 correctly. 6 Q. To give that kind of very confident answer to a court, 7 wouldn't you want to know of the existence of each of 8 the bugs and how they'd been resolved and whether, in 9 fact, there was any ongoing impact? 10 A. I didn't realise at the time that I needed to do that. 11 So at the time, no, I didn't think I needed to do that. 12 Obviously, with hindsight, I realised that maybe 13 I should have been doing more research but, at the time, 14 I felt that that was sufficient. 15 Q. The "that" in that sentence, meaning a general 16 confidence in the system and the way that it operated? 17 A. Yes, and the processes that were in place to actually 18 control things. 19 Q. So generally confidence in the system, plus processes 20 that you thought were working, allowed you confidently 21 to give a generalised view; is that right? 22 A. I think so, yes. 23 Q. During your employment within Fujitsu, did any of your 24 roles entail any overall responsibility for monitoring, 25 identifying or fixing bugs, errors or defects? 34 1 A. Well, we've talked about some of those already, so I'm 2 not quite sure what -- 3 Q. Did they always come to you? 4 A. Oh, no, no. I mean, they go to whoever was the relevant 5 person. So there were certain areas that -- normally, 6 they just went straight to the Development Teams and 7 stint get involved in things. I tended to get involved 8 if it was seen as being a broader issue or something 9 like that. Not necessarily me, there were other people, 10 also in particular areas, would get called with. 11 Q. How many other people? 12 A. I'm not sure, sort of another half a dozen, a dozen, 13 something like that. 14 Q. What were the circumstances in which you would get 15 called in? 16 A. It's a bit difficult to define, really. Certainly, when 17 we got on to things like Horizon Online, if it was seen 18 as having an impact on the accounts, then we realised 19 that that was a very serious problem and that I would be 20 getting involved in those, which is why I got involved 21 with what I see as the main serious issues in Horizon 22 Online. And I was involved -- 23 Q. Receipts and payments mismatch? 24 A. Receipts and payments mismatch, the local suspense issue 25 and Dalmellington, those are the ones that I see as 35 1 being serious issues with Horizon Online, and I got 2 involved in all of them because they saw me being 3 helpful in terms of my expertise and background for 4 those particular cases. 5 Q. That was my next question: why did they come to you? 6 A. Because I'd had experience of dealing with things like 7 that. I think it was felt that I was in a good position 8 of turning some of the technical jargon into something 9 that laypeople could understand better. Now, having 10 read my witness statement, you may disagree with that, 11 but it was felt that I was in a good position to be able 12 to actually explain the technical problems in a way that 13 could be understood and to do that, both in terms of 14 written notes and in meetings with the customer when 15 these problems occurred. 16 Q. What was the threshold for you becoming involved? 17 A. I'm not sure. It's -- I would get involved by probably 18 a phone call or an email from someone, either in the SSC 19 or from problem management, and they'd say "Gareth, can 20 you come and give us a hand, look at this problem and 21 see what's going on". 22 Q. For all the problems that didn't get referred to you, 23 where that trigger was not pressed, how did you become 24 aware of those, or didn't you? 25 A. Probably the answer is I didn't. 36 1 Q. When you started to become the man that the Post Office 2 relied on to give evidence in court, did anyone say, "We 3 need to make sure that Mr Jenkins knows about the things 4 that are not referred to him"? 5 A. No. 6 Q. Did it occur to you, "I might need, before going to 7 court, to find out about the things that have not been 8 referred to me, that aren't these level of particular 9 seriousness or which have financial or economic 10 consequences"? 11 A. That didn't occur to me but, again, as I've said before, 12 I was confident -- and possibly wrongly so -- that when 13 problems did occur, they were quickly fixed and they 14 weren't left to sort of fester in the system, to have 15 a larger impact. 16 Q. I think you've acknowledged it a couple of times 17 already, knowing what you know now, would you adopt the 18 same approach of relying on your confidence? 19 A. I think I'd have to say that, with hindsight, I would 20 have done things differently, yes. 21 Q. What interactions did you have with the third line of 22 support, the SSC, in relation to identifying and 23 rectifying bugs, errors and defects? 24 A. I'm not quite sure exactly what you mean by that 25 question, sorry. 37 1 Q. Yes. What was the nature of your communication with 2 third line support? 3 A. Usually, either by exchange of emails or telephone 4 conversations, and things like that. The third line 5 support team were in a locked area, so it was actually 6 quite difficult to actually go and -- I did occasionally 7 go and visit them and talk to them at their desks and 8 things like that but, generally, that was a difficult 9 thing to do because of getting through the security 10 systems to actually get to where their desks were. So 11 it was normally done by sort of phone or email or in 12 meetings, if they came out of their area to our slightly 13 less secure areas. 14 Q. Thank you. Can I turn to EPOSS and Riposte, please. 15 A. Yes. 16 Q. In relation to the EPOSS aspects of Horizon, you tell 17 us, it's paragraph 21 of your second witness 18 statement -- no need to turn it up: 19 "In summary until around 2003 I do not recall having 20 anything to do with EPOSS other than indirectly when it 21 had an impact on my work with the agents. From around 22 2003 onwards, I began to gain technical expertise and 23 a practical understanding of how EPOSS operated." 24 Is that -- 25 A. Yes, I'd agree with that. 38 1 Q. You've told us already that Escher was responsible for 2 provision of the Riposte software; is that right? 3 A. Correct, yes. 4 Q. And that that comprised of two main parts: the message 5 store and the desktop counter; is that right? 6 A. Correct. 7 Q. The latter of those, the desktop counter, was the basis 8 for the EPOS system? 9 A. Correct. 10 Q. I think you went to Escher in Boston for a week in 1996; 11 is that right? 12 A. Yes, I think I went back there for two or three ad hoc 13 trips after that, but that was my initial -- that was, 14 effectively, my first week on the Post Office Account, 15 when I got sent to Boston for a week to learn about how 16 Riposte worked. 17 Q. If you did not have anything to do with EPOSS, other 18 than indirectly, why did you go to Boston for a week in 19 1996? 20 A. Because of the -- that was talking more about how the 21 Riposte message store was working and, also, there was 22 specific interfaces that Riposte had that could be used 23 by the agents, so that was covered in the first week of 24 what I think was a three-week training course. It then 25 went on to how did the counter applications operate and 39 1 I wasn't particularly interested in that at the time, so 2 it was felt that, rather than stay in Boston for another 3 couple of weeks, I might as well come home. So I did. 4 Q. So is the answer that you went to Boston to learn how 5 the EPOS system worked, in particular how it might 6 interact with the work you were undertaking? 7 A. No. I went to Boston to learn how Riposte worked, to -- 8 and, therefore, how the agents could actually use it. 9 There may have been slight mention of how applications 10 could work in that first week but I think the detail of 11 that was really covered afterwards, and I just didn't 12 get involved in any of that. 13 Q. In relation to the development of EPOSS before the 14 rollout of Legacy Horizon, we've heard evidence from 15 David McDonnell; did you know him? 16 A. Vaguely. I have a vague recollection of him. 17 Q. He told the Inquiry: 18 "My understanding was that Gareth [that's you] 19 Jenkins worked alongside another Chief Architect under 20 Alan Ward." 21 Was that correct? 22 A. I certainly did have a dotted retail line relationship 23 to Alan Ward, who was the Chief Architect at that time. 24 I thought that David McDonnell actually suggested I was 25 the Chief Architect and he's certainly mistaken in that 40 1 belief. 2 Q. He continued: 3 "Gareth Jenkins' responsibility was specifically to 4 the EPOSS counter system." 5 Is that correct? 6 A. That is not correct. 7 Q. He said: 8 "As the Chief Architect, I would have expected him 9 to be much more involved in overseeing a lot of the 10 previous coding standards and methodologies, and things 11 like that, and certainly the design documents." 12 You've, I think, answered this question already: 13 were you the Chief Architect of Horizon? 14 A. No, I wasn't. 15 Q. Who was the Chief Architect of Horizon? 16 A. Alan Ward was at that time and I know he was quite 17 involved in the development of EPOSS at that time 18 because there were quite a lot of changes being made and 19 he used to regularly fly off to Boston for a couple of 20 days almost every week, to actually work with Escher on 21 changes to EPOSS. Now, exactly what there was changes 22 were, I've no idea; I was just aware that it wasn't 23 a good idea to try and find him on a Monday and 24 a Tuesday because he was probably in Boston. 25 Q. Okay, and if you could just slow down your evidence 41 1 slightly please. 2 A. Sorry. 3 Q. He told us that there was or had been: 4 "... a lack of formalised sign-off designs, a lack 5 of discipline, a lack of professional qualification in 6 key positions and a total disengagement of the Chief 7 Architect, Gareth Jenkins, [that] there were poor coding 8 standards, no methodology in place and no unit testing. 9 The issues were critical, making the EPOSS product 10 unstable. They were known to everyone in the building." 11 Taking those in turn, you've already told us that 12 you were not the Chief Architect of Horizon but, putting 13 that aside, in relation to EPOSS, did you know whether 14 or not there was an existence or a lack of formalised 15 sign-off designs. 16 A. I've no idea what they had. I know what we had in the 17 Agent Team and things were formally signed off, I had no 18 idea what sort of processes they were following in the 19 EPOSS Team. 20 Q. Did you know whether or not there was a lack of 21 discipline and a lack of professional qualifications in 22 the EPOSS Team? 23 A. I wasn't close enough to it to know either way. 24 Q. Did you know, one way or the other, whether they were 25 using poor coding and there was an absence of 42 1 methodology in place? 2 A. I have no idea. 3 Q. Did you know whether or not there was unit testing 4 within the EPOSS team? 5 A. I don't know. 6 Q. He says that those issues were critical and were known 7 to everyone in the building. You're telling us that -- 8 A. I would agree they were critical. As to being known to 9 everyone in the building, the other thing to remember is 10 that the EPOSS team was based in Feltham and I was based 11 in Bracknell. Yes, I used to visit Feltham a couple of 12 times a week but I wasn't, actually, based in the 13 building all the time so, therefore, I would have had 14 less knowledge of what was going on in the background. 15 Q. So these rather important facts that he told us about, 16 you were entirely unaware of? 17 A. Correct, and I'd agree with him they are important 18 facts. 19 Q. Were you aware of an EPOSS Task Force being set up by 20 ICL? 21 A. I don't think I was aware of it at the time. Obviously, 22 I'm aware of it now from the documents I've been shown 23 but I don't think I was aware of it at the time. 24 Q. Can we look, please, at FUJ00080690. If we just look at 25 page 2, please, we can see this was originally dated 43 1 18 September 1998 -- 2 A. Yes. 3 Q. -- and this finalised version is dated 14 May 2001. 4 A. Yes, I can see that. 5 Q. If we go -- 6 A. Yeah. 7 Q. -- back to page 1., it's a "Report on EPOSS PinICL Task 8 Force" and it: 9 "... reports on the activities of the EPOSS PinICL 10 Task Force, which was in place between 19 August and 11 18 September 1998 to reduce to manageable levels the 12 EPOSS PinICLs outstanding at that time." 13 You can see, although it's on the second page 14 described, I think, as a finalised version, the status 15 is still a draft. 16 A. Yes. 17 Q. The distribution does not include you. It's to 18 Messrs Austin, Bennett and McDonnell, but then 19 "Library"; can you see that? 20 A. Yes, I can see that. 21 Q. What was the nature of ICL's document library at this 22 time? 23 A. There was a document management system, I think -- 24 I can't remember exactly when that became formalised. 25 I think Matthew Lenton may have described something 44 1 about this when he appeared a couple of weeks ago and 2 there was a document management system to which all 3 documents were put and people could look at documents if 4 they wished to do so but, normally, you would only be 5 looking at documents that either you'd written yourself 6 or were needing -- being asked to review in that 7 library. 8 Q. So you wouldn't -- 9 A. I wouldn't have seen this document, unless someone 10 pointed it out and said, "What do you think of this 11 document?", and I'm not aware that they ever did. 12 Q. So it wasn't used as a knowledge base for those at your 13 level to peruse or look at your leisure? 14 A. I suppose I could have done, if I'd got nothing better 15 to do, but I don't think I would have been particularly 16 interest in what was happening with EPOSS at that time. 17 Q. I think you've seen this document subsequently -- 18 A. I've seen this document recently, yes. 19 Q. -- and you know -- I'm not going to take you through it, 20 we've seen it many times -- it gives a very damning 21 account of the problems with EPOSS? 22 A. I agree. 23 Q. Mr McDonnell told us in his evidence that he presented 24 this report to you; is that true? 25 A. I don't think he presented it to me. I think he 45 1 presented it to whoever commissioned it, which I think 2 I understand now is Terry Austin. 3 Q. And that you "denied the issues pointblank and ran off 4 to hide in Bracknell, and avoided contact with the 5 team"? 6 A. I don't think that's correct. I was -- as I said 7 before, I was based in Bracknell, not in Feltham. 8 I don't think I was hiding there particularly. I don't 9 think I had any involvement with the EPOSS PinICL Task 10 Force. 11 Q. He said that: 12 "We managed to get Gareth down to the counter team 13 twice." 14 Do you remember that? 15 A. I don't. I do remember having some involvement with him 16 with some issues with Logistics' visa system, which 17 I think is something he moved on to work with after he 18 left -- was no longer with the counter team, but I don't 19 recall having anything to do with him in terms of EPOSS. 20 Q. He said that his team tried to engage you in 21 a conversation about the missing API; do you remember 22 what that is? 23 A. I think that was to do with the LFS thing that 24 I mentioned previously. So I think that was in a role 25 that he took on after the EPOSS team. I can't remember 46 1 the gory details of it I'm afraid but I -- 2 Q. He said that you were very defensive of it and said, 3 "No, there's nothing wrong with it as it is". 4 A. I think it was more a case of pragmatism, in that 5 I think he wanted to totally change the way that the 6 Escher software interfaced with the applications. And 7 I could understand why he felt that there may have been 8 a better way of doing it but I could also see that there 9 was no way that Escher was going to go and totally 10 rewrite their software when it was basically operating 11 and doing what it needed to do. 12 The fact that there was a better way did not 13 necessarily mean that it would be justifiable to 14 actually change it. So I think there may have been some 15 sort of conversation along those lines but, as I say, 16 it's a very long time ago and I can't really remember 17 the details but that's a vague memory I have of that 18 area. 19 Q. He said that he tried to engage you to lend your 20 "political design weight" behind at least the cash 21 account being rewritten; is that correct? 22 A. I don't have any recollection of being involved with 23 anything to do with the cash account at that time. 24 Q. He said that he was unable to engage you to get you on 25 his side and to lend your persuasive weight to persuade 47 1 Terry Austin to rewrite the cash account; is that 2 correct? 3 A. I have no recollection of that and I can't see how 4 I would possibly have been able to persuade Terry that 5 it had something to do with the cash account when it was 6 nothing to do with my area. 7 Q. Did you have what he describes as "political design 8 weight": a certain amount of power? 9 A. Probably not so much at that time. I think maybe later 10 on then I -- that could be said but, at that time, my 11 main role was to do with the way the agents worked, 12 which was a fairly small cog in the system. It was 13 a fairly central cog and, yes, I think people did tend 14 to listen to me when I had technical ideas, but I think 15 that's over-stretching things, what you've just read 16 out. 17 Q. He said that you became evasive with him and he was 18 never able to persuade you to come back down to Feltham. 19 A. I just can't remember those sort of interactions. I've 20 described to the best of my recollection the only sort 21 of interaction that I might have had with him, and 22 that's a very, very vague memory, I'm sorry. 23 Q. Leaving aside your involvement and role in the way that 24 Mr McDonnell has described, would you agree with, 25 looking back now, what he said about the EPOSS team in 48 1 the late 1990s? He said the team was: 2 "... like the Wild West, there were no standards in 3 place, there were no design documents. The culture of 4 the Development Team was -- I wouldn't say it was 5 a holiday camp. It was a free format. There was no 6 structure, no discipline. It was crazy. I've never 7 seen anything like it." 8 A. I wasn't close enough to actually form any opinion on 9 that, I'm afraid. 10 Q. Do you remember going down to his team on at least 11 couple of occasions? 12 A. Not really. I -- as I say, the only recollection I have 13 is discussing issues to do with LFS, which was a later 14 role that he moved on to. I don't remember having any 15 real involvement with the EPOSS team at that sort of 16 time. 17 Q. That can come down. Thank you. 18 If you'd been asked a question at rollout time, to 19 what extent does the EPOS system function well and 20 effectively; what would your answer have been? 21 A. I don't know. 22 Q. You -- 23 A. I mean, my answer would have been "I don't know". 24 Q. You wouldn't know one way or the other? 25 A. Yeah. 49 1 Q. So if you were to give evidence about that, you wouldn't 2 be able to say one way or the other? 3 A. Correct. 4 Q. If you were to give evidence later on about events that 5 were happening in 2000, you wouldn't be able to say 6 whether EPOSS was functioning well or not? 7 A. Not in 2000, no. I believe it had stabilised by the 8 time I did get involved with it later on but, back in -- 9 what state it was in 2000, I don't know how stable it 10 was and, clearly, as part of the work I did in 2018, 11 I did find some fairly serious issues that had occurred 12 at that time. But -- 13 Q. What were the issues you found out in the course of the 14 Group Litigation -- 15 A. Yeah, that's right, yes. 16 Q. -- in 2018? 17 A. And those work ones that I'd not been aware of until 18 I came across them as part of the Group Litigation. 19 Q. Can you, rather than me doing it on -- it'll be late on 20 Thursday -- can you summarise what those are now? 21 A. The main one I can think of is the data tree build issue 22 that came -- that came around. That's the one that sort 23 of sticks in my mind at the moment. You may find others 24 that you may wish to prompt me about but that's the one 25 that sticks in my mind. 50 1 Q. Was that a surprise when you found this out in the 2 course of the Group Litigation? 3 A. That such serious problems had occurred and seemed to -- 4 though again, with that, the problems did get fixed but 5 I was surprised in some of the cases how long it took to 6 fix them and there were certainly some fairly 7 inappropriate comments in some of the PEAKs in terms of 8 whether things should be fixed quickly or not, which 9 I certainly wouldn't have agreed with if I'd seen them 10 at the time. But then it wasn't my job to see them at 11 the time. 12 Q. Was your surprise compounded by the fact that you had, 13 in the interim period, between the relevant events 14 happening with the data tree build failure and you 15 discovering them in 2018, you'd given evidence in 16 written witness statements, and on one occasion orally, 17 in court? 18 A. No, I didn't think it was a problem as far as that was 19 concerned because, at the times I was giving evidence 20 for, I believed that the EPOS system was stable and was 21 operating correctly. So the fact that there were 22 problems during the pilot and the rollout don't 23 necessarily mean that the problems carry on into the 24 system. I was confident in the way that problems were 25 being picked up and fixed and knew things were being put 51 1 into the system to actually manage the issues that are 2 being found early on. 3 Q. We're going to come back to the kind of answer that 4 you've just given quite a lot over the course of the 5 next two or three days but, in the answer that you've 6 given just now, do you agree that you are focusing on 7 whether it can be shown that an issue had an impact in 8 the case that you are looking at -- 9 A. Yes. 10 Q. -- rather than giving evidence about the existence of 11 bugs, errors and defects within Horizon, how quickly 12 they manifested themselves to Fujitsu, how promptly the 13 Post Office was told about them and the extent to which 14 subpostmasters had themselves been informed? 15 A. Yes, I was primarily looking at what was happening in 16 a particular branch at a particular time. 17 Q. Was that your mindset when you were giving evidence? 18 A. Yes, that's how I approached support, and I didn't see 19 the giving of evidence as being any different from what 20 I was doing in my day-to-day support job. What I'd been 21 asked to do in support, as saying this has happened, 22 should it have happened and, if not, what's gone wrong? 23 And I just approached things on the same sort of basis. 24 Q. Thank you. Can we turn to the IMPACT Programme. You 25 tell us in paragraph 20 of your witness statement that 52 1 you became involved in Project IMPACT, as it was called. 2 That was the name given to it by the Post Office and it 3 was essentially the replacement of the back-end 4 accounting system. Is that -- 5 A. Yes. 6 Q. -- an overall description of it? 7 A. Yeah. 8 Q. Given your role in the project, can you describe shortly 9 what Project IMPACT was designed to do, so far as the 10 back-end accounting system was concerned? 11 A. Post Office had a propriety in-house system called, if 12 I remember rightly, CBDB -- I can't remember what that 13 stands for -- where they used to process their accounts, 14 and they decided they wanted to go for a more industry 15 standard system and SAP was the industry standard, and 16 probably still is, for handling accounts, and so it 17 wanted to move on to that. 18 Now, CBDB worked on the basis that it would take the 19 information from the cash accounts that were signed off 20 each week and derive everything from that, which meant 21 that it was always about a week behind what was really 22 going on, whilst with an SAP system, they thought they 23 could get a much more realtime knowledge of what was 24 going on in the system. So with SAP they would know 25 what had happened yesterday and, therefore, be in a much 53 1 better position, particularly to control their cash 2 flow, because they saw the amount of cash flowing 3 through the post offices as being a fairly critical 4 measure because, obviously, cash had to be paid for, and 5 it was all to do with Treasury interest rates and things 6 like that that I didn't really understand. 7 But I understood that having a good handle on 8 exactly how much cash you had where made a big 9 difference in terms of Post Office finances. 10 Q. Can we look at your first witness statement, please, 11 WITN00460100, at page 13, please. It'll come up on the 12 screen. At paragraphs 45 and 46 you say: 13 "... I believe that ICL/Fujitsu was only involved in 14 one aspect of Project IMPACT, which was implementing 15 changes to Legacy Horizon that would enable it to 16 interface with the new back-end system. This was 17 implemented in two phases as part of the S60 release in 18 2004 and the S80 release in 2005. My involvement was to 19 design the changes which [Post Office] wanted. 20 "I remember that these phases of the work required 21 a major re-engineering of the accounting processes in 22 each [Post Office] branch, [for example] removing the 23 cash account and introducing the branch trading 24 statement." 25 A. That's correct, yes. 54 1 Q. So, as the counters specialist, which is what you were, 2 you redesigned part of the Legacy Horizon system that 3 produced the accounts for the subpostmaster to sign off; 4 would that be fair? 5 A. I oversaw that. I think the main detail was done by 6 other designers but my main role was actually 7 identifying exactly what the output needed to be of the 8 system. So, yeah. 9 Q. So you were the lead designer -- 10 A. Yes. 11 Q. -- of this part of the IMPACT Programme -- 12 A. Correct. 13 Q. -- ie that part of the programme which redesigned the 14 accounts for subpostmasters to sign off? 15 A. Yes. 16 Q. Can we look, please, at POL00038878. If you look at the 17 foot of the page, you'll see it's dated 3 March 2004? 18 A. Right, yes. 19 Q. If we go to the top of the first page, it's described as 20 "Branch Trading Reporting, Management and Control and 21 Transaction Management: Conceptual Design". Is, 22 essentially, this a design document for the IMPACT 23 Programme? 24 A. Not really. I'd call it a requirements document. So 25 Post Office decided that the term "conceptual design" 55 1 was what should be -- the document be called but, to me, 2 it was a requirements document. This was setting out 3 what Post Office wanted to happen and this was actually 4 written by Post Office, though aided by a Fujitsu 5 requirements analyst called Phil Boardman, who I believe 6 has been in front of the Inquiry a year or so ago. But 7 this -- 8 Q. And you: you contributed to this? 9 A. I was involved in the workshops doing it. I don't think 10 I wrote any of the text in there but I was involved in 11 commenting on it and helping put together the diagrams. 12 But it was basically understanding what Fujitsu wanted 13 to do, and my involvement was really a case of saying, 14 well, if you want to do that, is that practical? Would 15 it be better to do it in a slightly different way to 16 make it more practical from a (unclear) point of view? 17 Q. Just scroll down. We see your name as a contributor. 18 A. Yes. 19 Q. Are you saying you didn't contribute to this? 20 A. Well, I contributed ideas to it but I don't think 21 I actually wrote any of the words in it. 22 Q. Can we go to page 22, please, and paragraph 6.7, 23 an "Overview". If we just read this together: 24 "This area of functions has the purpose of providing 25 mechanisms to make adjustments to branch accounts, to 56 1 correct errors and ensure branch accounts align with the 2 Post Office Accounts within POLFS. Various mechanisms 3 are available to identify errors that require 4 adjustments, and the discrepancy management functions 5 may be initiated from various places across the 6 business. The main areas will be from within the 7 Branch, from [Post Office] clients or centrally via 8 distributing electronic transaction corrections. These 9 corrections will replace the current error notice 10 processes and should not involve any manual paperwork or 11 processing. They will be received and actioned via 12 Horizon and will be distributed more quickly, 13 potentially only days after an error is recorded. 14 "The analysis has also identified requirements to 15 more tightly control and police the use of the suspense 16 account within the branch accounts. Only a limited 17 subset of the existing suspense account products will be 18 retained. The contractual requirements for agents to 19 make good unknown errors in branch accounts will be used 20 instead." 21 What did you understand that second part of 6.7 to 22 mean? 23 A. I understood that they wanted to limit the use of the 24 suspense accounts. My understanding was that there were 25 a number of suspense accounts and postmasters could 57 1 actually post money to the suspense accounts for some 2 sort of business reason which I never really did 3 understand, and they wanted to reduce the amount of 4 money that was being posted into the suspense accounts. 5 Now, exactly why they wanted to do that, that was part 6 of Post Office's business for trying to save money, and, 7 now, in what way it did that, I didn't really 8 understand. 9 Q. Were you told why they wanted to "more tightly control 10 and police the use of the suspense account"? 11 A. Not in so many words. What I believe was behind a lot 12 of this was they were trying to reduce the number of 13 staff that they had in Chesterfield back office, which 14 were monitoring things like this. So they were trying 15 to actually automate things and reduce the flexibility 16 on the postmasters which I thought was a business 17 decision for Post Office and nothing for me to get 18 involved in. 19 Q. Were you told that the suspense account was considered 20 by the Post Office to be a vehicle for postmasters to 21 hide fraud? 22 A. I can't remember. There were various discussions about 23 behaviours of postmasters and things like that but -- 24 and I know that Post Office were quite keen to get rid 25 of the suspense account. In fact, they didn't get rid 58 1 of it altogether but I know they were keen to reduce its 2 usage and there was, indeed, some reduction on its 3 usage. 4 Q. Given that reduction on usage, if there was 5 a discrepancy in a subpostmaster's account, after 6 Project IMPACT, the subpostmaster would have to accept 7 it and pay it before they could roll over into the new 8 trading period, correct? 9 A. That is what I understood the Post Office wanted the 10 system to do, so yes. 11 Q. And, if they didn't pay, the Post Office would, in the 12 language of this document, enforce contractual 13 requirements; did you understand that? 14 A. I didn't really consider that sentence particularly and 15 I'm not sure that I really understood what that actually 16 meant. And my focus was on what exactly was it I had to 17 actually implement in the way that Horizon worked, and 18 what I took from that was I had to simplify the menus 19 for the suspense accounts, which was actually just 20 a reference data change. 21 Q. Given some of the suspense account products were to be 22 removed and the consequence of that would be for 23 subpostmasters to have to accept discrepancies and pay 24 them before they could roll over into new trading 25 periods, was there any consideration, to your knowledge, 59 1 of the importance of the accuracy of the figures 2 produced by Horizon? 3 A. Well, I always felt it was important that the figures 4 produced by Horizon were accurate and I wasn't aware 5 there was any doubt or challenge to that at that stage. 6 So, therefore, I didn't see that as being an issue as 7 part of what we were doing here. 8 Q. So is the answer: no, that issue didn't even arise, that 9 the change that we're making, the removal of this 10 facility, means that it is very important that the 11 accounts that are to be signed off by subpostmasters are 12 100 per cent accurate? 13 A. Yes, and I always felt that they should be 100 per cent 14 accurate. 15 Q. Of course they should be but were you satisfied that 16 they were 100 per cent accurate the whole time? 17 A. I don't know that I'd actually thought it through in 18 those sort of the terms at the time. 19 Q. We've heard evidence from Susan Harding, who told us 20 that the business drivers for the IMPACT Programme were 21 to reduce the costs to the business, increasing 22 accounting efficiency and reducing the losses to the 23 Post Office. Do you agree that they were the business 24 drivers that you were told about? 25 A. I can't remember the details of the drivers but, 60 1 certainly, I know that Post Office were trying to reduce 2 their costs. I've not twigged so much about the 3 reducing losses but they were certainly trying to reduce 4 their costs on the back end. 5 Q. Were you told that a business driver was to improve the 6 Post Office's debt recovery from subpostmasters? 7 A. I don't specifically remember that one but I can't say 8 that I hadn't seen it in a document. 9 Q. You tell us in your first witness statement -- no need 10 to turn it up, it's paragraph 41 -- that the effect of 11 the project was to reduce staff costs? 12 A. Yes. 13 Q. Is that what you were told, as the driver, or as being 14 the driver -- 15 A. That's my memory -- 16 Q. -- namely: this is all about reducing the number of 17 people at Chesterfield or elsewhere that need to be 18 looking at suspense accounts, rather than the reduction 19 of losses and the improvement of debt recovery? 20 A. That's my memory but I can't be definitive because it's 21 an awfully long time ago now. 22 Q. You tell us in your witness statement -- paragraph 43 of 23 your first witness statement -- that the decision to 24 remove the ability of subpostmasters to post 25 discrepancies to a suspense account was because the Post 61 1 Office "took what I assumed to be a business decision to 2 remove this functionality". Was the only thing you were 3 told about the business decision that it was to reduce 4 staff costs? 5 A. I don't know that I would have taken too much notice as 6 to why they were doing things. My main focus on -- was 7 what was it they wanted changed in the system, not why 8 they were doing it. 9 Q. Ms Harding told us that it was agreed during the design 10 of the IMPACT Programme that the suspense account would 11 be removed because, historically, it was used by 12 subpostmasters to hide discrepancies in their accounts. 13 Were you told that? 14 A. I can't remember explicitly, exactly, what I was told or 15 what I wasn't told. 16 Q. What understanding did you have as to how the local 17 suspense account had worked historically? 18 A. We need to be careful here about terminology because one 19 of the things introduced here was something called the 20 "local suspense account". 21 Q. I'm talking about at or before 2005, what was your 22 understanding as to how the local suspense account 23 worked? Not the -- 24 A. That's where I'm getting confused. We introduced a new 25 concept as part of IMPACT called the local suspense 62 1 account so, as far as I'm concerned, there wasn't 2 a local suspense account prior to IMPACT. But I think 3 I know what you're referring to. 4 Q. Yes. What's the answer? 5 A. I didn't know how it was used or what it was used for. 6 I just assumed this was part of the business processes 7 that happened. I think I understood that, in order to 8 post something to local suspense account, you did need 9 to have permission. How that actually operated -- 10 presumably through NBSC -- but exactly how that operated 11 I didn't know and it didn't really concern me, 12 particularly. 13 Q. Do you remember being told by the Post Office that 14 subpostmasters "hid their discrepancies" in the suspense 15 account? 16 A. I can't remember, sorry. 17 Q. Do you remember any evidence of subpostmasters hiding 18 their discrepancies in suspense accounts? 19 A. I don't think I would have seen any evidence of anything 20 like that, one way or the other. 21 Q. Were you given any information as to why subpostmasters 22 might hide their discrepancies in suspense accounts? 23 A. Not that I can remember. As I say, I wasn't that 24 interested in the whys. I was interested in what it was 25 that I was being asked to do in the system, not why it 63 1 was being done. 2 Q. You had no professional curiosity in that? 3 A. I mean, I went to a number of workshops about these 4 things and I sat through number of workshops, but 5 exactly what was discussed in them, because this is over 6 20 years ago now, I just can't remember the details. So 7 I may have been aware of some of those things at the 8 time, I just have no memory of it now, I'm afraid. 9 MR BEER: Thank you. 10 Sir, that's an appropriate moment for the morning 11 break. Can we break until 10.20 (sic), when we'll turn 12 to Mr Castleton's case. 13 SIR WYN WILLIAMS: 11.20, I think. 14 MR BEER: 11. Yes, quite right, sir. 15 SIR WYN WILLIAMS: All right then, 11.20. 16 (11.08 am) 17 (A short break) 18 (11.20 am) 19 MR BEER: Good morning, sir, can you see and hear us? 20 SIR WYN WILLIAMS: Yes, I can. 21 MR BEER: Thank you very much. 22 Mr Jenkins, thank you. Mr Jenkins, can we turn to 23 Lee Castleton's case, please. 24 A. Yes. 25 Q. Can we start, please, by looking at FUJ00152573 and by 64 1 going to page 13, please. Thank you. You can see here 2 a letter dated 18 November from Bond Pearce -- who were 3 the Post Office's solicitors in the case against 4 Mr Castleton -- addressed to Fujitsu; can you see that? 5 A. I can. 6 Q. If we just read the first page in summary, it says to 7 Fujitsu, in relation to Mr Castleton's branch at Marine 8 Drive: 9 "We act on behalf of the Post Office. [Between 10 those dates] Mr Castleton was a subpostmaster at the 11 Marine Drive Post Office. He was strictly responsible 12 for the safe custody of cash and stock and was obliged 13 to make good all losses caused through his own 14 negligence, carelessness or error and losses of any kind 15 caused by his assistants. 16 "Between [some dates] net losses of [£27,000-odd] 17 occurred at the [branch]." 18 He was suspended and then dismissed: 19 "The Post Office has now issued a claim against him 20 to try to recover these net losses. [He's] issued 21 a counterclaim for wrongful termination of his 22 contract." 23 Then 2, "Mr Castleton's defence": 24 "[His] case is that any shortfall is entirely the 25 fault of problems with the Horizon computer and 65 1 accounting system at Marine Drive ... and that the [Post 2 Office] wrongly terminated his ... contract in respect 3 of which he suffered losses not exceeding £250,000." 4 Three documents are attached. In fact, the second 5 one has some attachments itself: 6 "Bentley Jennison [they were the authors of the 7 report in the second attachment] state that deficiencies 8 have probably been brought forward despite the fact that 9 they have been entered onto the suspense account entry. 10 They suspect this is because the Horizon system, despite 11 the suspense account entry, has failed to recognise the 12 entry on the daily snapshots. They have drawn this 13 conclusion through looking at the discrepancy of 14 [£3,500-odd] on 26 February 2004. They then suggest 15 this double accounting could have continued over 16 a number of weeks and that as such, [his] defence 17 'appears to hold potential merit based on the limited 18 documentation' they have so far reviewed ... 19 "Mr Castleton believes that if he can obtain further 20 documents, such as the daily snapshots, he will be able 21 to undertake a manual reconciliation of the cash account 22 in order to substantiate his belief that the losses are 23 not real but attributable to computer error. We attach 24 an email from Fujitsu to Richard Benton at the Post 25 Office dated 5 May in which Fujitsu state: 'It is 66 1 possible that they are not accurately recording all 2 transactions on the system' and that there was no 3 evidence whatsoever of any system problem." 4 Over the page: 5 "Please could you review W Mr Castleton's experts' 6 reports and prepare a formal report dealing with the 7 following points ... 8 "1. We need to explain to a judge who will know 9 nothing about Horizon exactly how it works [et cetera]. 10 "2. Precisely what steps Fujitsu took to examine 11 the Horizon system at the Marine Drive [branch]. 12 "3. Whether there have been any similar or serious 13 problems with the Horizon system at [Marine Drive] since 14 [his] suspension and dismissal. 15 "4. Whether you believe the suggestion put forward 16 by Mr Castleton's experts is likely to be correct ... 17 "5. If there have been any human errors in 18 recording the transactions, could an explanation be 19 that: 20 "(a) There was nothing wrong with Horizon because it 21 simply reflected the information entered onto it; but 22 "(b) If staff entered the wrong numbers into Horizon 23 there may have been no real loss (even though Horizon 24 would show a loss), because there could be human error 25 in accounting accurately recording transactions. 67 1 "If so, would that be a likely explanation? 2 "6. Any other information you believe may be 3 relevant." 4 Then scroll down, please. Paragraph (4), "Duty to 5 the Court": 6 "As a result of the instruction you may be asked to 7 give evidence before the Court. Whilst the [Post 8 Office] will be liable to pay your fees, in preparing 9 your report and giving evidence, your overriding duty 10 will be to help the Court on the matters within your 11 expertise. 12 "You agree to meet the requirements of the Civil 13 Procedure Rules Part 35 [and a copy is enclosed] and 14 that your report will: 15 "1. Be addressed to the Court and not to the Post 16 Office ... 17 "2. Confirm that you understand your duty to the 18 Court and that you have complied and will continue to 19 with that duty. 20 "3. Contain a statement setting out the substance 21 of all material facts and instructions (whether written 22 or oral) on the basis on which your report is written. 23 This statement should summarise the facts and 24 instructions given to you which are material to the 25 opinions expressed in the report or upon which those 68 1 opinions are based and if any of the facts are within 2 your own knowledge, which they are. 3 "4. Contain a chronology of relevant events; 4 "5. Contain a statement of Truth in the following 5 form [and gangs it's set out]. 6 "6. You should note that proceedings for contempt 7 of Court may be brought against you if you make a false 8 statement and report verified by a Statement of Truth 9 without an honest belief it was true. 10 "7. It must contain a declaration that the report 11 has been prepared in accordance with the Code of 12 Guidance on Expert Evidence [which was also enclosed]. 13 "8. Give ... qualifications. 14 "9. Give details of any literature or any other 15 material which you rely on in making the report. 16 "10. So who carried out any test or experiment 17 [et cetera]. 18 "11. Give the qualifications of the person who 19 carried out any such tests [et cetera]. 20 "12. Where there is a range of opinion on the 21 matters dealt with [that that change should be 22 summarised]. 23 "13. Give reasons for your own opinion. 24 "14. Contain a summary of the conclusions reached 25 including any qualifications ... 69 1 "Given the fundamental importance of meeting these 2 requirements, you should endeavour in your report to be 3 not only accurate but complete. You should mention all 4 matters which you regard as being material to the 5 opinions you express and draw the Court's attention to 6 any matter to which you are aware which might adversely 7 affect the validity of those opinions. This applies in 8 relation to the factual matters to which you refer and 9 also to the opinions which you express. 10 "You should not include in your report anything that 11 is suggested to you by anyone without forming your own 12 independent view. 13 "If, on reading the report of any other expert in 14 this matter, or for any reason, you consider, at any 15 stage, that any existing report of yours requires 16 correction or qualification you will immediately notify 17 us in writing." 18 Then (5): 19 "In performing all your duties for which the client 20 will pay, you will owe a duty to the client to act with 21 the professional standards of skill, care and diligence 22 adhered to by experienced and competent consultants 23 acting as expert witnesses. 24 "You will take reasonable care of any documents, 25 [et cetera] 70 1 "You confirmed that you: 2 "1. Are an independent party and not the client's 3 employee or agent, other than at the material time 4 Fujitsu was responsible for looking after the Horizon 5 system; 6 "2. Know of no reason why you should not act as 7 a witness for the Post Office in relation to the 8 dispute; 9 "3. Will advise us in writing immediately if [there 10 is] any conflict between your interest and the Post 11 Office's interests ..." 12 We are going to see in a moment that you commented 13 on some the passages from this letter. Was this letter 14 sent to you? 15 A. I've no recollection of that. I recognise the last bits 16 now as being what an expert's duties are, but the first 17 time I was made aware of what those were when I was 18 first put in touch with solicitors in 2020/2021, as part 19 of the police investigation into my conduct. 20 Q. If the letter had -- you say you've got no recollection 21 whether or not it was sent to you? 22 A. I mean, I think I would have remembered if it had been 23 sent to me because I can see there that it's clearly set 24 out what the duties are and I wasn't aware of any of 25 those duties, until my solicitor pointed them out to me 71 1 when we first got engaged at the end of 2020. 2 Q. When you are giving your answers, Mr Jenkins, you should 3 refrain, unless you're doing it deliberately, from 4 disclosing communications between your current solicitor 5 and you, which involve either the seeking of legal 6 advice by you or the provision of legal advice to you. 7 A. Okay, sorry. 8 Q. Because, if you do that, you might be said to have 9 waived your privilege -- your legal professional 10 privilege in those -- and I or somebody else can ask you 11 about those. 12 A. Okay. I'm sorry. 13 Q. So I'm trying to help here. 14 A. Yes, I understand. 15 Q. If you wanted to give an answer of the kind that you've 16 just given, it would be without waiving privilege by 17 saying, "I only learnt that in 2020 or 2021". 18 A. Okay, yes. 19 Q. Okay? 20 A. Yeah. 21 Q. It is, of course, open to you to waive privilege -- 22 A. Yeah. 23 Q. -- but I don't want you to do so accidentally. 24 A. No, I don't want to do so either. 25 Q. Okay. You're reasoning that you didn't see this letter 72 1 because, if you had have seen the letter, you would have 2 learnt earlier than you say is the case about the 3 existence of an expert witness's duties? 4 A. Correct. 5 Q. If you had received this letter, presumably you would 6 have read all of it? 7 A. I would certainly have skimmed through all of it, yes. 8 Q. You wouldn't have read half the letter: the first page 9 but not the second page? 10 A. I don't think so. I don't recognise any of it. 11 Q. You agree that Bond Pearce have set out here, just 12 reading it, the duties of an expert witness, what they 13 entail, very clearly, haven't they? 14 A. I agree, yes. 15 Q. In a very easy-to-understand way, haven't they? 16 A. Yes. 17 Q. Can we turn, please, to FUJ00154713. If we look at the 18 email at the foot of that page first, thank you, it's 19 an email dated 8 December 2005 about Marine Drive from 20 you to Brian Pinder? 21 A. Yes. 22 Q. What was Brian Pinder's function at this time? 23 A. He was the Manager of the Security Team, I believe. 24 Q. Just tell us what the Fujitsu Security Team did? 25 A. They were responsible for overall security, so making 73 1 sure that the systems were secure, but they were also 2 involved with communicating with Post Office in terms of 3 providing what was called prosecution support or 4 litigation support. So they were providing things like 5 ARQ data and support of things like this. 6 Q. Can we just see how this email works because it's 7 slightly difficult to understand which bit is you 8 speaking, which bit is Mr Pinder speaking and which bit 9 is the questions from the Bond Pearce letter of 10 18 November 2005. Okay? 11 A. Yes, and I suspect the original would have been slightly 12 easier to decode because it would have been -- 13 Q. Coloured? 14 A. -- multicoloured, yes. 15 Q. Yes, we don't have the native version -- 16 A. Yes, I understand that. 17 Q. "Brian", I think that's you writing, isn't it? 18 A. I think it is. 19 Q. Then the words in dark black or darker black: 20 "I have been asked by the Fraud Investigation Team 21 to answer several questions, to assist an enquiry 22 regarding a [Post Office] employees (subpostmaster) 23 alleging that the Horizon system may have lost his 24 money. I can manage 5 of the 6 questions but wonder if 25 you could explain or provide an answer or opinion to the 74 1 following question, the answers are being directed back 2 to Graham ..." 3 Would you understand that to be Graham Ward? 4 A. I do now. I'm not sure I would have done at the time. 5 Q. "... to formalise our response." 6 Is that part in black there something that Mr Pinder 7 wrote? 8 A. I think so. Looking at that page, I think the bits that 9 I would have written is the "Brian" at the top and "This 10 is certainly true", which is a bit further down, I can't 11 see certain but that's my guess. 12 Q. Trying to decode it, then, that second paragraph then, 13 after the word "Brian", looks like it's Mr Pinder having 14 sent that to you and then you've cut it into your email 15 back? 16 A. Correct. 17 Q. Then there's a part in lighter text, almost grey: 18 "If there have been human errors in recording the 19 transactions, could an explanation be that: 20 "(a) There was noting wrong with Horizon, because it 21 simply reflected the information entered on to it; but 22 ..." 23 Just stopping there. Is that likely to be what 24 Mr Pinder has included in his original email to you? 25 A. Yes, I believe so and I recognise that as being in the 75 1 previous document which you showed me, yes. 2 Q. One of the six questions -- 3 A. Yes. 4 Q. -- in the Bond Pearce letter -- 5 A. Yes. 6 Q. -- of 18 November. Then you reply, is this right: 7 "This is certainly true (ie Horizon simply reflects 8 the information entered into it)." 9 A. Correct. That would have been my comment. 10 Q. So, in relation to if there have been human errors in 11 recording the transactions could an explanation be there 12 is nothing wrong with Horizon because it simply 13 reflected information entered into it, and you say, 14 "This is certainly true (ie Horizon simply reflects the 15 Horizon entered into it"? 16 A. Yeah, that is what I would have said at the time and 17 I stand by that today as well. 18 Q. That is a possible explanation? 19 A. Yes. 20 Q. Were you able to say that without investigation of the 21 facts of this case? 22 A. Yes, because Horizon is going to record what is actually 23 entered into it, whether that reflects what happened in 24 the real world, and if you do a typo when you put 25 information into Horizon, then it's going to accurately 76 1 record that typo. 2 Q. Then part (b) of the original question: 3 "If staff entered the wrong numbers into Horizon 4 there may have been no real loss (even though Horizon 5 would show a loss), because there could be a human error 6 in accurately recording transactions." 7 Then scroll down. You say, I think, is this right: 8 "Again, this could be true. However if there is 9 some sort of misentering of data into Horizon, then 10 there would be another corresponding error which should 11 be picked up, eg as a stock error or some AP Client 12 being credited with an incorrect amount. Also, such 13 error should show up as part of the balancing process." 14 Again, you're there saying this could be 15 an explanation; is that right? 16 A. I am. 17 Q. Then question 3 was: 18 "If so, would that be a likely explanation?" 19 You give quite a long reply, and you say: 20 "It is a possible explanation, but without doing 21 a detailed analysis of everything that has gone on in 22 the branch it is difficult to speculate as to what has 23 happened. Certainly the most likely explanation, is 24 misoperation or fraud. However I appreciate that that 25 is not sufficient for a prosecution. Without 77 1 understanding what exactly is alleged (by [Post Office] 2 and the subpostmaster), I don't feel I can add anything 3 further." 4 So are you saying there, firstly, that you need to 5 see some underlying data in order to say what is likely 6 to have happened? 7 A. Yes. 8 Q. Because, otherwise, it's mere speculation -- 9 A. Correct. 10 Q. -- ie the answers you've given to parts (a) and (b) are 11 speculation? 12 A. Yes. 13 Q. You say that "It is certain that the most likely 14 explanation is misoperation or fraud". Why was it 15 certainly the case that the most likely explanation was 16 misoperation, ie by the subpostmaster, I think, or 17 fraud? 18 A. I'm not sure why I would have said that at the time. 19 I mean, looking back now, I think that's just one of 20 many different options but I accept those are the words 21 I used at the time. But I think I would stand by the 22 fact that I would need to look into exactly what had 23 happened. 24 Q. Lots of people have said to me "I accept that they're 25 the words that I used at the time", and obviously they 78 1 are -- 2 A. Yeah, sure. 3 Q. -- because they are, they're on the screen? 4 A. Yeah. 5 Q. Why did you use them? 6 A. I don't know, is the simple answer. 7 Q. Did they disclose the way you were thinking in December 8 2005, ie the most likely cause of what is being 9 described here is mistake by the subpostmaster or fraud 10 by the subpostmaster? 11 A. My feeling was that that was more likely than an error 12 in Horizon, yes. But I'm including the option of it 13 being a mistake. 14 Q. But you can't say that without looking at the data, can 15 you? 16 A. Which is why I go on to say that I needed to look at the 17 data. 18 Q. You say you appreciate that that is not sufficient for 19 a prosecution. Had you understood this to be a criminal 20 case? 21 A. Well, the fact that the request was coming from the 22 Security Team suggested that there probably was. 23 Q. Did you, at this stage, understand the difference 24 between criminal and civil proceedings? 25 A. No. 79 1 Q. Did you know that there were different types of courts, 2 some which tend to deal with money and some which tend 3 to deal with crime, sending people to prison and the 4 like? 5 A. Not really, no. 6 Q. Did you understand that a prosecution tends to refer to 7 criminal cases? 8 A. I'm not sure I really thought about it. As I say, I'd 9 not really appreciated any real difference and so I was 10 probably using fairly loose language. After all, this 11 is an internal email. 12 Q. There's then a part of the text which says, "I have 13 spoken with Dave Baldwin", et cetera. 14 A. Yes. 15 Q. Is that paragraph again Mr Pinder? 16 A. That I would expect it to be, yes. 17 Q. "I have spoken to Dave Baldwin on this and he 18 recommended I speak with you. I'll also be liaising 19 with him on my final response as we are fundamentally of 20 the opinion that Horizon does not 'eat money' merely 21 accounts for its placement." 22 Then you finish, I think this is you at the end 23 saying: 24 "Happy to discuss further if you want to look at 25 some more specific areas of Horizon." 80 1 A. That sounds like me, yes. 2 Q. The answer to the third question is an important one, 3 isn't it, because it says the answer to the first two 4 are mere speculation, without looking at the data? 5 A. Yes. 6 Q. Can we go to the top of the page, please. Later that 7 day, Mr Pinder writes back to you, and says: 8 "Gareth 9 "[For your information] 10 "This is the response I intend to send to Graham re 11 your question if that's okay." 12 Then "Comments from Gareth Jenkins", question 1: 13 "If there had been human errors [then] (a) ..." 14 You can see your text is faithfully reproduced, yes? 15 A. Except that the last point has not been faithfully 16 reproduced. 17 Q. No, no, just look at (a) to start with. 18 A. Yes. 19 Q. That has been faithfully reproduced, hasn't it? 20 A. Yes. 21 Q. Then look at (b), that's been faithfully reproduced, 22 hasn't it -- 23 A. Yes. 24 Q. -- your answer, yes? 25 A. Yes. 81 1 Q. Then (c) -- or I'm going to call it (c) -- the third 2 part of the question: 3 "If so, would that be a likely explanation?" 4 He has written: 5 "Not able to comment on this." 6 A. Yes. 7 Q. The detailed and important answer you gave to the third 8 part of the question has been changed to "No comment", 9 hasn't it? 10 A. Yes. 11 Q. Do you know why that was? 12 A. No. 13 Q. It's an important qualifier to the answers to the first 14 two questions, isn't it? 15 A. I realise that now. I probably didn't appreciate that 16 at the time. 17 Q. Why not? 18 A. I don't know, is the simple answer. 19 Q. Can you think of a good reason why the careful 20 explanation you've given to question 3 has been changed 21 into "No comment"? 22 A. No, it may have been considered that it was just too 23 complicated an answer but -- 24 Q. What's complicated about an answer "You need to look at 25 the data to give an accurate picture"? 82 1 A. I don't know, I think you'd have to ask Brian Pinder. 2 Q. I'm asking you at the moment because what -- 3 A. Yeah. 4 Q. -- this is is an email to you, essentially saying, "Is 5 it okay if I send this as your response"? 6 A. I probably hadn't picked up the fact that it had taken 7 away the qualification that I had put in the original 8 email. 9 Q. Do you read documents carefully? 10 A. I try to, yes. 11 Q. You're saying you wouldn't have -- or you obviously 12 didn't pick this up at the time? 13 A. I clearly didn't pick that up on the time. Why he was 14 changing that to a "no comment" I wasn't particularly 15 clear. Perhaps I should have taken it up to him as to 16 why he was doing that but I assumed that he knew why -- 17 what he wanted to respond, and unable -- maybe he felt 18 that I wasn't in a position to make the comment on that 19 last part. I just don't remember. 20 Q. But you had, hadn't you? 21 A. I had commented on it, yes, but why he'd decided he 22 didn't want to pass that on, I don't know. 23 Q. I'm not so much asking for his take; I'm asking why you 24 assented to this approach by not saying, "Well, hold on, 25 you've missed out an important qualifier"? 83 1 A. Perhaps I didn't realise at the time how important that 2 qualifier was. 3 Q. Can we go back to FUJ00152573. 4 SIR WYN WILLIAMS: Before we do, Mr Beer, I'm slightly 5 puzzled -- and it may just be me -- about how this email 6 the one that's on the screen, begins. "For your 7 information", and then this: 8 "This is the response I intend to send to Graham re 9 your question, if that's OK." 10 Now, to me, that reads as if Mr Pinder is going to 11 send this response to Graham, if it's okay with 12 Mr Jenkins, because it's the question that Mr Jenkins 13 has asked him. Am I misreading this? 14 MR BEER: No, that's exactly right, sir, I think. 15 SIR WYN WILLIAMS: Is that consistent with the 16 interpretation that was given to the preceding emails, 17 which suggested that it was Mr Pinder who was seeking to 18 ask questions of Mr Jenkins, rather than Mr Jenkins 19 asking questions of Mr Pinder; do you see my point? It 20 may be that I'm just misunderstanding. 21 MR BEER: Sir, I think you may have misunderstood the first 22 email that we looked at, if we just go down to that. 23 SIR WYN WILLIAMS: Yes. 24 MR BEER: As I read this, there must have been an earlier 25 email that we haven't got in which Mr Pinder sent that 84 1 first paragraph "I've been asked by the Fraud 2 Investigation Team", and then the second paragraph which 3 says, "If there have been human errors", if that's 4 marked, and then the paragraph at (a) if that can be 5 marked, and the paragraph at (b), and then Mr Jenkins 6 has replied in the darker text. 7 SIR WYN WILLIAMS: Yes. Well, that's as I understood what 8 you were explaining, and that did or could make sense, 9 but it's the juxtaposition of that with the beginning of 10 the next email that I'm struggling with. Anyway, I'm 11 sorry to interrupt your questioning of Mr Jenkins. 12 Maybe it'll all become clear to me. 13 MR BEER: I'll try, sir. 14 If we go back up to the top page, do you understand 15 that this is Mr Pinder saying to you everything below 16 the words "Comments from Gareth Jenkins" and including 17 the words "Comments from Gareth Jenkins" is what I, 18 Brian Pinder, are going to send to Graham -- 19 SIR WYN WILLIAMS: Yes, I do, but it relates to, and it's 20 just those three words "re your question", and it's the 21 other way around, isn't it, on the first two emails? 22 MR BEER: Sir, I think what that -- your -- I don't want to 23 give evidence here, but your -- 24 SIR WYN WILLIAMS: Maybe I should just ask Mr Jenkins. 25 What do you think -- 85 1 A. I think the "re your question" -- my understanding is 2 that Graham Ward had asked Brian Pinder six questions. 3 He'd identified one of those questions for me to respond 4 to, namely question 6, I think it was, and, therefore, 5 what he's saying is, "In response to your -- the 6 question that I've allocated to you, this is the answer 7 I intend to send to Graham Ward in response to that 8 specific question". 9 SIR WYN WILLIAMS: So your interpretation of the phrase 10 "your question" is not that you asked the question, it's 11 the question that he allocated to you? 12 A. Correct. That's how I understand it -- 13 SIR WYN WILLIAMS: Right. 14 A. -- and that's how I would have understood it at the 15 time. 16 SIR WYN WILLIAMS: Fine. Well, then I follow. Thank you. 17 Sorry -- 18 MR BEER: So that should read, "I intend to send to Graham 19 the question that has been allocated to you, if that's 20 okay"? 21 A. That is how I've read it at the time, yes. 22 SIR WYN WILLIAMS: Fine. Then I do understand now. Thank 23 you. 24 MR BEER: Thank you. 25 Can we go back, please, to FUJ0015 -- 86 1 A. Before we leave that, can we go down, please, to the -- 2 my response to the third question? 3 Q. Yes, absolutely. 4 A. It starts -- again, it said, "It is a possible 5 explanation, but without detailed analysis it's 6 difficult to speculate". So I think what Brian had 7 interpreted from that, therefore, I wasn't going to 8 comment on it and that's why I would have been happy 9 with the "I'm not happy to comment". I suspect that is 10 probably what -- why I was happy with that being 11 replaced with "unable to comment" because I'm 12 speculating there and, therefore, it is not a definitive 13 answer. 14 Q. Doesn't that overlook completely -- 15 A. But I accept the fact that the qualification would have 16 been useful information to have provided. 17 Q. Doesn't it completely overlook the fact that the main 18 point to question 3 is you need to look at the data? 19 A. Yes, yes, I accept that now. But that may well have 20 been why I was happy to allow the change to have been 21 made at the time. 22 Q. Okay. Can we go back, please, to FUJ00152573. This is, 23 if we just look at the foot of the page, Brian Pinder's 24 document -- if we go up, please -- addressed to Graham, 25 which is going to be Graham Ward. This is the response 87 1 of Mr Pinder. He says: 2 "In response to your email request dated 28 November 3 concerning the questions raised at paragraph (3) of the 4 above document ..." 5 The "above document" being the Bond Pearce letter 6 that we've looked at. 7 You will see the answer to question 1. Question 2, 8 it refers to, in the answer, a response from Anne 9 Chambers, which is attached. Then, if we scroll down, 10 please, to question 5, you'll see that the full question 11 in three parts is set out and the reply from Mr Pinder 12 is "Gareth Jenkins' response: attached". 13 If we go to page 3, please. So this was the 14 attachment to Mr Pinder's document. 15 A. Yes. 16 Q. Can you see that it's essentially his email back to you, 17 the "FYI" email -- 18 A. Yes. 19 Q. -- including your signature block at the bottom, which 20 he did, in fact, include in his reply to you? 21 A. Yes. 22 Q. So the shortened reply has been sent back to the Post 23 Office? 24 You were able to comment, weren't you, whereas your 25 reply has gone back to the Post Office saying that 88 1 you're not able to comment? 2 A. Yes. 3 Q. Was the reason for that because it was thought that 4 an investigation of what the subpostmaster was saying 5 might not be in Fujitsu's best interests? 6 A. I don't know what was the reason for that change. 7 I would have been quite happy to investigate and look at 8 exactly what had happened in the branch and, in fact, 9 I believe, at some later stage, I did assist Anne in 10 doing some investigations there. 11 Q. Was the reason for the different reply that 12 an investigation might reveal issues that Fujitsu didn't 13 really want to speak about? 14 A. Not that I'm aware of. 15 Q. Thank you. That can come down. 16 I think you subsequently attended a conference on 17 6 June 2006 with representatives of Bond Pearce and 18 others; do you remember? 19 A. I remember seeing the documents about it and have vague 20 recollections of the meeting. 21 Q. Let's look at the documents, please, POL00071427. We 22 can see this is an attendance note in the case of Lee 23 Castleton, dated 6 June 2006. We were previous looking, 24 to give you some context, at November and December -- 25 A. Yes. 89 1 Q. -- 2005. This is at Fujitsu's HQ in Bracknell. In 2 attendance were, if we scroll down, solicitors, the 3 first three people, then Brian Pinder, Peter Sewell, 4 Andy -- I think that's Dunks -- 5 A. Yes. 6 Q. -- I think he's called "Dunce", Anne Chambers, Naomi 7 Ellis and you, a "Distinguished Engineer of Fujitsu"? 8 A. Yes. 9 Q. I'm not going to go into the detail of this attendance 10 note. Do you recall any discussion in this meeting 11 about you or Anne Chambers giving evidence? 12 A. I think there was discussion about evidence being given 13 but I can't remember the details of it and I do know 14 that Anne eventually ended up giving evidence in the 15 case and I didn't. 16 Q. You remember the second page of the Bond Pearce letter 17 and onto the third page of the Bond Pearce letter -- 18 A. Yes. 19 Q. -- an extensive explanation of the duties of an expert? 20 A. Yes. 21 Q. Do you know what had happened to that? 22 A. No idea. 23 Q. Were those issues discussed at this conference? 24 A. I have no recollection of any discussion of those there. 25 As I say, all I can vaguely remember is the fact that 90 1 there was a meeting and I met a couple of solicitors. 2 Q. Had the guidance that Bond Pearce had given about the 3 provision of expert evidence by a Fujitsu employee been 4 forgotten by this stage? 5 A. I don't know whether it had been forgotten or what but 6 I don't remember any mention of it. 7 Q. In any event, your recollection is that that wasn't gone 8 over again in the context of either you or Anne Chambers 9 giving evidence? 10 A. No. 11 Q. Thank you. Can we turn -- this is 6 June 2006 -- to 12 three days later, to 9 June 2006. FUJ00154722, and look 13 at the email at the bottom part of the page. If we 14 scroll up just so we can see from and to, thank you. 15 From the solicitor at Bond Pearce, Stephen Dilley, to 16 Mr Pinder, copied to Graham Ward and Mandy Talbot. So 17 not you at this stage. 18 A. Yeah. 19 Q. "... I would like to thank you and your team for finding 20 time to meet us on Tuesday." 21 That would be the date of the conference we just 22 looked at: 23 "We found the meeting to be really worthwhile, 24 helpful and productive. We will circulate a note of the 25 meeting shortly, just for everybody's record. 91 1 "In the meantime, I attach a letter received on 2 7 June from Mr Castleton's solicitors. I have 3 highlighted the two paragraphs that I would like you to 4 have a quick look at. Basically they state the Horizon 5 system is equip with a facility via which its operating 6 software can be updated remotely via the connection to 7 the [Post Office's] head office and that the system 8 sometimes went offline when there are software updates. 9 They have therefore asked for documents relating to the 10 updating of the software of the system. My view is that 11 they are making a stab in the dark, but please could you 12 ask your team whether they would have or could obtain 13 any records of software updates?" 14 So this is a question about remote accessing of the 15 system. 16 A. A sort of remote access. I think I have explained 17 different types of remote access in one of my witness 18 statements. 19 Q. We're going to come to that, probably tomorrow, but 20 a species of remote access? 21 A. Yes. 22 Q. Now, if we look, please, at FUJ00152605, and if we 23 scroll down, please, and again, and again. We'll see 24 that email -- if we just scroll down a bit further, 25 we'll see the email we've just looked at. 92 1 A. Yes. 2 Q. Then if we scroll up, please, Brian Pinder sends that on 3 to you and Penny Thomas: 4 "I would like to run this by you first as I hope you 5 might be able to either answer the query (hopefully) or 6 point me in the right direction?" 7 Then up, please. Your reply of 9 June 2006, you 8 say: 9 "It is correct that we do update the software 10 remotely and that a software update could result in the 11 desktop being closed and restarted. However, even then 12 I don't see how it would result in losses as is being 13 claimed. Also software updates are relatively rare and 14 normally only take place between 20.00 and 01.00 and 15 again between 4.00 and 7.00 am with occasional 16 extensions at the weekends. This puts them well outside 17 the normal operational hours of Horizon. 18 "I think he is clutching at straws. However I guess 19 we do need to cover this is angle. 20 "I would expect the system to record details of all 21 software updates and when they took place at which bans. 22 However I don't know how long these are retained and 23 whether they are included in the Audit Trail so you can 24 check out the history for Marine Drive in 2004." 25 Then you suggest a place to start. 93 1 In the provision of this information, which is 2 a reply about a species of remote access, had the 3 guidance in the Bond Pearce letter been drawn to your 4 attention about the provision of expert evidence? 5 A. No. I just had the question as shown in the email trail 6 that you are showing me. 7 Q. So that letter of 18 November never made its way through 8 to you? 9 A. No, I'm sure that it couldn't have done because I would 10 have done things differently, not necessarily in this 11 case but certainly in later cases, if I'd been aware of 12 those responsibilities. 13 Q. Can we move on, please, to FUJ00154727, and start with 14 page 2, please. An email from Mr Ward to Mr Pinder and 15 Mr Sewell; can you see that? 16 A. Yeah. 17 Q. "Stephen Dilley (representing the Post Office) is asking 18 if it is possible that someone undertakes an analysis of 19 the figures recorded on transaction logs ... It appears 20 that the solicitors for Castleton are saying that 21 they've compared the transaction logs with the cash 22 accounts for week 42 themselves ... and that they don't 23 match. They conclude that Horizon is therefore only 24 recording half the [transactions]." 25 Then reading the second and third paragraph: 94 1 "Given this is a 'test' case and that the integrity 2 of the Horizon system is being challenged, my own 3 opinion is that this exercise should be completed by 4 an 'expert' from Fujitsu. However, it will obviously be 5 extremely time consuming as all the figures recorded 6 will need to be methodically and carefully analysed. It 7 is also important we complete the analysis and respond 8 formally to the points raised as soon as possible. 9 "My first question is: 10 "1. Is this is an exercise that Fujitsu could 11 undertake, possibly by Gareth perhaps, who would 12 presumably have a thorough understanding of the figures 13 recorded on both the transaction logs and the figures on 14 an office cash account? 15 "2. If the answer is yes, how soon could this 16 exercise be performed and a formal response prepared? 17 Will there be a cost ..." 18 Then if we scroll up, please. We'll see there that 19 Mr Pinder forwards the email to you on 31 July, asking: 20 "Is this something you can provide?" 21 A. Yes. 22 Q. Then if we scroll up to the top of the page, your reply 23 of 31 July: 24 "I guess it is the sort of analysis that I could do 25 if required. However it is fairly time consuming and 95 1 the problem is in adjusting my priorities to find the 2 time to do it. I don't know how many transactions are 3 involved and I would need copies of the cash account 4 before and after the period to carry out the analysis. 5 I might need to get hold of the reference data that was 6 current at that time (which is probably quite 7 difficult). 8 "Given the volume of the data, I would expect this 9 to take anything between 2 days and a week (it is 10 difficult to tell in advance given that I'll probably 11 need to feel my way), and I'll probably need help from 12 Penny in carrying out extracts from the relevant message 13 stores ... 14 "As for timescales, I can't spend more than a couple 15 of hours on it this week and probably can't start in 16 earnest until the week of 14 August [then] I'm on leave 17 ... 18 "Anne Chambers might be able to do a similar 19 analysis (and she may well have some tools to help ...). 20 However, she's on leave this week." 21 So you're essentially expressing concerns about your 22 capacity. 23 A. Yes. 24 Q. By this time, had any of the guidance that we saw in the 25 letter of 18 November 2005 been given to you about 96 1 providing expert evidence? 2 A. I'm pretty sure it hadn't. I can't say definitively 3 because it's so long ago but, as I say, I think I would 4 have known if I had seen it. 5 Q. Can we look, please, at FUJ00152290. If we scroll down, 6 please, you say there on 7 August to Mr Sewell, in 7 relation to the Castleton case and transaction logs, 8 second paragraph: 9 "I spoke with Anne this morning and we've agreed 10 that once the CP ..." 11 Is that a change -- 12 A. Change proposal. 13 Q. "... is approved, that she will carry out the initial 14 analysis (based on transactions we get from Penny) then 15 we will review them together and I'll front up any 16 report we present to [the Post Office]." 17 Again, by this stage, we're now in August 2006, had 18 the guidance given in the 18 November 2005 letter been 19 provided to you? 20 A. Not that I'm aware of and I don't think that's the 21 letter that's referred to in an earlier email that's 22 mentioned in this email. 23 Q. No, that's right. 24 A. I think that's a letter to somebody different. 25 Q. Can we look, please, at FUJ00152292. 17 August 2006 97 1 now, about Marine Drive, from Anne Chambers to you. She 2 says: 3 "I've spent some time recalculating the CAP42 ..." 4 The Cash Account Period 42: 5 "... cash account for Marine Drive and addressing 6 the points in the letter from the solicitor. Hope it 7 makes sense to provide you with a starting point not 8 quite at the very beginning. 9 "Subsequent cash accounts, if needed, should be much 10 easier now I have the mappings set up and some idea of 11 what numbers need to go where. 12 "I'm now going to look at the loss made in week 42 13 and demonstrate that it was due to the difference 14 between system holdings and declared holdings. 15 "If this isn't at all what you wanted, please let me 16 know -- I don't really know what I'm doing! 17 "This hasn't had my full attention, lots of people 18 are on leave ... Also, yesterday I got my witness 19 statement which is (as I expect you found) full of 20 things I didn't say or do, including all those PowerHelp 21 calls." 22 So Ms Chambers had been asked by you to do the work 23 or it had been agreed by you that she should do the 24 work; is that right? 25 A. Yes, as I understand it, she did the initial 98 1 calculations. I think what she sent there is 2 a spreadsheet that she put together and a note of what 3 she'd actually done and I then went through and checked 4 both her methodology and the conclusions she came to, 5 and then I think, at some later stage, I wrote it up in 6 a slightly different form than the form that she'd 7 passed it to me in. 8 Q. Why were you going to, in your words, front it up, front 9 the report? 10 A. I can't remember now. That's clearly something we'd 11 agreed between us at the time but I can't remember what 12 the -- what was behind that at the time. 13 Q. Can we turn to FUJ00122279, please, and scroll down 14 please. Thank you. An email of 2 August from 15 Mr Dilley, the solicitor, to Brian Pinder and others, 16 not including you, with the subject "First draft 17 Statement of Gareth Jenkins". He says to Mr Pinder: 18 "I refer to our previous correspondence and attach 19 the first [draft] of a ... witness statement and Exhibit 20 for Gareth Jenkins. [It's] designed to: 21 "Explain what Horizon is; and 22 "Comment on Mr Castleton's allegations about the 23 Horizon system. 24 "I am preparing a second statement for Anne (to 25 follow shortly) which will deal with the call logs. 99 1 "[Can you] look at the first draft, make any changes 2 [you feel] are appropriate and also answer my questions 3 in bold italics. [He'll then revise the statement with 4 counsel." 5 If that's scrolled up, please. It's forwarded to 6 you, for your information and action. 7 A. Yeah. 8 Q. Can we look, please, at FUJ00122283. You reply the next 9 day at 3.30-ish and say: 10 "I've had a go through this. It isn't a complete 11 statement and I won't have time to sort out the 12 outstanding bits for a while. I think we probably need 13 to get the detailed analysis done before this can be 14 completed. 15 "I'm happy for you to send it as it is to Stephen if 16 you like but you may decide it isn't complete enough for 17 that. Also I may be able to research some of the areas 18 (such as auditing) that I'm claiming ignorance of if 19 that is required." 20 So you're essentially saying that "Here's 21 an incomplete statement, I haven't got a chance to 22 complete it yet" -- 23 A. Yes. 24 Q. -- "there's still some analysis work to be undertaken"? 25 A. Correct. 100 1 Q. Again, by this stage, had the letter of 18 November and 2 the expert duties points in it been drawn to your 3 attention? 4 A. No. 5 Q. Thank you. That can come down. I think you were 6 subsequently informed by Mr Pinder that you wouldn't be 7 required to give evidence in Mr Castleton's case; is 8 that right? 9 A. That is what I understand from the documents I was 10 emailed. I certainly remember I didn't give evidence. 11 Q. Do you remember Mr Pinder informing you directly that 12 you would not be required to give evidence in 13 Mr Castleton's case? 14 A. I've seen an email which effectively says that, yes. 15 Q. Before the break can we just look at that, please. 16 FUJ00154733. I think this is, if we scroll down, 17 please, and again and again, and just go back to the 18 top, please. I think, again, as is or as was your 19 practice, you have cut into an email the contents of 20 Mr Pinder's email, and replied to it? 21 A. Yes. 22 Q. So the first part of the text in the two lightly shaded 23 paragraphs is what Mr Pinder is saying; is that right? 24 A. That's my understanding of it and it makes sense reading 25 it that way. 101 1 Q. He says: 2 "Just been chasing Stephen up re your attendance and 3 any matters still outstanding for us [Post Office 4 Account], as follows; (my words) 5 "He states that although you [that's you, 6 Mr Jenkins] would probably make a good witness, it is 7 for evidential reasons that you cannot be called. To do 8 with evidence of 'opinion', 'expert' evidence and 'real 9 evidence', et cetera, et cetera (complicated legal 10 issues nothing to do with personalities)." 11 You say: 12 "Fine (I won't try and understand what this means!)" 13 A. Yes. 14 Q. Was that the limit of the explanation that you received 15 as to why you were not being called as a witness in the 16 Lee Castleton case? 17 A. As far as I can remember, yes, I'm not aware of any 18 other briefing I got given. 19 Q. Did you have an understanding of what the distinctions 20 drawn in this email between opinion and expert evidence 21 and real evidence were? 22 A. No, hence my comment about I won't try and understand 23 what this means. 24 Q. Was that the end of it, so far as you were concerned? 25 A. Yeah. As far as I was concerned, it just meant that 102 1 I didn't need to be involved in the case any more and 2 therefore I could get on with my day job. 3 Q. You, in fact, did have a continuing involvement in the 4 case because you answered some queries? 5 A. Yes. 6 Q. But you didn't give evidence? 7 A. Correct. 8 Q. The references to those, I'm not going to turn them up, 9 are FUJ00154747 and POL00069822. 10 In what capacity were you continuing to be involved 11 in the case? 12 A. I believe I was asked to do some further analysis. As 13 I say, I can't remember now what the details were but 14 I've seen emails suggesting that I was asked some 15 specific questions which I responded to, and I think 16 I may have produced another paper as a result of some of 17 that investigation, but I can't remember now what I did 18 at the time. All I can go by is the emails that the 19 Inquiry has shown me. 20 Q. That document can come down. Thank you. 21 You tell us in your witness statements that you 22 didn't see Anne Chambers' "Afterthoughts" document, 23 either at the time or until this Inquiry, essentially. 24 A. I've no memory of seeing that. 25 Q. Given that the plan was that you should present the 103 1 evidence that constituted the work she had undertaken, 2 you were going to front up the report -- 3 A. Yeah. 4 Q. -- and then she had ended up giving evidence, rather 5 than you at trial, did you not discuss with her how that 6 came about? 7 A. I think it was really as part of this thing, that it was 8 a case of it had been decided that, because Anne had 9 done the original analysis and I think she had done the 10 analysis -- I think there may have been a call with her 11 back in 2004 because I think her analysis was partly 12 based on work that she'd actually done at the time and, 13 therefore, that was why she needed to give the evidence, 14 because she'd actually done the original analysis and, 15 therefore, me giving evidence, just reporting what she'd 16 told me, didn't really work. 17 So it was something like that, but I just don't 18 remember the details, and that was probably what I took 19 as being behind what was shown in the last email you 20 showed me. 21 Q. Was that explained to you at the time, ie that Anne 22 Chambers is the one that's been selected to give 23 evidence because she could give some evidence of primary 24 facts, ie work that she had undertaken in answering 25 a Helpdesk query in 2004: that's why she's the witness 104 1 of choice? 2 A. I think I may have heard that at some point but I just 3 can't remember the details, sorry. 4 Q. Did you have any discussion with her about the wider 5 issues raised by her document? 6 A. What do you mean by -- 7 Q. She was critical of the process by which she had become 8 involved in the case -- 9 A. Are you talking about the afterthoughts document? 10 Q. Yes. 11 A. No. 12 Q. Were you aware that she was an extremely reluctant 13 witness? 14 A. Yes, I think I was aware she was reluctant to be 15 a witness. 16 Q. Why was she reluctant to be a witness? 17 A. I think it just took her outside her comfort zone. 18 I think that would apply to any person, really, going 19 before the legal process and she certainly didn't feel 20 very comfortable with it. 21 Q. Did she raise with you her complaints about the manner 22 in which she had been pulled into the case? 23 A. No, I don't believe so. 24 Q. Did she raise with you her complaints about the process 25 by which disclosure had been given? 105 1 A. No, I think she may have mentioned something about 2 having something sprung on her about NT event logs being 3 different from logs, but that's the only vague 4 recollection I have, because I think she -- didn't she 5 have to go back about a month after the original trial 6 to present some further evidence or something? 7 Q. That's right. Is that something you spoke about at the 8 time with her? 9 A. I was aware that she had to go back and that she wasn't 10 happy with it, but not the gory details of it. 11 MR BEER: Thank you. 12 Sir, that's just coming up to 12.20 now, can we 13 break now until 12.35? 14 SIR WYN WILLIAMS: Yes, of course. 15 MR BEER: Thank you, sir. 16 (12.19 pm) 17 (A short break) 18 (12.35 pm) 19 MR BEER: Good afternoon, sir, can you see and hear us? 20 SIR WYN WILLIAMS: Yes, thank you. 21 MR BEER: Thank you. 22 Mr Jenkins, in your -- I'm going to turn to the 23 capacity in which you gave evidence in witness 24 statements and then orally in the Seema Misra case -- 25 A. Yeah. 106 1 Q. -- just as a general topic, as a theme. 2 A. Okay. 3 Q. In your witness statements provided to the Inquiry, on 4 a number of occasions you explain you were not the Chief 5 Architect for Horizon, correct -- 6 A. Correct. 7 Q. -- nor, indeed, a lead engineer for the Post Office 8 Account? 9 A. No. 10 Q. You explain that you were not the only person involved 11 in the development of Horizon nor the maintenance of it 12 when it was up and running? 13 A. Correct. 14 Q. Why did the Post Office rely on you specifically as its 15 expert witness? 16 A. I don't know, is the simple answer. I did have a fairly 17 good overview knowledge of the whole of Horizon because 18 of really going back to my role in the agent team, 19 because the whole point of the agents was to sit in the 20 middle and see the difference between what was happening 21 at the counter and what was happening at the back end 22 and, obviously, with the work I did on IMPACT, I got 23 a much more detailed knowledge of how the counter 24 operated. So, from that point of view, I probably was 25 one of the people who had a good overview knowledge of 107 1 how the Horizon system worked. But I don't think I was 2 necessarily the only person. 3 Q. Was there anyone who was a better placed witness than 4 you? 5 A. Not necessarily. I think there was some email exchange 6 in early 2010, when there was -- other names did get 7 suggested to as to being able to get involved but, in 8 the end, it was me that got picked to actually get 9 involved. 10 Q. Did you ever discuss the role that you were performing 11 as a witness with anyone senior at Fujitsu? 12 A. They were clearly aware that I was being asked to do 13 these things and they accepted that it was something 14 that needed to be done and that I could probably do it 15 as well as others but I don't know that we had a formal 16 "Yes, this is something that needs to be part of your 17 job", but it sort of evolved, and then it did become, 18 effectively, part of my job, moving forward. 19 Q. Did you ever discuss with anyone senior at Fujitsu the 20 demands being placed upon you by the Post Office? 21 A. Yes, because I had to see how this scheduled in with the 22 work that I was doing, and it was felt that at the sort 23 of level that we thought was required of the work, then 24 it could be fitted in. But, clearly, now understanding 25 what maybe I should have been doing in terms of the role 108 1 of an expert, then that would have been maybe 2 a different situation. It's difficult to speculate. 3 Q. Was the focus on your capacity and the money that the 4 Post Office would be charged for the work that you did? 5 A. I don't think the money came into it because the Post 6 Office were paying for me regardless of what I did. So 7 it was really a case of they could pay for me to do 8 designs of new bits of the system or they could pay for 9 me to support prosecutions, and it was really up to 10 them. They were paying the same either way, so it 11 didn't really make any difference and certainly made no 12 difference to my salary. 13 Q. So was the focus, then, on the issues of your 14 capacity -- 15 A. Yes. 16 Q. -- rather than the nature of the function that you were 17 being asked to perform? 18 A. Yes, I think that's probably fair. 19 Q. During your involvement in the various criminal cases, 20 did you feel under any pressure placed upon you by the 21 Post Office or its lawyers publicly to refute any 22 suggestion by the defence that there were issues with 23 the reliability of Horizon? 24 A. There were certainly cases where they were trying to put 25 words into my mouth which I didn't want to say and 109 1 I think there are examples of emails where I've said, 2 "Well, I agree with what this defence statement is 3 saying and nothing really to add to it". 4 Q. What do you think at the time, at that pressure being 5 placed upon you by the Post Office? 6 A. I just took it as being the way these things happened. 7 There was clearly, in some cases, it was -- things 8 seemed to be all happening last minute and I just 9 assumed that's how things happened in that sort of 10 environment. 11 Q. Never mind the timing so much. I'm thinking more of 12 people were trying to put words in your mouth? 13 A. I wouldn't allow them to put words in my mouth unless 14 I agreed with them. 15 Q. Did you raise that with anyone senior at Fujitsu? 16 A. I think I did -- well, I certainly had conversations 17 with David Jones back in February 2010. 18 Q. He was a Post Office lawyer? 19 A. No, he was a Fujitsu lawyer. 20 Q. Sorry, I meant a Fujitsu lawyer. 21 A. Yes, he was a Fujitsu lawyer. I think that was really 22 instigated by Penny, rather than myself. She felt that 23 Post Office were getting to be very demanding of Fujitsu 24 in general, and me -- and her to some extent -- in 25 particular, because not only was there the case for 110 1 Mrs Misra, there was also a case to do with Mr Hosi 2 going on at the same sort of time, and she felt that it 3 was important that we had some sort of legal cover, for 4 want of a better word, of the way that it was being 5 approached. 6 And so she set up this meeting with David Jones 7 which lasted about an hour or so, something like that, 8 where basically what -- my memory coming out of that is, 9 "Well, just tell the truth and what happened, and get on 10 with it and we'll support you as you need it". 11 Q. Was there any discussion there as to the capacity in 12 which you were giving evidence? 13 A. Not that I can recall. 14 Q. Was there any discussion then as to the duties that you 15 may owe the court, given the nature of the evidence that 16 you were providing? 17 A. No, I don't believe so. 18 Q. I think it's right, isn't it, that the Post Office and 19 its lawyers made a series of direct approaches to you on 20 a number of occasions? 21 A. Yes, they did, and I tried to forced them back into the 22 way that they should have been communicating with me, 23 they should have been communicating with me through 24 Penny and the Prosecution Support Team, and there were 25 a number of occasions where I was approached directly 111 1 and, in most of those cases, I then forwarded the emails 2 on to Penny and tried to get some sort of control back 3 on the process. 4 Q. Can we just look at some examples of those, please, 5 Mr Jenkins. POL00097138. If we scroll down, please. 6 There's an email underneath this one, no need to look at 7 it for now. You reply on 16 November to Rachael Panter; 8 do you remember who she was? 9 A. She was a lawyer working for Cartwright King, I believe. 10 Q. You say: 11 "Rachael, 12 "Can't be you use the report I have already sent 13 you? There is no mention of the case on the report. 14 "You should really be addressing such requests 15 through Post Office rather than directly to myself. 16 "As far as I know there is no commercial cover in 17 place for me to spend any time on such activities (and 18 that includes the case of Nemesh Patel)." 19 Then, if we scroll up, please, James Davidson, 20 a Delivery Executive, says to Ms Panter, and you're 21 copied in: 22 "I am concerned at the engagement approach being 23 taken here, we are fully on board to support but all 24 approaches must come through Post Office by the correct 25 change process." 112 1 Is this is an example of you pulling the Post Office 2 up on direct approaches to you? 3 A. Indeed it was, and if you notice on the previous email, 4 although I've been sent the email directly by Rachael 5 Panter, when I responded, I copied in both Penny Thomas 6 and James Davidson, so they were aware of the fact that 7 I'd been approached. 8 Q. What was the issue in your mind with a direct approach 9 from the Post Office? 10 A. I think -- well, there's two or three issues, really. 11 The main one was that I felt that the Prosecution 12 Support Team, namely Penny, needed to be aware of what 13 was going on and that we actually had a single point of 14 contact so that Fujitsu was aware of what Post Office 15 was requiring them to do in terms of legal processes, 16 and so on. 17 Q. Why? 18 A. Just so that the right people knew what was going on. 19 And then there was separately the question of making 20 sure that Post Office was being charged correctly for 21 the use of time, and so on. Again, the charging didn't 22 really come down to anything to do with what I was being 23 paid or anything like that, but I needed to be able to 24 allocate my time so that it was being accounted for 25 correctly. 113 1 Q. So just going back to the first reason you gave there, 2 or the first issue, why was it important that the 3 Fujitsu Prosecution Support Team were sighted on these 4 exchanges? 5 A. So that they knew what -- basically, so they knew what 6 was happening. It was their responsibility. There was 7 supposed to be a line of communication between the Post 8 Office Legal Department -- or not the Legal Department, 9 their investigators -- Jane Owens is the name that 10 I think of in that role -- and the Fujitsu team, to 11 actually know what was going on to coordinate the work 12 so as to be aware of what was happening and what was 13 going on. 14 Q. What was the importance of that, though? Was it written 15 down somewhere that that's what had to occur? 16 A. I think it might have been but I wasn't aware of that 17 but it just felt that that was the right process for 18 doing things. I mean, my work had to be accounted for. 19 I had management telling me what I needed to do. Having 20 these odd requests out of the blue that were going to 21 take up time, I needed to make sure that people were 22 aware that these were happening and that my priorities 23 could be allocated correctly between the various demands 24 of my time. 25 Q. Was it in your mind that a prosecutor has duties to 114 1 retain, to record and to disclose material? 2 A. I wasn't aware of that, no. 3 Q. Was it in your mind that Fujitsu, as a third party to 4 a prosecution, might be required to disclose material to 5 a court? 6 A. No, I'd not thought of it that way at all. No. 7 Q. Did you ever come to understand that you might owe 8 personal duties of disclosure to a court? 9 A. Only when it was explained to me in 2020 what an expert 10 needed to do. 11 Q. Did you ever come to understand that Fujitsu may owe 12 duties of disclosure? 13 A. No, as far as I was -- understood it, the disclosure was 14 something that Post Office needed to worry about, not -- 15 it wasn't something that affected me or Fujitsu. 16 Q. What was your understanding of the Post Office's duties 17 of disclosure? 18 A. I didn't really understand what they were but I was 19 aware that there were -- that I was told that there had 20 been various hearings requesting various bits of 21 disclosure, which they seemed quite happy had gone to 22 say that they did not need to actually disclose the 23 things that they had been asked to disclose. 24 Q. Was that in the Seema Misra case? 25 A. Yes. I wasn't concerned about the details; I was just 115 1 aware that such discussions were going on in the 2 abstract. 3 Q. In any of those discussions, were the extent of Fujitsu 4 or your own duties of disclosure discussed? 5 A. No. 6 Q. Can we move forwards, please, to FUJ00153986 and look at 7 page 8, please, and if we scroll down, please. That's 8 excellent. An email from Jane Owen on 16 November 9 2012 -- which is the same day we were looking at -- at 10 the moment to Penny Thomas, copied to Jayne Bradbury. 11 That's not a name that's familiar to me, do you -- 12 A. I don't recognise the name at all, I'm afraid, sorry. 13 Q. Anyway, Jane Owen, Security Manager in the Post Office, 14 says: 15 "Hi Penny ... 16 "I am a little out of the loop with this now that 17 I have changed roles. I asked Mark about raising 18 an invoice (he has the budget) to cover the expert 19 witness requirements for one of Sharron's cases ... 20 which I think the email below will relate to ... back in 21 April an order was raised ... but we were never invoiced 22 ... Do you know anything about that at all? Gareth has 23 been approached directly by an external solicitor and 24 Jayne will need to start discussions with our team 25 leaders on how we prevent this from happening going 116 1 forward as it is clearly unacceptable that he be 2 'ambushed' in this way." 3 Did you feel like you were being "ambushed" by the 4 Post Office, that's the word? 5 A. I wouldn't have necessarily used that word but, clearly, 6 having this email out of the blue from Rachael Panter in 7 November when the last contact I had had about any of 8 these cases was about a month before that was a bit of 9 a surprise and, clearly, I felt that we needed to have 10 some sort of control as to how my work was to be 11 scheduled and how Post Office was managing this and 12 I see this as being Post Office trying to actually get 13 that -- some control put back in there. 14 Q. But this is all about capacity and formalisation between 15 two business partners. It's not to do with the 16 substance of the evidence that you're to give or any 17 procedural duties that attach to it? 18 A. No, or sorry, yes. That's -- yes. Yes. 19 Q. That's correct what -- 20 A. That's correct what you said, yes. 21 Q. Can we move forwards please to POL00229801. 22 Sorry, FUJ00229801. I think I said POL. Thank you. 23 If we can turn to page 3, please, and scroll down, 24 please. Thank you. This is an email exchange between 25 you and Angela van den Bogerd, can you see that, of the 117 1 end of January 2013? 2 A. Yes, I see that. I don't understand the context yet 3 but, presumably, we'll get there. 4 Q. To give you a bit of context, I'm not sure it's 5 completely necessary but if we just scroll down, please, 6 and scroll down still further, and again. Thank you. 7 This is about the Helen Rose Report. If you scroll up, 8 please, now. 9 A. When you say Helen Rose Report I'm aware of two Helen 10 Rose Reports. Are we talking about Lepton here? 11 Q. Yes, I think so. 12 A. Okay. 13 Q. If we scroll up to there. Thank you. I think Ms van 14 den Bogerd is asking you questions about it. 15 A. Yes. This looks again like an email from Angela van den 16 Bogerd to myself, where I've actually inserted with the 17 GIJ prefix my responses to her various questions. 18 Q. Yes, and if we just scroll up, please. It's the second 19 paragraph of your reply. In the first paragraph you 20 say, "I've added replies with Gareth Idris Jenkins, 21 GIJ". 22 A. Yes. 23 Q. In the second paragraph, you say: 24 "... this is all outside of our official remit. 25 I probably shouldn't be doing this investigation for you 118 1 in this way (though if you ask for formal help, it is 2 likely to be me who gets involved!)." 3 We saw the exchange of the previous November, where 4 there was an attempt at formality made. Had that 5 failed? 6 A. I'm not sure. I think what I -- it's a case of trying 7 to understand the amount of work involved. I've been 8 asked some specific questions. I probably spent about 9 half an hour or an hour replying to this email so 10 I thought it was simpler to just reply to the email and 11 then point out that this should really go through 12 official channels, and that is why I'd copied Penny in 13 on the response because I saw hers being the official 14 channel. 15 So I wasn't trying to be awkward; I thought let's 16 just answer the question because that's fairly 17 straightforward. If it was going to take me a long 18 amount of time, then I would have just said, "Let's get 19 this sorted out officially". 20 Q. Why was it outside your official remit? 21 A. Because my job was designing new changes to the system, 22 not answering questions from random people in Post 23 Office. 24 Q. Why did you answer random questions from people in Post 25 Office? 119 1 A. Because I felt I could actually answer them and I felt 2 that, if it did go through the official channels, it 3 would probably be me that would end up being asked about 4 them. I think at this stage, I may have already had 5 some sort of contact with Helen Rose about this but 6 I can't remember exactly the chronology. 7 Q. By this time, you'd already raised a red flag, so to 8 speak, in November -- 9 A. Yes. 10 Q. -- about direct contact but you were perpetuating it 11 here? 12 A. Yes. 13 Q. Do you adopt an informal "trying to help" approach? 14 A. I try to be helpful when I can, yes, but I'm aware that 15 I also need to try to put some sort of control in, which 16 is why I copied Penny on this to make her aware that 17 I was being asked these sort of questions. 18 Q. Because, of course, you don't know what's going to be 19 done with what you said in your earlier email, do you, 20 ie the use to which it's going to be put? 21 A. No. 22 Q. Can we move on. That can come down, please. You tell 23 us in your third witness statement a series of things -- 24 I'm going to try and summarise them -- about your 25 approach when giving evidence. Firstly, would this be 120 1 right, that you gave evidence without an understanding 2 that you might be an expert witness in the legal 3 sense -- 4 A. Correct. 5 Q. -- and that you were, therefore, lacking in 6 understanding that you were subject to the duties 7 imposed upon an expert witness, including duties of 8 disclosure? 9 A. Correct. 10 Q. You tell us in your witness statements that you lacked 11 that understanding in Ms Misra's case, that's 12 paragraphs 233 and 329 of your third witness statement. 13 Mr Allen's case, paragraph 619; and in Ms Sefton and 14 Ms Neald's case, paragraph 689. 15 A. Correct. 16 Q. The second thing -- and, again, I'm trying to summarise 17 a large body of text here -- is this right: you say you 18 were never provided with written instructions to be 19 an expert witness in any case in which the Post Office 20 asked you to give evidence? 21 A. Correct. 22 Q. That includes, Mr Thomas' case, paragraph 313; 23 Ms Misra's case, paragraph 329; Mr Allen's case, 24 paragraph 619; Ms Sefton and Ms Neald's case, 25 paragraphs 689; and Mr Ishaq's case, paragraphs 643 and 121 1 646? 2 A. Correct. 3 Q. Thirdly, is this right: you say you were not informed on 4 how to interact with a defence expert or to prepare 5 a joint statement with another expert? 6 A. Correct. 7 Q. That applied in both Ms Misra's case, paragraph 478, and 8 Mr Ishaq's case, paragraph 666? 9 A. Correct. 10 Q. Then, fourthly: you say that you were never asked by the 11 Post Office to provide a witness statement that 12 contained an expert witness's declaration? 13 A. Correct. 14 Q. That applied in Ms Misra's case, paragraphs 233, 466 and 15 483; Mr Allen's case, paragraph 619; Ms Sefton and 16 Ms Neald's case, paragraph 689; and Mr Ishaq's case, 17 paragraph 669? 18 A. Correct. 19 Q. You tell us in your witness statement, it's your third 20 witness statement, about circumstances in which you 21 asked for guidance in the run-up to Seema Misra's trial 22 as to what was expected of you as a witness. 23 A. Yes. 24 Q. By that time, had you suffered from a lack of confidence 25 in the management of the requests coming from the Post 122 1 Office? 2 A. Well, they were certainly fairly chaotic, a number of 3 emails I was getting, and I was certainly aware that 4 I was getting asked the same question multiple times, 5 having already answered questions, and there was even 6 some occasions when I responded saying, "Well, here's 7 the email I sent to you yesterday answering the same 8 question". So, yes, I could see that it -- that there 9 was a lack of organisation, is maybe a polite way of 10 putting it, in terms of the way things were being 11 managed. 12 Q. The email exchange that you had -- I'm not going to go 13 to it now, we'll go to it later -- when you asked for 14 guidance "What's expected of me as an expert witness", 15 and you got a rather asinine reply, didn't you? 16 A. I can't remember that particular email but that doesn't 17 surprise me at all. 18 Q. That one where "You've just got to tell the truth". 19 A. Yeah. 20 Q. That one. 21 A. That's going back to -- that was Mr Thomas' case 22 I believe, yes, yes. 23 Q. Yes. 24 A. Yeah. 25 Q. Were you concerned about that, the basic nature of the 123 1 reply you got, the lack of guidance you received? 2 A. I just took it at face value, I think, rather than 3 questioning it. If I just had to tell truth then, well, 4 I'd do that anyway so that wasn't really telling me 5 anything new. 6 Q. Why do we not see you asking any of your own managers at 7 Fujitsu for assistance in this regard, "I'm being 8 dragged into court proceedings here, please help me as 9 to what is required of me"? 10 A. I thought I was being told what was required of me by 11 Post Office, so I didn't really understand that I wasn't 12 being told the right things. So I just trusted what 13 I was being asked by Post Office and didn't see the need 14 to involve my management. They were clearly aware I was 15 spending my time doing these things but I didn't feel 16 I needed any further guidance from them, in that 17 I thought what I am looking for is guidance about legal 18 things, therefore I'll probably get it better from the 19 lawyers that I'm talking to in Post Office than managers 20 in Fujitsu who know no more about the law than I do. 21 Q. But you got no guidance from the Post Office? 22 A. Well, I realise that now. I thought I was getting some 23 sort of guidance but it probably wasn't really very good 24 guidance. 25 Q. What sort of guidance did you think you were getting 124 1 from the Post Office at the time? 2 A. Tell the truth to complete a witness statement. 3 Q. Why do we not see you asking the lawyers at Fujitsu, 4 with whom you did correspond relatively frequently for 5 greater support and guidance on the role that you were 6 to perform? 7 A. I think that came after about 2010, sort of in 8 Mrs Misra's case, and I did seek guidance from David 9 Jones and I had that meeting with him in February and, 10 for a while, then he actually acted as a buffer between 11 me and Post Office lawyers, and so on. Then later on, 12 I was involved with Jean-Pierre (sic), who did give me 13 a bit of further guidance but not the true guidance that 14 maybe I should have had in terms of what an expert's 15 duties were. 16 Q. Can we look, please, at FUJ00152866. If we scroll down, 17 please -- thank you -- an email exchange between you and 18 David Jennings. Who was Dave Jennings? 19 A. I'm not quite sure. Reading the context of that email, 20 I think he was someone within Customer Service but 21 I can't remember exactly what his role was. I don't 22 recall the name at all now, I'm afraid. This looks to 23 be to do with booking time and concerns about what time 24 I was charging to Post Office for activities in support 25 of Penny. 125 1 Q. So, again, this is one of those emails where you set out 2 his questions and then answer them? 3 A. Yes. 4 Q. You say: 5 "Dave, 6 "In response to your questions ..." 7 Then his question is: 8 "I understand you act from time to time to support 9 our prosecution ... activity -- sort of an 'expert 10 witness' I suppose. 11 "I have a few initial questions on this ..." 12 Then, stopping there, he describes you as "sort of 13 an expert witness". Did you regard yourself as a sort 14 of expert witness? 15 A. Well, I was clearly an expert in how Horizon operated 16 and that's what I assume that meant. I didn't 17 understand what it's meaning was in the legal sense, and 18 didn't until about 2020. 19 Q. He says: 20 "How frequent is this typically and how much notice 21 is there ..." 22 You say: 23 "It's ad hoc ... I tend to have various questions 24 from Penny a couple of times a month. 25 "How much effort ... 126 1 "It varies ... 2 "Where do you book the time ..." 3 You say: 4 "I have a code for prosecution support", and you 5 give the codes. 6 "Does anyone else in Requirements ... do this 7 activity?" 8 You explain and say you're not sure how they book 9 their time: 10 "Is there anyone else who should cover this activity 11 as well/instead of you?" 12 You answer: 13 "'Should', then probably Yes. 'Could', then 14 probably No!" 15 What did you mean by that? 16 A. Well, I think what I was suggesting is that maybe 17 someone should be formally doing -- performing this sort 18 of role but I was also accepting the fact that I was 19 probably in the best position to actually do that but 20 I certainly didn't want it as full-time job because 21 I was enjoying myself doing the design work that was my 22 day job, if you like, so I was seeing this as something 23 that I was fitting in on the side between that. So 24 that's really what I was trying to summarise there. 25 Q. This seems to be focused on the commercial aspects of 127 1 the arrangement, whether there's appropriate commercial 2 coverage in place -- 3 A. Yes, I think that must have been Dave Jennings' role, 4 was looking at the commercials of it because I think he 5 comes back to say that I'd spent about a week on it over 6 the past year, or something like that. 7 Q. Was there any equivalent set of exchanges over the basis 8 of which you were giving evidence or your status when 9 you were giving evidence? 10 A. I'm not quite sure what you mean by that? 11 Q. Were there any exchanges internally with Fujitsu, to 12 your knowledge, over the status of your evidence? 13 A. Do you mean, in terms of whether I was an expert or not? 14 Not that I'm aware of. As I say, what usually happened 15 is if there was going to be some sort of engagement of 16 my time for a particular case, then they would arrange 17 with Post Office to have some sort of commercial cover 18 for a number of days of my time and, if necessary, that 19 would get extended so that I had something to put the 20 time against. But exactly how those -- I left that 21 really to Penny to sort out, as long as I had the time 22 code to book to on a Friday afternoon, then that was all 23 I was really worried about. 24 Q. Can we look, lastly before the lunch break, at 25 POL00097123, and look at page 9, please, and scroll 128 1 down, and scroll down. Thank you. 2 An email from Ms Jennings: 3 "The case [that's Patel] due for trial ... Can you 4 ... please put this date in your diaries." 5 Scroll up, please. You reply: 6 "... I'm not aware of this case or what might be 7 required of me. 8 "I'm not aware of any other outstanding cases which 9 I might be involved in." 10 Scroll up. Reply from Sharron Jennings: 11 "... Gareth 12 "This is the one you supplied the expert report and 13 witness statement for the week before last. Apologies 14 for not explaining that properly in a previous email. 15 It was a blanket email for all witnesses! It is unclear 16 at this stage who will be required as witnesses and 17 which evidence will be accepted without the need for 18 attendance. I just thought if I let everyone know they 19 can pencil it in and then I can let you all know nearer 20 the time." 21 Again, there's a reference there to you supplying 22 an expert report. Did that trigger anything in you? 23 A. No, because as far as I was concerned, what an expert 24 report meant was a report by someone who understood what 25 Horizon was about. 129 1 Q. Okay. Then if we scroll up, please. You say: 2 "Thanks for the clarification ... If I am required 3 to go to court ... I need some more background on the 4 specific case and exactly what's being alleged. 5 I appreciate that is not covered by my statement, but if 6 I need to be an expert witness, I need to understand 7 what is happening." 8 In what sense were you referring to yourself as 9 an expert witness there? 10 A. Again, as an expert on how Horizon operated. 11 Q. That's not quite how it reads, is it? 12 A. I just didn't understand the concept of an expert in the 13 legal sense because I've got no legal background or 14 training. 15 MR BEER: Thank you. 16 Sir, that's an appropriate moment. Can we take 17 a break until 2.00, please. 18 SIR WYN WILLIAMS: Yes. So we will take a break until 2.00 19 please. Thank you. 20 MR BEER: Thank you very much, sir. 21 (1.08 pm) 22 (The Short Adjournment) 23 (2.00 pm) 24 MR BEER: Good afternoon, sir. Can you see and hear us? 25 SIR WYN WILLIAMS: Yes, thank you. 130 1 MR BEER: Good afternoon, Mr Jenkins. We were dealing with 2 the theme of the capacity in which you gave evidence in 3 criminal proceedings and your understanding of the 4 capacity in which you were giving evidence. 5 A. Yes. 6 Q. Can we continue that theme, please -- 7 A. Okay. 8 Q. -- by looking at FUJ00152872 and if we look at the top 9 half of the page, please, we'll see this is an email 10 from you to Penny Thomas of 19 December 2009 with 11 an attachment which in fact follows this document, and 12 you say: 13 "Penny, I attach the expert witness statement with 14 my comments." 15 What you do, essentially, is attach Charles 16 McLachlan's second report in the Seema Misra case with 17 your comments on it. In that first line of the email, 18 you say: 19 "I attach the Expert Witness statement with my 20 comments." 21 You understood that Mr McLachlan was an expert 22 witness? 23 A. Yes. 24 Q. Did you understand you were an expert witness in the 25 equivalent capacity? 131 1 A. Not in the terms I now understand the capacity to be. 2 Q. We're going to see in a moment you prepared documents in 3 answer to his and you engaged in a joint discussion with 4 him. 5 A. Yes. 6 Q. Did you understand you were performing a different 7 function than he was? 8 A. I was clearly in a -- at least, I certainly felt I was 9 in a different position because I was not independent in 10 the same way that he was, in that I was an employee of 11 Fujitsu and, therefore, was effectively part of the POL 12 prosecution team. 13 Q. Did you think you were part of the POL prosecution team? 14 A. Yeah, I think I probably did, because that's how the POL 15 lawyers were treating me. 16 Q. If we scroll forwards, please, to the report which 17 starts at page 4, you'll see his second report and then, 18 if we go over the page, please, to page 5, you'll see 19 there's an index. Just note for the moment -- we're 20 going to come back to it in a moment -- section 5 is "My 21 Duties to the Court". If we go over to page 6, please. 22 You'll see that he sets out, to start with, his 23 instructions; can you see that? 24 A. Yes, I can. 25 Q. If we keep scrolling through this, if we do it at quite 132 1 a slow pace, you'll see that he sets out his 2 qualifications? 3 A. Yes. 4 Q. Then if we scroll on, please. These are all 5 qualifications and experience. Keep scrolling, keep 6 scrolling, and keep going, and keep going, and again. 7 Then at 1.3, he sets out the confidentiality of the 8 document, and he makes some caveats at paragraph 1.4: 9 "This report should not be read as expressing any 10 opinion on factual matters which depend on disputed 11 testimony of the witnesses of fact, or legal issues. It 12 ... reflects my understanding of the position." 13 Then, over the page, he then sets out at 14 paragraph 1.5 his sources of information, and scroll on, 15 please. A big list of the documents that he's been 16 given. Then over the page, and keep going. He then 17 sets out at point 6 of his introductory paragraphs the 18 scope of his work. He describes himself as an expert 19 witness, not as a witness of fact. He then sets out 20 what he's done; looked at the documents that have been 21 provided to him; what he hasn't done, in 1.2; what he 22 hasn't had access to, in 1.3; where he attended, in 1.4 23 and 1.5. Then scroll down. Then over the page and at 24 1.7 he says: 25 "I have prepared an independent and objective report 133 1 addressed to the Court." 2 He sets out what his company's fees are dependent on 3 or not dependent on, and then he describes the structure 4 of his report. 5 Then if we can go to page 35, please. He sets out 6 his duties to the court. He understands his overriding 7 duty is to the court and that he has complied with that 8 duty and will continue to do so. Then, if you just scan 9 to yourself what he says in paragraphs 2, 3, 4, 5 and 6, 10 and scroll on. These are very similar to the letter of 11 18 November 2005, aren't they? 12 A. Yes, I agree. 13 Q. Then, over the page, please. The expert declaration 14 continues. Now, all of those parts that I've shown you, 15 you would have read? 16 A. I will probably have skimmed through them, yes. 17 Q. You will have read them? 18 A. I will have read them, yes. 19 Q. When you read them, did you think, "Hold on, I'm giving 20 evidence about the very same matters that Mr McLachlan 21 is giving evidence. I don't say any of these things in 22 my witness statements or documents." 23 A. No, because I thought I was in a different position 24 because I thought I was actually -- well, for a start 25 I was providing witness statements, not an independent 134 1 report and so I didn't think I was -- and I wasn't 2 independent, like he was. 3 Q. Never mind the independence point, what about all of 4 things that he said as to all of his qualifications, his 5 experience, his instructions, the documents that he had 6 been shown, the scope of his report, the caveats within 7 it? 8 A. Yeah. 9 Q. Did you think, "Hold on, I've done none of that"? 10 A. I probably realised I hadn't done any of that but 11 I didn't see that it was -- what I did was what I was 12 asked to do, which was provide a witness statement or, 13 at this stage, all I was doing was providing comments on 14 his report. 15 Q. When you read this, and you thought, "Well, I've given 16 witness statements in the past, in the capacity as 17 somebody with expertise about Horizon, I've not included 18 any of these things in any of the documents I've 19 produced, should I have done so"? 20 A. That didn't occur to me at the time. Maybe it should 21 have done but it didn't. 22 Q. You said a moment ago that you thought that you were 23 appearing or performing a function in a different 24 capacity. Did that actually occur to you at the time, 25 ie, you addressing your mind to the fact that you were 135 1 a witness from Fujitsu, therefore lacked independence 2 and, therefore, the nature of the documents that you 3 produced fell to be differently constructed? 4 A. I'm not sure that I actually analysed it that far. 5 I was basically doing what I'd been asked to do, which 6 was comment on the report and it then was suggested that 7 it was presented in terms of a witness statement. So 8 that's what I did. 9 Q. Can we go forwards, please, to POL00029411. This is one 10 of Mr McLachlan's reports, it's 50 pages long, with very 11 lengthy exhibits to it, again in the Seema Misra case. 12 One is dated 4 October 2010. You were provided with 13 this report, weren't you? 14 A. Yes. 15 Q. Did you read it? 16 A. I didn't necessarily read the introductory bits to it, 17 I was more concerned with the technical information that 18 was in there and commenting on that. I may have read 19 the whole lot, I can't remember now, but what I was 20 really interested in was the technical information in 21 the report rather than things like the expert 22 declaration and things like that because I didn't see 23 that was particularly relevant to what I was being asked 24 to do at the time. 25 Q. Again, do you now remember that: that you skipped over 136 1 or were not interested in the parts of the documents 2 that set out an expert's duties and the like? 3 A. I can't -- sorry, I can't remember exactly what I did at 4 the time but, as I say, what I would have been 5 interested in and what I knew I was being asked to 6 address was the technical aspects of the report. 7 Q. By this time, this is October 2010, so shortly before 8 the trial commenced, you were aware that this was 9 a criminal prosecution of Seema Misra? 10 A. Yes. 11 Q. A serious matter? 12 A. Oh, yes. 13 Q. Were you conscious of the seriousness of the matters, 14 certainly for her, by this stage? 15 A. Yes, I suppose I was. 16 Q. You would, therefore, presumably have wished to 17 approached your duties conscientiously and with care? 18 A. Oh, yes. 19 Q. Can we look, please, if we go over the page, again, the 20 index page, and then over the page, please, and again. 21 There's an introduction and, if we go over the page, and 22 again, he sets out a "Summary of findings". If we go to 23 page 20, please and read this together, section 4 of his 24 report is points of difference between him and you. He 25 says: 137 1 "I submitted my draft report to Gareth Jenkins, the 2 expert from Fujitsu instructed by the Post Office for 3 his review on 1 October 2010." 4 That's right, isn't it: he sent you a draft of his 5 report? 6 A. Yes. 7 Q. "We then had an opportunity to discuss points of fact 8 and opinion over the phone. Jenkins was able to advise 9 me of some errors of fact based on his extensive 10 understanding of the Horizon system which I have not 11 sought to dispute and I have incorporated corrections 12 arising from these errors into this final report. 13 Jenkins also indicated some areas where he held 14 a different opinion and this section seeks to set out 15 explicitly where our opinions differ side by side. I am 16 relying on his annotations to my draft report to 17 faithfully represent this position for the convenience 18 of the court. I apologise if, any of the individuals 19 items I have inadvertently misrepresented him or omitted 20 comments of importance." 21 Then 4.2: 22 "In relation to 2.2.1 Transaction Corrections, 23 Jenkins is of the view that because the subpostmaster 24 accepts the Transaction Corrections and has 25 an opportunity to raise issues for correction this is 138 1 not relevant to the case. In my view, if Transaction 2 Corrections are incorrect or omitted then it is 3 necessary for the subpostmaster to be able to have the 4 evidence and training to contest them. The fact that 5 there is a process for Transaction Corrections means 6 that data entry errors are recognised as occurring. The 7 ... system does not automatically provide the evidence 8 (in the form of vouchers) to enable the subpostmaster to 9 easily raise or contest Transaction Corrections. Nor 10 have the Post Office provided evidence that demonstrates 11 that the training of Misra equipped her to deal with 12 transaction corrections effectively." 13 Then another example, if we go over the page, 14 please: 15 "In relation to 2.2.1 ... Jenkins is of the view 16 that because the subpostmaster accepts the remittances 17 and has an opportunity to raise issues for correction 18 this is not relevant to the case. In my view, there is 19 testimony from other subpostmasters that the end-to-end 20 remittance process introduces incorrect data into 21 Horizon, and because I have had no opportunity to 22 investigate this, I am unable to include this as 23 a source of problems at West Byfleet." 24 So you engaged in a discussion with him -- 25 A. Yes. 139 1 Q. -- and he provided this document, which sets out his 2 views and yours, where they differ? 3 A. Yes. 4 Q. You would have read this? 5 A. Yes, I would have done. 6 Q. Again, did you not think that you were giving evidence 7 in the same capacity as Mr McLachlan? 8 A. Well, I didn't see it in quite the same way because, 9 again, I was there as an employee of Fujitsu and, 10 therefore, very much as part of the Horizon system, 11 whilst he was whereas a totally independent person. So 12 I didn't see it as being quite the same. 13 Q. You were being treated as the expert from Fujitsu that 14 had been instructed by the Post Office, as he sets in 15 his report? 16 A. Yes, okay. 17 Q. Can we go to page 27, please. He sets out the terms of 18 reference for this exercise. 19 A. Yes. 20 Q. He sets out his instructions in point 1, point 2 his 21 qualifications, and it follows the same format as the 22 previous report. I'm not going to take you -- 23 A. Yeah, yeah. 24 Q. You would have read this, wouldn't you? 25 A. Yes, I probably would have skimmed through that because 140 1 I didn't think that was particularly relevant to the 2 technical things I was being asked to comment on but 3 I probably would have read through it. 4 Q. Did you not understand that you were the subject of the 5 same duties as is set out over pages and pages in 6 Mr McLachlan's report? 7 A. No, I'd not understood that. 8 Q. Did you ask anyone, "Am I subject to the same duties as 9 this other man keeps writing pages and pages about in 10 his reports?" 11 A. No, I didn't ask anyone. Perhaps I should have done but 12 I didn't. 13 Q. Can we move on, please. FUJ00156248. It's an email of 14 1 October, that's the date, if you remember, that 15 Mr McLachlan said that he had sent you his report? 16 A. Yes. 17 Q. From you to Jarnail Singh, Penny Thomas and then 18 Jean-Philippe -- who I think you mentioned earlier -- 19 Jean-Philippe Prénovost; who was Mr Prénovost? 20 A. He was a lawyer within Fujitsu. 21 Q. And Tom Lillywhite, who was Mr Lillywhite? 22 A. He was Penny's boss and Head of Security at that time, 23 so effectively the roles that Brian Pinder had earlier. 24 You say: 25 "Jarnail [third paragraph], 141 1 "As discussed, this is what I've received from 2 Charles McLachlan the defence expert in the West Byfleet 3 case. 4 "I'm concerned about the tone and some of the things 5 being attributed to me. I'm in the process of 6 annotating the document with my thoughts/comments. 7 "Charles [Mr McLachlan] spoke to me earlier this 8 morning and has arranged to discuss my views with me at 9 4.00 pm this afternoon. At this point ... I'm reluctant 10 to have that conversation without some further guidance. 11 "When I get to the end of the document, I'm happy to 12 pass on my annotated copy ..." 13 Then something else. 14 So you were asking Mr Singh, is that right, for 15 guidance on Mr McLachlan's report when you first saw it 16 and you got, in fact, 14 pages into it -- 17 A. Yes. 18 Q. -- is that right? 19 A. Yes. 20 Q. It goes to the lawyer at the Post Office, Mr Singh, and 21 to Fujitsu Legal -- 22 A. Yes. 23 Q. -- Mr Prénovost; is that right? 24 A. Yes. 25 Q. Was this a plea by you for some guidance before your 142 1 planned call with Mr McLachlan later that day? 2 A. Yes, clearly, yes. 3 Q. You say you were reluctant to have the conversation 4 without guidance -- 5 A. Yes. 6 Q. -- further guidance? 7 A. Yes. 8 Q. What guidance were you looking for: guidance about what? 9 A. What I should be saying, how I should conduct the 10 conversation, and so on. And I can't remember what 11 guidance, if any, I did actually receive but I think 12 I did have the call with Charles McLachlan later on that 13 afternoon. 14 Q. Can I press you on what guidance you were looking for? 15 A. How I should approach things, so it was really a case of 16 I thought there were things in that document that we'd 17 already discussed before, and I thought we'd agreed that 18 they weren't relevant, and here they were coming back up 19 again. So it was a case of why is that happening? 20 Q. So was it about the substance of what Mr McLachlan was 21 saying in the report? 22 A. Yes. 23 Q. Rather than your role, the role that you were 24 performing, that you were seeking guidance? 25 A. No, it was the substance. 143 1 Q. How could they each help you, the two lawyers, on the 2 substance of what he was saying? 3 A. It was really a case of why was he coming back with 4 things that I thought we'd already discussed and what 5 I should do about that, and how to approach that in 6 terms of discussing things with him, as far as I can 7 remember. 8 Q. You tell us in your third witness statement at 9 paragraph 463, that Mr Singh did not reply to this 10 email. 11 A. Okay, I can't remember the detail at this point but 12 I clearly had researched it at the time I was writing 13 that witness statement. 14 Q. Can you recall whether any guidance was forthcoming from 15 anyone else, including Mr Prénovost at Fujitsu? 16 A. I can't remember, I don't believe I've seen any emails 17 that suggest that; whether there were any phone calls or 18 not I just can't remember. 19 Q. By the time you were sending this you hadn't finished 20 reading; you were a third of the way in? 21 A. Yes. I mean, the point was I'd received this probably 22 first thing in the morning, so I thought let's -- whilst 23 I'm reading it, then that gave people a chance to react 24 because I'd already established the fact that I -- that 25 Charles McLachlan wanted to talk to me at 4.00 this 144 1 afternoon, therefore let's give people warning, so they 2 have a chance to say and -- see what they can do to help 3 me in preparation for that call later on this -- that 4 day. 5 Q. Did you go back to Mr Singh or Mr Prénovost when you'd 6 finished reading the report? 7 A. I think at some point I sent an annotated copy of the 8 report out. I'm not sure exactly who to -- probably to 9 the same people. I can't remember the exact sequence of 10 the emails that -- 11 Q. That's the one we looked at earlier, the annotated 12 version of the report? 13 A. Yes, but I can't remember exactly who I sent it to and 14 what stage in the proceedings. No, I think the one we 15 looked at earlier was one from about two or three days 16 later. 17 Q. Okay. You think on this day you sent an annotated 18 version out? 19 A. I believe so but I can't be 100 per cent sure. 20 Q. Did you at any time ask anything about the duties of 21 an expert witness about which Mr McLachlan spoke 22 extensively in his report? 23 A. I doubt it. I was more concerned about the technical 24 detail rather than things like that, that I saw as being 25 administrative parts of the report. 145 1 Q. Did you chase Mr Prénovost to say, "Look I'm doing this, 2 I'm still doing this without any guidance, can I get 3 some help, please"? 4 A. I don't remember, I'm afraid. 5 Q. Or did you just plough on? 6 A. I probably just ploughed on but I don't know. 7 Q. Did you have an in-person meeting with Mr Singh, 8 Mr Longman and Mr Tatford about this? 9 A. The following week I did, yes. 10 Q. I think we've got no note of that meeting; is that 11 right, so far as you know? 12 A. I don't remember seeing any note of that meeting. 13 I certainly haven't seen one as part of the Inquiry. 14 Q. I don't think you've got a recollection of what was 15 discussed at that meeting? 16 A. No, my main recollection is coming up to London to have 17 a meeting in barristers' chambers, which is something 18 I'd never done before, and meeting -- I think that was 19 probably the first time I'd met all three of them, so -- 20 though I'd spoken to them on the phone before that. But 21 exactly what was discussed at the meeting, I can't 22 really remember. 23 Q. So is it the case that when you were being -- looking at 24 the matter generally -- sent these documents from 25 Mr McLachlan and you read the parts in them that 146 1 concerned his instructions, the documents with which 2 he'd been provided, the scope of the work that he had 3 undertaken, the caveats to the work he had undertaken, 4 the limitations of it, and quite extensive passages 5 about the duties of an expert, that didn't trigger 6 anything at all with you? 7 A. No, I don't think it did. 8 Q. Can we turn forwards, please, to FUJ00124313. This is 9 a document in a different case. 10 A. Yes. 11 Q. This is concerning the prosecution of Kim Wylie, and 12 it's a document prepared by you, we can see the date, 13 19 February 2013; can you see that? 14 A. I can see that, yes. 15 Q. So it's two and a bit years later? 16 A. Yes. 17 Q. What you do, if we just summarise what you do in this 18 report, it comments on essentially requests for 19 disclosure that were being made by the defence in 20 Ms Wylie's case. What you do is you set out passages 21 from an expert report by Michael Turner, and then you 22 comment on them? 23 A. That seems to be what this is doing, yes. 24 Q. Okay. You say in the introduction "I have been asked to 25 report on the document "Report: Disclosure Requests 147 1 Computer Evidence" provided by the defence in the case 2 of [Kim Wylie]. 3 "In order to do that, I have copied in the report 4 below in blue font and added my comments in black font." 5 Again, I don't think we've got the -- 6 A. Yes, I appreciate that. 7 Q. We'll have to work our way through it: 8 "In summary, it would appear that the Defence expert 9 is looking for as much information as he can to carry 10 out a detailed analysis. I have no problem in him doing 11 that and I am happy to assist in such an analysis (as 12 I have done in the [past] with other defence experts), 13 since the data requested is proprietary to Horizon or 14 Horizon Online and is unlikely to be understood easily 15 without some guidance. 16 "Any such analysis is likely to require a lot of 17 time and effort to analyse and therefore incur 18 considerable cost and elapsed time." 19 Then 2, "Disclosure Requests Computer Evidence", 20 "Terms of Reference". Now, this is you setting out 21 passages from Mr Turner's report. 22 A. Yes, this is just a cut and paste of Mr Turner's report. 23 Q. You'll see that he sets out his instructions and then, 24 at 1.3, he sets out his qualifications. Over the page, 25 he annexes a CV in Appendix A -- I'm not going to look 148 1 at that -- he sets out a statement of independence and 2 then underneath "Statement of Independence", you've 3 written: 4 "I note that the CV refers to another Horizon case, 5 R v Julia Richards, Winchester Crown Court. I have no 6 knowledge of that case." 7 That's you, adding that in? 8 A. Yes. 9 Q. Then back to him: 10 "I have made a declaration at the end of this is 11 report." 12 Then he continues. I'm not going to go through all 13 of this. The part that's indented is him and the part 14 that's not indented is you; is that right? 15 A. That's my belief, yes. 16 Q. Can we go on to page 6, please, and scroll down, 17 "Expert's declaration", you'll see that Mr Turner had 18 set out the expert's declaration. 19 A. Yes. 20 Q. Yes: 21 "I understand that my overriding duty is to the 22 court ... 23 "I have set out in my report what I understand from 24 those instructing me to be the questions in respect of 25 which my opinion as an expert is required. 149 1 "I have done my best ... to be accurate and complete 2 ... All of the matters on which I have expressed an 3 opinion lie within my field of my expertise. 4 "I have drawn to the attention of the court all 5 matters, of which I'm aware, which might adversely 6 affect my opinion. 7 "Wherever I have no personal knowledge, I have 8 indicated the source of factual information. 9 "I have not included anything in this report which 10 has been suggested to me by anyone, including the 11 lawyers ... without forming my own independent view ... 12 "Where ... there is a range of reasonable opinion, 13 I have [set out] the extent of that range ... 14 "At the time of signing I [believe] it to be 15 accurate. I will notify those instructing me if ... 16 I subsequently consider that the report requires any 17 correction or qualification. 18 "... this report will be evidence that I give under 19 oath ... 20 "I confirm that ... the facts in the report are 21 within my own knowledge", et cetera. 22 Then you've added, "Standard stuff. No comment 23 required", haven't you? 24 A. Yes. 25 Q. In this context, "standard" means usual or normal, 150 1 doesn't it? 2 A. It means I'd seen something similar before, for example 3 in the case of Mr McLachlan but I hadn't really taken 4 much notice of it and I didn't really think any of that 5 applied to me, wrongly. 6 Q. Is that what that means, you're saying? 7 A. That's -- 8 Q. -- it's rather more than those five words set out? 9 A. That was what I think I felt at the time, that I'd seen 10 that sort of stuff in other reports but I didn't see 11 that any comment was required and I didn't realise that 12 it applied to me because I wasn't making such 13 declarations. 14 Q. That's not what you say here, is it? 15 A. No. 16 Q. You say it's standard stuff and "standard stuff" here, 17 would you agree, in context, means it's normal or usual? 18 A. What I meant by that was that I'd seen something similar 19 to that before, for example in Professor McLachlan's 20 reports. 21 Q. So on your account, it indicates that it was something 22 with which you were familiar? 23 A. I had seen it before. 24 Q. Therefore it was standard? 25 A. For independent experts' reports. 151 1 Q. So going back to paragraph 1 of this list, you 2 understood that his overriding duty was to the court, 3 yes? 4 A. Yes. 5 Q. That's standard, usual stuff? 6 A. Yes. 7 Q. Did you understand that your overriding duty was to the 8 court? 9 A. Well, I took "overriding duty to the court" being just 10 to tell the truth. 11 Q. So the answer is, yes, you did understand that you had 12 an overriding duty to the court? 13 A. Yes. 14 Q. Paragraph 2, you were familiar with that. That was 15 standard and usual, that the experts should set out in 16 their report the questions in respect of which the 17 expert's opinion was sought? 18 A. I'm not clear that I would necessarily have considered 19 all those points in detail when I wrote "standard stuff" 20 at the end. I was just commenting in general on 21 section 4 is an expert declaration and I'd seen expert 22 declarations before. So I'm not -- I doubt if I would 23 have gone through looking at each of those paragraphs in 24 detail and considering what they said. It was a case 25 of, yes, I'd seen an expert declaration at the end of 152 1 reports before and I saw that as a standard thing 2 without going into detail as to exactly what it said. 3 Q. In relation to paragraph 3, which you regarded as 4 standard or usual, did you understand that your duty in 5 preparing documents was to be accurate and complete? 6 A. Well, yes, certainly it needed to be accurate and 7 complete but to say whether I took detailed note of 8 exactly what it was saying in there at the time, I think 9 it's unlikely. I think I just saw that as an expert 10 declaration and that comment was referring to that as 11 a section rather than to the individual words within it. 12 Q. When you were preparing statements or reports that were 13 annexed to statements, did you understand that you were 14 under a duty to mention all matters which you regard 15 were relevant to the opinions which you expressed? 16 A. I don't know that I did understand that because I'm not 17 sure that I necessarily did. 18 Q. When you provided statements to the court or reports 19 that were annexed to statements to the court, did you 20 understand you were under a duty to draw to the 21 attention of the court all matters which might adversely 22 affect your opinion? 23 A. No, I didn't understand that at the time. 24 Q. You thought you were entitled to keep to yourself 25 matters that might adversely affect your opinion? 153 1 A. I was -- well, I was asked to comment on documents and 2 reports, and that was all I thought I needed to do, and 3 that's what I tried to do. 4 Q. Answer the narrow question that was asked of you? 5 A. Yes -- 6 Q. You -- 7 A. -- and no one had told me that I needed to do more than 8 that. 9 Q. Did you understand yourself to be under a duty to 10 indicate in your documents submitted to courts where you 11 have no personal knowledge of a matter mentioned? 12 A. No, I wasn't aware that I had any need to do that. 13 Q. You weren't aware that you had to indicate the source of 14 your information if it wasn't direct knowledge? 15 A. No. 16 Q. Did you think, looking at number 6, that you were 17 permitted to include matters in your documents submitted 18 to the court -- if we go over the page -- which had been 19 suggested to you by anyone else, including the lawyers? 20 A. No, because, certainly as we've seen in some of the 21 cases, then there were comments made on my reports by 22 the lawyers, and I actually changed things there as 23 a result of some of those comments. So, therefore, 24 I didn't realise that there was anything wrong with 25 doing that. Obviously, if I disagreed with the 154 1 comments, then I would not change things but if I was 2 happy to agree with the comments and that it didn't 3 detract from what I was trying to say, then I was happy 4 to accept suggestions as to how my wording could be 5 improved. 6 Q. We're going to come tomorrow to look, in some detail, at 7 the occasions when you changed your evidence as a result 8 of the suggestions made to you by others. 9 A. Yes. 10 Q. In particular, lawyers. 11 A. Yes. 12 Q. You didn't understand that you were under a duty not to 13 include in your documents suggestions made to you by 14 others? 15 A. No, I didn't understand that at the time. 16 Q. Did you understand that you were under a duty to set out 17 a range of expert opinion in relation to a matter where 18 there was a range of expert opinion? 19 A. No, I'm not sure that necessarily applied but, no, 20 I certainly wasn't clear -- I wasn't aware of that duty. 21 Q. Why do you say you're not sure that that necessarily 22 applied? 23 A. Well, you're talking about a range of opinion then I'm 24 not sure that there were cases where there was a range 25 of opinion, if -- 155 1 Q. On all the matters that you gave evidence, the answers 2 were binary? 3 A. Well, okay, I certainly wasn't aware that I needed to 4 set out anything to do with a range. 5 Q. In relation to the other matters mentioned in 6 paragraphs 8, 9 and 10, are your answers the same: that 7 you wouldn't be familiar with those duties -- 8 A. No. 9 Q. -- as applying to you? 10 A. No, all I thought I had to do was answer the questions 11 that I was being asked to answer and that they had -- 12 obviously the answers had to be truthful. 13 Q. None of these obligations can have been new to you, 14 because you wouldn't have described them as "standard 15 stuff", would you? 16 A. They weren't new to me but I don't think I'd taken in 17 the detail of them at all. I just -- as I said earlier, 18 the "standard stuff" was probably more to do with the 19 title of the section "Expert declaration", rather than 20 the details of the ten individual points within that. 21 Q. You weren't saying the title is standard stuff; you're 22 saying the content is standard, aren't you? 23 A. The content is -- yes. But I hadn't really considered 24 in detail what the content actually was saying. 25 Q. Why not? 156 1 A. Because I was concerned with the technicalities, so it 2 was the other bit -- I'm a technician, so I was more 3 concerned about the facts of how the system was working, 4 rather than the other -- the legal niceties, if you 5 like. 6 Q. Well, both Messrs McLachlan and Turner were technical 7 experts too, weren't they? 8 A. Yes. 9 Q. Did you think, "Well, hold on, I keep reading these 10 reports from other guys, and they keep saying this stuff 11 at the beginning and the end of their reports about the 12 duties that they're under. Does any of that apply to 13 me"? 14 A. Well, I wasn't producing reports; I was producing 15 a witness statement. 16 Q. Do you think that's the answer to the question: because 17 the document on which you were typing or was being typed 18 for you, is headed up "Witness statement", you needn't 19 have abided by any of these requirements? 20 A. It didn't occur to me that this would apply to me. 21 I clearly understand now that that's wrong but, I mean, 22 that was where I was at the time. 23 Q. That can come down. Thank you. 24 You tell us in your third witness statement -- no 25 need to turn it up, it's paragraph 331 at page 110: 157 1 "I knew that, like any other witness, I was required 2 to be truthful, although it would never have occurred to 3 me to be anything else. The approach that I took was 4 the one that I'd asked -- was asked to take by Post 5 Office lawyers, David Jones, a lawyer at Fujitsu, also 6 read at least some of my witness statements." 7 You deal at length in that witness statement with 8 limited instructions you say you were given and your 9 limited understanding of an expert's duties. You were, 10 however, aware that you were being asked to produce 11 technical evidence on which a criminal prosecution was 12 based, weren't you? 13 A. Yes. 14 Q. You would have been aware that the outcome for the 15 subpostmaster in question, the outcome that the Post 16 Office was seeking, was the criminal conviction of 17 a subpostmaster? 18 A. Yes. 19 Q. You'd be aware of all of the consequences that such 20 a conviction would entail, including a possible term of 21 imprisonment? 22 A. Yes. 23 Q. You signed a statement of truth on each of your witness 24 statements, didn't you, a declaration? 25 A. Yes. 158 1 Q. I think you acknowledge, or we'll find tomorrow that you 2 acknowledge, that the witness statements that you signed 3 presented a partial picture of what you understood about 4 Horizon's issues and faults because you were answering 5 only the narrow questions that you had been asked? 6 A. That's correct, yes. 7 Q. How did you feel able to sign witness statements with 8 a declaration on them which presented only a partial 9 picture of what you understood about Horizon's issues 10 and faults? 11 A. I don't think I was being asked about Horizon's issues 12 and faults. What I was being asked to do was address 13 the questions in the reports that I was addressing. 14 Q. You would know that witnesses, when they come to give 15 evidence, are asked to tell the truth and the whole 16 truth? 17 A. Yes. 18 Q. Did you think you were only required to tell the truth 19 in your witness statements? 20 A. I was talking about the -- those aspects that related to 21 the Horizon system and I believe I did tell the truth 22 and the whole truth, as far as the Horizon system was 23 operating in the specific branches at the specific times 24 that I'd looked at data. 25 Q. But you didn't feel under any compunction to volunteer 159 1 information about other faults or system defects about 2 which you had not been asked? 3 A. I didn't think that they were relevant in those 4 particular cases. 5 Q. Just expand on that. You're saying that you took 6 a conscious decision not to mention them because of your 7 own assessment of relevance? 8 A. Not a conscious decision. As far as I was concerned, 9 the system was behaving correctly in the branch at the 10 time and I'd seen nothing to show that it wasn't and, 11 therefore, other issues that I may have been aware of 12 were ones that had been -- that had gone on in the past, 13 or in the future, and had been fixed and did not apply 14 to the branch at the time that I was considering. 15 Q. We'll come to the detail tomorrow but, in the light of 16 the answers you've just given, are you saying that, in 17 every case in which you provided a witness statement, 18 you had undertaken or caused to be undertaken a detailed 19 examination of the data relating to that branch? 20 A. I had undertaken an analysis of the data, not 21 necessarily a completely thorough analysis of it, but 22 what I felt at the time it was sufficient to show that 23 things were working okay at that time. 24 Q. What about the generic witness statements that you 25 supplied later? 160 1 A. Those, I think, were done at a much higher level. They 2 were giving a -- they were effectively saying that 3 Horizon was behaving as it should and that information 4 could be found from the audit trail to show what had 5 actually happened in the branch to look at any more 6 detailed question. 7 Q. Can we look, please, at your third witness statement at 8 page 139. Again, we're going to deal with Ms Misra's 9 case in detail tomorrow and on Thursday but can we look 10 at paragraph 404, please. You're dealing here with 11 Ms Misra's case, and you say: 12 "I have already explained that I had referred to the 13 event timeout/locking issues causing missing 14 transactions in Legacy Horizon and that Fujitsu had 15 drawn these to Mr Singh's attention. I don't believe 16 that [the Post Office] asked me whether it should 17 disclose these (or any other Horizon problems) to 18 Mrs Misra. I haven't seen any emails from around this 19 time which suggest that. I hadn't been involved in 20 responding to defence requests for disclosure in 2009, 21 nor the abuse of process application made by the defence 22 in February 2010. I don't think my email of 1 March 23 2010 is about disclosure at all." 24 So you say here you can't recall the Post Office 25 ever raising whether they, the Post Office, or you 161 1 should disclose the issue of locking problems to 2 Ms Misra? 3 A. Correct. 4 Q. By this time, you knew that, if you were to sign 5 a witness statement in Ms Misra's case, you would have 6 to sign a statement of truth on it? 7 A. Yes. 8 Q. You knew that those statements were being used as 9 evidence in court to support a criminal prosecution? 10 A. I had sought guidance, I had said that, until we had 11 actually looked at the data, that I was not able to say 12 that there weren't any problems in Horizon and, when 13 we -- when I came to get -- see the data, which was in 14 early March, then there was a check done of the NT 15 events, which was the way that this event timeout 16 problem manifested itself and, therefore I was able to 17 include that as having been a problem in that branch. 18 Q. By this time, you had got -- this is 2010 -- wider 19 knowledge of bugs, errors and defects in the Horizon 20 system than, in the event, was reflected in your witness 21 statements in the Seema Misra prosecution? 22 A. I'm not quite sure what you mean by that. 23 Q. So, by March onwards, March 2010 onwards, when you made 24 witness statements in Seema Misra's case, you didn't 25 disclose in those witness statements all bugs, errors 162 1 and defects about Horizon of which you knew? 2 A. I didn't think I needed to do that because what 3 I thought I needed to worry about were ones that had 4 occurred at that particular branch at the time in 5 question. 6 Q. Did you consider reflecting in your witness statements 7 or making qualifications in your witness statements, to 8 set out exactly what you had been asked to do and not 9 asked to do, and make clear that you had not been asked 10 about other issues? 11 A. That didn't occur to me, I'm afraid. 12 Q. Looking back now, do you think that a reader of your 13 witness statements could reasonably gain the view that 14 you were setting out all that you knew about bugs, 15 errors and defects in Horizon? 16 A. I don't think I was ever asked to consider bugs, errors 17 and defects. What I was asked to do was comment on the 18 reports produced by Professor McLachlan. 19 Q. We've seen you on a number of occasions provide 20 information through an informal channel, through emails, 21 to Post Office? 22 A. Yes. 23 Q. Did you have any understanding or expectation of whether 24 that information provided informally would be disclosed 25 to the defence in the criminal prosecutions you were 163 1 involved in? 2 A. I didn't know what -- I didn't understand how the 3 process worked in that respect. I mean, clearly, if 4 someone had come back to me and said, "Well, in that 5 case you need to put something about it in your witness 6 statement", I would have done that, but no one did. 7 Q. Did you think to include in your witness statements the 8 fact of your informal communications with both Post 9 Office and Fujitsu about the substance of the evidence 10 that you were to give? 11 A. No, that never occurred to me. 12 Q. Or the detail of those communications? 13 A. No, I didn't think that was relevant. 14 Q. What did you think was going to happen to the 15 information in those communications? 16 A. I didn't know what was going to happen to them. As 17 I say, if someone had told me that I needed to do 18 something about it, I would have done something about 19 it, but no one came back and said I needed to do 20 something formal about that -- that sort of exchange. 21 Q. That document can come down. Thank you. 22 By the end of the Seema Misra trial, would you agree 23 that you have, by then, had a number of opportunities to 24 see whether the Post Office was adopting an approach of 25 seeking to tweak your evidence for its own interests? 164 1 A. I didn't think that at the time. Having looked back at 2 things now, then I can understand that that may have 3 been happening but, at the time, I thought everything 4 that was happening was just a legitimate tidying up of 5 the statements to make them more readable. 6 Q. Are you really saying, bearing in mind we're going to 7 look at these tomorrow, that you thought that what 8 happened in both the Hughie Thomas case and the Seema 9 Misra case was a tidying up exercise to make your 10 statements more readable? 11 A. And clarifying that, I don't think any of the changes 12 that were made actually changed the technical background 13 to what was happening. Yes, it changed the wording and 14 the way things were explained to some extent but 15 I think, in terms of they were still setting out the 16 facts as I understood them at the time. 17 Q. I'm thinking more about the things that the Post Office 18 was trying to get you to include and, by the time we get 19 to October 2010, had you not formed the view that the 20 Post Office had tried to alter the substance of your 21 evidence, both in the Hughie Thomas case and in the 22 Seema Misra case and, in some cases, you'd resisted and 23 sometimes you'd not? 24 A. I'd not seen it that way. I mean, looking back it now, 25 I can understand exactly what you're saying. But, at 165 1 the time, I didn't see it as being anything strange or 2 unusual. 3 Q. Can we look, please, at your witness statement, your 4 third witness statement, at paragraph 466, please, which 5 is on page 162. Scroll down, please. Thank you. 6 466, thank you. We're going to come back, again, to 7 the detail of this later, but you say, on that time and 8 date, you emailed a draft statement to Mr Tatford, 9 Mr Singh and Mr Longman. This is in Seema Misra's case. 10 A. Yes. 11 Q. "I used the same template as I had for all my previous 12 statements. I specifically asked for their feedback as 13 to whether it was correct in terms of approach and 14 style. No one suggested that I needed to add any sort 15 of expert declaration to it. Mr Tatford made comments 16 on this draft which are visible at [and then you give 17 a reference]. He appeared to want me to make some 18 points more strongly in favour of the Post Office than 19 I had done. In particular, he wanted me to say it 20 looked as though Mrs Misra had stolen the money rather 21 than it was incompetence." 22 Then if we go forward to paragraph 468, which is 23 over the page. You say: 24 "I think overall my responses to Mr Tatford's 25 comments demonstrate that I was not comfortable with 166 1 some of the points he tried to press." 2 So, by this time, you had had the experience of 3 Graham Ward in the Hughie Thomas case seeking to press 4 amendments to your witness statement, hadn't you, back 5 in 2005? 6 A. Yeah, I mean, I was going to say that's about four or 7 five years earlier, so I probably wasn't really 8 conscious of that at this time. 9 Q. Had you not remembered -- I mean, you didn't make 10 witness statements in many cases, I think it's 15 11 overall? 12 A. Yeah. 13 Q. Had you not remembered Mr Ward being dissatisfied with 14 your use of language in a witness statement and making 15 a marked-up version for you suggesting parts being 16 deleted? 17 A. I probably didn't remember that in 2010, no. 18 Q. But, in any event, here you say that Mr Tatford wanted 19 you essentially to harden up the case against Mrs Misra? 20 A. I'm saying that now, looking back at the email 21 exchanges. I wouldn't necessarily have put it in that 22 terms back in 2010. 23 Q. Why not? A barrister saying, "Please include in your 24 witness statement", if we just go back to what you said 25 at the foot of the page, 466: 167 1 "In particular he wanted me to say it looked as 2 though Mrs Misra had stolen the money rather than it was 3 incompetence." 4 A. Yeah. 5 Q. It doesn't take hindsight to realise that that is 6 a barrister asking you to harden up the case -- 7 A. Yes, and I refused to do so -- 8 Q. I'm asking what effect this had upon you. 9 A. I don't know, I just assumed that this was normal 10 practice. I had no experience to compare it with. 11 Q. Did it not make you feel uncomfortable? 12 A. Yes, it probably did make me feel uncomfortable, and 13 I resisted it. 14 Q. Why did it make you feel uncomfortable? 15 A. Because he was trying to change what I was saying to say 16 what he wanted to say, rather than what I wanted to say. 17 Q. Did this not cause you to think that there is a need for 18 caution here in my dealings with the Post Office, in 19 particular when giving evidence on behalf of the Post 20 Office, that their barrister is trying to get me to say 21 something that I can't and don't want to say? 22 A. No, I think it just made me make sure that I did stick 23 with my guns when I disagreed with what I was being 24 asked to do. 25 Q. Again, did you not consider whether you ought expressly 168 1 to qualify your statements to spell out what you were 2 saying about Horizon not being your complete knowledge 3 about Horizon integrity? 4 A. That didn't occur to me at the time. 5 Q. As I've said, you later went on to produce so-called 6 generic statements -- 7 A. Yes. 8 Q. -- in a series of cases, essentially without access to 9 the complete underlying data? 10 A. Yes, because I saw those as being high-level statements. 11 I was expecting to get into more detail at some later 12 stage. As I say, when I first produced that, I had no 13 knowledge as to anything that was involved, and 14 I assumed that that was normal practice. Again, I now 15 understand that it wasn't and was probably totally 16 inappropriate but, at the time, I didn't realise that. 17 Q. Did your experience in the Hughie Thomas case and the 18 Seema Misra case, and the approach of the Post Office in 19 particular in the Seema Misra case, not suggest to you 20 that you ought to proceed with extreme caution in 21 producing such generic statements? 22 A. I didn't realise at the time that anything was actually 23 improper or going wrong with what was happening. I did 24 get -- certainly in some of the cases -- with the 25 generic statement, I did actually seek guidance from 169 1 Fujitsu lawyers as well as Post Office lawyers as to 2 whether this was the right thing to do and whether 3 I should carry on making those sort of statements, and 4 they seemed to be supporting what I was doing. 5 Q. Or was it the seeming success of the approach in the 6 Misra case created a false sense of confidence for you 7 and for the Post Office; it resulted in a conviction, 8 after all? 9 A. I don't know that I considered it that way. Again, the 10 generic statement was about two years later, so I was 11 probably looking at that in isolation rather than 12 looking back as to what had happened before. 13 Q. Can we just look back, please, to your attitude to 14 witness evidence in a different case by looking at 15 FUJ00083741. If we just scroll down, please. Can you 16 see that the email from Mr Dunks to you and Penny Thomas 17 of 28 October 2010 concerns Porters Avenue, which 18 I think is Mr Hosi's branch, yes? 19 A. Yes, I believe so. 20 Q. "It looks as if everything has been sent at one time or 21 other. I can send all if wanted", et cetera. 22 Then I just want to see your reply. If we scroll 23 up, you say: 24 "Andy, 25 "Can you let me have them all. I've got to do 170 1 another 'expert report' and this time I want to actually 2 read your logs properly!" 3 You're writing here to Mr Dunks and Penny Thomas, 4 yes? 5 A. Yes. 6 Q. Well-known colleagues? 7 A. Yes. 8 Q. You're purposely using the term "expert report" by 9 putting it in inverted commas, aren't you? 10 A. Yes. 11 Q. Was that what had been requested of you? 12 A. I can't remember exactly what had been requested of me 13 at that time. 14 Q. But you're putting it in inverted to commas, like it's 15 a term of art, yes? 16 A. I can't remember -- I tend to be fairly liberal with 17 inverted commas and capital letters, and things like 18 that, in my emails. 19 Q. Isn't this is an indication that you realised that what 20 you were being asked to do was not simply give evidence 21 because you had expertise in an area, namely the Horizon 22 system, but you were being asked to provide an expert 23 report just like the ones we've read? 24 A. No, I don't think so. I think it was more to do with 25 the fact that it was a report as an expert on Horizon. 171 1 Someone who actually understood the Horizon system well. 2 Q. Your evidence is that, although you say here that you've 3 got to do another "expert report", in inverted commas, 4 you didn't understand what that meant, ie the provision 5 of an expert report, or that it came with specific 6 duties? 7 A. That's correct. 8 Q. By this time, you'd produced statements for use in 9 Hughie Thomas' case and earlier in the year in Seema 10 Misra's case? 11 A. Correct. 12 Q. You say here: 13 "... this time I want to actually read your logs 14 properly!" 15 Have you not considered the logs properly in those 16 other cases? 17 A. I don't know that I -- I certainly had not read in -- 18 been shown any logs to do with Mr Thomas' case. I was 19 aware in Mrs Misra's case that there were some logs, 20 which I thought had been handled by Andy Dunks and, when 21 I listened to his evidence in court, I realised that 22 there was some things that he was saying that maybe 23 I should have looked into and I actually did some 24 investigation that evening as to what I had heard about 25 the logs and I realised that, actually, maybe in future, 172 1 I should have actually taken more notice of what the 2 Helpdesk calls were involved, hence this comment here. 3 Q. It was a flippant comment, wasn't it? 4 A. I don't know that it was a flippant comment; it was more 5 a case of saying I'd learnt from the fact that I'd been 6 taken by surprise by some of the things that Andy was 7 saying in Mrs Misra's case as to what he'd actually 8 found in the Helpdesk calls that I'd not been aware of 9 and realised that maybe I should have taken more notice 10 of those and, therefore, moving forward, I wanted to 11 learn from that. 12 MR BEER: Thank you. 13 Sir, can we break until 3.15, the afternoon break. 14 Thank you very much, sir. 15 (3.01 pm) 16 (A short break) 17 (3.15 pm) 18 MR BEER: Good afternoon, sir, can you continue to see and 19 hear us? 20 SIR WYN WILLIAMS: Yes, thank you. 21 MR BEER: Thank you. 22 Mr Jenkins, you've told us today that you were never 23 provided with written instructions as to the duties and 24 responsibilities of an expert witness in the cases in 25 which you were instructed -- 173 1 A. Correct. 2 Q. -- and that you've no recollection of anyone explaining 3 to you that you were subject to certain duties as 4 an expert or being given oral instructions as to what 5 those duties were? 6 A. Correct. 7 Q. Have you read the witness statement of Warwick Tatford, 8 the prosecution barrister in the Seema Misra case? 9 A. I have. 10 Q. Did you listen to or watch his evidence or read the 11 transcript -- 12 A. I was actually here for that particular event. 13 Q. -- for that day. Can we look at his witness statement 14 to start with, please, WITN09610100, and page 5, please. 15 Second line down: 16 "At the time of the Misra case I ensured that the 17 Callendar Square bug was disclosed and made it very 18 clear to those instructing me that enquiries should be 19 made with Fujitsu about any other problems and any other 20 bugs should be disclosed. I made this clear also to 21 Gareth Jenkins, the expert instructed by the Crown, and 22 I made it very clear to Mr Jenkins that he was under 23 a duty to provide frank disclosure of Horizon problems 24 to the defence expert instructed in that case." 25 A. I don't believe we had that sort of conversation. Also, 174 1 I believe that Mr Tatford sent that information to one 2 of the Post Office Investigators, who then asked me 3 a slightly different question, rather than the full 4 question that Mr Tatford had intended for me to be 5 asked. 6 Q. That's certainly the case because we can see that on the 7 documents. We'll explore those tomorrow as we've 8 explored with other witnesses. 9 A. Yes, but I've no recollection of being asked to talk 10 about other bugs or that I would have considered 11 irrelevant anyway, by Mr Tatford or anyone else. 12 Q. I want to get your evidence clear on this: he says that 13 he made this clear, ie that any other problems or any 14 other bugs should be disclosed to you, and that he made 15 it very clear to you that you were under a duty to 16 provide frank disclosure of Horizon problems to the 17 defence expert. Are you saying that he is incorrect? 18 A. I'm saying that I had not understood that from him. 19 Q. Did you have a conversation with him about the extent to 20 which you needed to provide disclosure of Horizon 21 problems to the defence expert? 22 A. I've no recollection of such a discussion. 23 Q. By that, do you mean that it did not occur or that it 24 may have occurred but, because of the passage of time, 25 you no longer remember it? 175 1 A. I don't believe it would have occurred because, if it 2 had, I would have done something about it. 3 Q. You appreciate there's a bit of reverse engineering 4 going on there -- 5 A. Yes, I understand that, yes, but, as I say, I have no 6 recollection of it and, if I'd been told I needed to 7 disclose other problems in Horizon, then I would have 8 made efforts to find out about other problems in 9 Horizon, even though I thought that they would have been 10 irrelevant. 11 Q. Hold on. A moment ago you said, "I wouldn't have told 12 the defence expert about problems in Horizon" because 13 you wouldn't have believed that they were relevant to 14 Seema Misra's case? 15 A. Well, that's right. So, therefore, if I'd been told 16 explicitly I should have done, then I would have done 17 something about it, but that's why I'm saying that would 18 not -- I don't believe that I was told that I needed to 19 do that. 20 Q. So his evidence here is in error, is it? 21 A. I can't say categorically that -- I -- that is not my 22 understanding of what occurred at the time and I think 23 I would have behaved differently if I had been briefed 24 in the way that he suggests that he briefed me. 25 Q. What do you remember about a conversation with 176 1 Mr Tatford about the extent of your disclosure duties to 2 the defence expert? 3 A. I don't remember having any conversation along those 4 lines. 5 Q. Can we turn to INQ00001094, and turn to page 17, please. 6 If we look at internal page 68, which is bottom right, 7 line three onwards. This is Mr Tatford giving evidence: 8 "You say in your statement -- I'm not going to turn 9 it up -- you took great pains in all your conversations 10 with Mr Jenkins to make sure he understood the duties of 11 an expert witness? 12 "Answer: Yes. 13 "Question: You explained it was his overriding to 14 assist the [that should be 'court'] -- 15 "Answer: Yes. 16 "Question: -- to give an opinion that was objective 17 and unbiased, and that that duty overrode any obligation 18 that he might feel to the party calling him: the Post 19 Office. You explained that it was his duty to disclose 20 anything that might undermine his position and that he 21 should be entirely open with both the Post Office, as 22 prosecutor, and Professor McLachlan, about any Horizon 23 problems? 24 "Answer: Oh, yes, because the -- I had asked 25 previously in my advice for Fujitsu to be contacted and 177 1 to inform us of any problems and I saw Mr Jenkins as 2 an obvious route to doing that. That's how I saw 3 things. And it seemed to me particularly from the 4 feedback I was getting from the defence, that this 5 approach was working. 6 "Question: ... you've referred to 'the defence' ... 7 "Are you saying that [was a conversation] with 8 a colleague in chambers -- 9 "Answer: No ... 10 "It was perfectly clear to me that he found it 11 helpful to work with Mr Jenkins. It received to fit his 12 way of doing things because his way of approaching 13 things was to suggest hypotheses which needed somebody 14 to help him with. They needed to sit down together and 15 it's absolutely clear that they did that, from the 16 evidence they gave at the trial. 17 "Question: Before we get into the detail of the 18 communications between you, the Post Office and Fujitsu 19 and Mr Jenkins, if you were mindful of these expert 20 duties and the need to make them crystal clear in 21 somebody who did not enjoy functional independence from 22 the party that was calling them, and you explained them 23 to Mr Jenkins, how is it that every witness statement 24 which the Post Office sought to rely on from Mr Jenkins 25 omitted any of the necessary inclusions for an expert 178 1 report? 2 "Answer: Because most of -- until the last 3 statement, all of those were responses to Professor 4 McLachlan. It is important to bear in mind that there 5 had been an abuse of process", et cetera. 6 Did Mr Tatford make clear to you the duties that he 7 spoke about? 8 A. I don't believe he did. 9 Q. If we just go back to them, the previous page, bottom 10 right a duty "to give an opinion that was objective and 11 unbiased"; did he tell you that he have a duty to do 12 that? 13 A. I don't believe so. 14 Q. Did he tell you that that duty overrode any obligation 15 that you might feel to the party calling you, the Post 16 Office? 17 A. I don't believe so. 18 Q. Did he explain to you that it was your duty to disclose 19 anything that might undermine your position? 20 A. No, I don't believe so. 21 Q. Did he explain to you that you should be entirely open 22 with the Post Office and the defence expert about any 23 Horizon problems? 24 A. I don't believe so. 25 Q. That can come down. Thank you. 179 1 Had you been made aware of your role, your formal 2 role, as a prosecution expert witness in the legal 3 sense, in respect of each of the Post Office's criminal 4 prosecutions and the legal duties that that role 5 entailed, how would you have sought to have done things 6 differently? 7 A. Well, I think it would have required an awful lot more 8 effort than was actually put into the various cases. 9 The best example I can think of is that there was 10 a whole team of us that did quite a lot of analysis of 11 previous bugs as part of the GLO proceedings in 2018, 12 and it required something of a similar degree of effort 13 to do that, if it had been understood that that was 14 required at the time. 15 Q. Did that occur to you at the time, that, "We're putting 16 all this effort and money into defending the Post Office 17 in a civil claim for damages, we didn't do this when we 18 were prosecuting dozens of subpostmasters"? 19 A. No, that didn't occur to me at that time, no. 20 Q. Quite striking, isn't it? 21 A. Looking back at it now, I can understand that but, as 22 I say, that was not something that occurred to me. As 23 I say, in all cases we were, or I was, responding to 24 specific requests to do specific bits of work and I was 25 being asked to do a lot more work in 2018 than I had 180 1 been in previous cases. 2 Q. In your witness statement, it's your third witness 3 statement, at paragraph 17, no need to turn it up, you 4 say you approached the various criminal cases where you 5 provided evidence in the same way as you approached 6 a problem in the day-to-day course of your work? 7 A. Yes. 8 Q. That's something you've said in summary today as well -- 9 A. Yes. 10 Q. -- and that you focused on the question of what the data 11 in a given case might show about the specific branch? 12 A. Yes. 13 Q. Rather than any other wider problems with Horizon? 14 A. Yes. 15 Q. You also say that you didn't understand that other 16 issues within the Horizon system, including issues that 17 had affected other branches, might be regarded as 18 important by a court? 19 A. Correct. 20 Q. You say that's all in the context that it is inevitable 21 that a system like Horizon would have bugs, errors and 22 defects in it? 23 A. Correct. 24 Q. Are you saying that in determining what had caused 25 a particular issue at a branch, you would only consider 181 1 the data that you had been specifically provided with? 2 A. Including the information we were given about what was 3 happening at the time but, yes, that was basically what 4 we'd look at. 5 Q. Would that data have been provided to you by the post 6 office or by Fujitsu? 7 A. Fujitsu. 8 Q. If you'd been conducting a fourth-level review, 9 a service support review, which is how you described you 10 were approaching this task, would you have looked only 11 at data from within the branch? 12 A. Yes. 13 Q. Not at any wider issues that had been identified 14 elsewhere in other branches, for example on that day or 15 in that week? 16 A. No, because each branch was effectively operating 17 independently. I mean, if it was a competence issue -- 18 if there was some sort of central fault, say, for 19 example, the interfaces of the banks had gone down or 20 something like that then, obviously, that would affect 21 many branches but a particular -- most -- most issues 22 really affected a specific branch at a specific time, 23 and that was my -- where I was coming from. 24 Q. I think at all times you knew that Horizon had bugs, 25 some of which would disclose themselves as they were 182 1 occurring, but some of which might only be discovered 2 long after the relevant event? 3 A. I felt that they would be visible at the time if you 4 looked at the events associated with the thing in the 5 branch. I think I later found out that there were some 6 cases where it wasn't that obvious until later but, 7 certainly, if you look at where I was coming from in 8 2010, then my experience was that, by looking at the 9 events and the message logs from the branch at the time, 10 then you would be able to detect if there'd been 11 a problem. 12 Q. But I think, more generally, you knew that Horizon had 13 bugs within it, some of which you would know about but 14 some of which may only be discovered a while after they 15 had had an operative effect? 16 A. I'm not sure that that was the way I was looking at 17 things back in 2010. 18 Q. Do you look at things differently now, then? 19 A. I am aware that there were a couple of bugs with Horizon 20 Online that did not manifest themselves immediately at 21 the time that they occurred but I'd not had that 22 experience of similar sort of occurrences on Legacy 23 Horizon. 24 Q. So you thought your job was to provide an explanation as 25 to whether something had gone wrong in a particular 183 1 branch, by looking narrowly at the data relating to that 2 branch? 3 A. Yes. 4 Q. If you concluded that the underlying cause of an issue 5 was a bug in Horizon, would you have had any hesitation 6 in saying so? 7 A. No. 8 Q. If you believed that you didn't have sufficient 9 information to arrive at a view, you needed additional 10 information, would you have had any hesitation in saying 11 so? 12 A. I -- all I said was I could find no -- what I was 13 basically saying was I could see no sign of any 14 problems. So I never said that there weren't any 15 problems. All I was saying was that I had not seen any 16 sign of any problems so, in effect, I was saying that 17 was the view that I was giving. 18 Q. That's not really an answer to the question, if you 19 believed that you didn't have sufficient information to 20 arrive at a view, and that you needed more information 21 in order to do so, would you have had any hesitation in 22 saying so? 23 A. If I -- I'm not very good at hypotheticals, so it's 24 a case -- I felt that I had enough information to say 25 that there wasn't -- that I had been unable to see 184 1 a problem. 2 Q. So, in every case in which you gave a witness statement 3 or evidence, you thought that you had full and complete 4 information, upon which to say there were no problems 5 with Horizon? 6 A. I wouldn't go that far. I made it clear in some of the 7 cases with the generic witness statement that that was 8 just at a generic level and I was expecting to be asked 9 to look at specifics and, as it turned out, I wasn't 10 asked to look at specifics so I was giving a general 11 level statement of the general behaviour of Horizon, not 12 the specific case of exactly what had happened in that 13 branch. I was expecting to be asked to look at specific 14 data, because my understanding was that in order to 15 prosecute someone, then it was necessary to provide the 16 data. 17 Q. If it was your view that there was more than one 18 explanation or possible explanation for an issue, and 19 one of those issues involved there being a bug, error or 20 defect, would you have had any hesitation in saying so? 21 A. I don't think so. 22 Q. Did you understand that it was part of your duty to say 23 that "There were a range of possibilities here, and my 24 assessment is this is the most likely, this is the least 25 likely", for example? 185 1 A. I didn't understand that, no. 2 Q. Thank you. Can we turn, then, to a separate topic: 3 bugs, errors or defects. Can we start with some general 4 questions, please. We've heard some evidence about 5 bugs, errors or defects that Fujitsu believed could be 6 lived with, ie did not believe disrupted the system to 7 a material extent, and we've heard some evidence that 8 bugs were not investigated for their root cause., 9 including where bugs were caused by Riposte and were 10 believed to be the possibility of Escher. Were you 11 aware of that? 12 A. I was aware that things had not been always thoroughly 13 investigated but I did not believe that such things had 14 actually caused a direct impact on the branch accounts 15 at the time. 16 Q. What investigations did you carry out to satisfy 17 yourself that they, the bug did not have an impact on 18 branch accounts? 19 A. I -- there were a number of cases in the early 2000s 20 where there were unexpected events occurring and, in 21 those cases, I looked at what was actually happening. 22 I mean, some of them were to do with the correspondence 23 servers, others were to do with the branches and, in the 24 cases I looked at, I couldn't see that there was -- they 25 were mainly happening out of hours and, therefore, there 186 1 wouldn't have been anyone doing anything at the branch 2 at the time and, therefore, I saw the thing as being 3 benign. 4 Q. Can we look, please, at FUJ00083596, and start with 5 page 2, please. If we scroll down, it's an email from 6 you to Brian Orzel. Can you remind us of who he was? 7 A. Brian Orzel was part of the EPOSS Development Team and 8 he was the main interface between the EPOSS Team and 9 Escher in terms of how the EPOSS counter software 10 operated. 11 Q. You say: 12 "Brian, 13 "You've obviously cracked the technology to enable 14 a list of PinICLs to be extracted to a file. How is it 15 done? Alternatively, can you generate a list of the 16 current Escher Development PinICLs into a file for me 17 please. We need to go through an exercise of checking 18 all outstanding PinICLs (at least the Bs and Cs) ..." 19 Is that a categorisation of their severity? 20 A. Yes. 21 Q. "... so as to factor them into [I don't know what this 22 is] M1R/S10 drops ..." 23 What were they? 24 A. They were -- I can't remember exactly what they were but 25 I mean we're talking about 2000/2001 sort of timescale, 187 1 in terms of when they were. There was a whole load of S 2 releases that started with S10 and worked up to S60 and 3 S80, which we've talked about before in terms of IMPACT, 4 and so on. 5 Q. So they are updates, effectively? 6 A. So they're updates, yes. 7 Q. "My belief is that many of them can be 'lived with' in 8 which case they need to be downgraded to D (or at the 9 highest C)", et cetera. 10 If you knew that there were Riposte issues that 11 could only be explored fully by reproduction or by 12 investigation by Escher, which you tell us about in your 13 witness statement, how could you be confident that 14 PinICLs of this kind could be lived with? 15 A. By looking at the effects that they had at the time, and 16 looking at the details of the individual PinICLs. 17 Q. What do you mean by "looking at the effects"? 18 A. Looking at what the description of what had actually 19 happened at the time, whatever was in the detail of the 20 individual PinICLs. 21 Q. Didn't that need, in order to be done to a satisfactory 22 state, reproduction of the fault or investigation by 23 Escher itself? 24 A. Not necessarily. It was a case of looking to see 25 exactly what the problem was and what was behind it to 188 1 see whether it was something that -- what the impact of 2 the problem was. 3 Q. Can we go to page 1, please. He replies: 4 "As to the PinICLs themselves, I think it makes good 5 sense to postpone them from last minute fix releases, 6 but given the amount of money we pay for support, when 7 we have a full regression cycle like (presumably) S10, 8 I would want them ALL fixed, including the D priorities. 9 No exceptions. Chris ..." 10 I think that's Chris Wannell? 11 A. Correct. 12 Q. "... has the final say." 13 So he appears to disagree with you and says he wants 14 them all fixed but says the final call is for Chris 15 Wannell? 16 A. Yes. 17 Q. What would Chris Wannell say? 18 A. He was the manager who had the managerial interface with 19 Escher and did the negotiations with Escher. I'm not 20 sure what his exact role was. 21 Q. Are you able to recall what happened with these bugs, in 22 particular, did Fujitsu live with them as you suggest or 23 were fixes found by Escher that applied? 24 A. I think some were lived with and some were fixed but 25 I can't remember which ones. 189 1 Q. Were there simply some bugs, then, even where the root 2 cause were unknown that Fujitsu and you were willing to 3 live with? 4 A. It depended on exactly what the impact of it was, and 5 that would have been done by a bug-by-bug evaluation, 6 and this starts like the start of a process where 7 an investigation was done into the individual PEAKs and 8 PinICLs to see exactly what the issue was, and I seem to 9 remember there was a later email which identified at 10 least one, that was an enhancement request, so it wasn't 11 exactly a bug, as such, it was just something that it 12 would be nice if the Riposte software did something 13 differently. 14 Q. So what was the outcome here? 15 A. That some things got fixed and some didn't but I can't 16 remember the details of the individual cases. 17 Q. The ones that didn't get fixed, was that without 18 an investigation of the root cause? 19 A. There would have been an investigation as to what the 20 systems were and whether that was something that was 21 likely to cause harmful effects in the system but 22 exactly the details of such investigations, clearly 23 I don't remember that at this stage. 24 Q. In order to say that there were no harmful effects, 25 didn't that require the bug to be reproducible by 190 1 Escher? 2 A. Not in all cases, I don't think. In some cases, yes it 3 would need to be done but, in others, then it would be 4 fairly clear from the description of it as to whether 5 this was -- what the harm was that it was causing. 6 Q. But it was the case that Mr Orzel's suggestion that they 7 all be fixed was not carried into effect? 8 A. I suspect not but I can't actually remember all the 9 details. 10 Q. Thank you. Can we move on to 2005, please. 11 FUJ00086315. Thank you. We're in August 2005, an email 12 to you from Martin McConnell; can you recall who Martin 13 McConnell was? 14 A. He was one of the developers in the first Development 15 Team -- 16 Q. I'm so sorry. It's from Mark Scardifield copied to 17 Martin McConnell to you. 18 A. Mark was the -- was the Development Manager in charge of 19 the EPOSS team at that time and Martin McConnell was one 20 of his developers. 21 Q. It's about a PEAK PC0121925 and release S80, shared SU 22 Cash declaration after -- what does the rest of it mean, 23 after "mig [greater than] TP"? 24 A. I can't remember exactly what that meant but, with the 25 introduction of IMPACT, one of the changes was to 191 1 actually migrate from branches producing a cash account 2 each week to producing a trading period every month. So 3 it was -- so the software had to be able to support both 4 working in the old way against the cash account and 5 then, when some reference data was changed for a branch, 6 it then started working and producing monthly trading 7 periods. 8 Q. So this is about, again, after S80 migration; is that 9 right? 10 A. It's -- this was during the testing of the S80 software 11 so, in other words, before it had gone live, and it was 12 saying that the software on that particular test rig was 13 configured that it was working in the branch trading 14 mode, as opposed to the cash account mode, is how 15 I would -- 16 Q. Mr Scardifield says in summary: 17 "Martin can see evidence of two different types of 18 failure. Firstly a failure of Message Port to alert 19 counter code when a new transaction has been inserted 20 into message store. And secondly a Riposte query 21 failing. My suspicion is that something (and I think we 22 have now eliminated archiving because there is no trace 23 in the Event Log) is hogging CPU time and the Riposte 24 errors are a secondary effect. There is some evidence 25 for this because at about the time of the second 192 1 failure, BUSY.EXE has logged an event, which it does 2 when there is a sudden change in system resources ... 3 "Needless to say this problem(s) cannot be 4 reproduced at will. 5 "we have run out of ideas on this one so any 6 thoughts appreciated." 7 Is this is an instance of a Riposte problem arising 8 which wasn't reproducible and which Fujitsu didn't 9 understand? 10 A. This was a system issue, and I believe we decided in the 11 end that there was some configuration had been set up 12 incorrectly on that particular test rig, which was 13 the -- the underlying cause of the problem. We were 14 unable to reproduce it. We did change some 15 configuration parameters, so as to make the -- if it 16 were to reoccur, to make the effects -- make it much 17 less likely to happen but I don't think we ever found 18 any evidence of that actually occurring in a live 19 system. 20 Q. In relation to a separate issue, you tell us in your 21 witness statement, your second witness statement, that 22 the impact of the Riposte lock had been much wider than 23 you had previously realised, and it was only by doing 24 work in 2018/2019, that you realised the full scope of 25 the Riposte lock issue, and that you became aware it 193 1 wasn't just the Callendar Square branch that had 2 suffered from discrepancies, but other branches where 3 there were counter occurrences of the timeout waiting 4 for the lock problem. Okay? I'm reading directly from 5 your witness statement. 6 A. Yeah, yeah. 7 Q. Okay. Would you accept that there may have been impact 8 arising from problems which weren't investigated at the 9 time but which may have been discovered with further 10 contemporary investigation and support? 11 A. I think a lot of effort was put into investigating this 12 particular problem and it was not possible to reproduce 13 the issue, and we did make some configuration changes to 14 the way in which the software worked to protect against 15 the fact that -- that the system performance could cause 16 issues. 17 Q. Can I turn; that can come down, thank you -- to the 18 first bug that I want to look at, data tree build 19 failure. 20 A. Okay. 21 Q. Can we start with some building blocks, this is for more 22 my benefit than anyone's: in summary, a node is a unit 23 of data structure that contains data? 24 A. Um ... 25 Q. That's not a very good start! 194 1 A. Right. The data tree, it -- 2 Q. You're starting far too high up. I'm starting much 3 lower down in my level of understanding. So a node 4 contains data -- 5 A. Yes. Okay, yes. 6 Q. -- and a data tree is a structure of nodes? 7 A. Yes. 8 Q. It's a hierarchical structure of connected nodes? 9 A. Correct. 10 Q. There is a root node, which has child nodes to it -- 11 A. Correct. 12 Q. -- and each of which children can have further children? 13 A. Correct. 14 Q. Each node can be connected to one or more child nodes? 15 A. Correct. 16 Q. The root node has no parent? 17 A. Correct. 18 Q. Aside from the root node, each node can only have one 19 parent node? 20 A. Correct. 21 Q. Legacy Horizon created summary accounts by building data 22 trees and harvesting the information? 23 A. I'm not sure I'd have used the term "harvesting" in that 24 case but I think that's a reasonable word to use. 25 Q. In Legacy Horizon, transaction data was stored as 195 1 a message or messages in Riposte? 2 A. Correct. 3 Q. When Legacy Horizon generated reports or prepared 4 accounts, it would scan the message store for relevant 5 transactions. 6 A. Correct. 7 Q. The relevant messages identified by Horizon would be 8 stored in one of the nodes of the data tree? 9 A. One or more of the nodes of the data tree. There 10 would -- there would be number of different relevant 11 nodes it would be stored under. 12 Q. So the data tree in these circumstances was used to 13 populate and to organise the data for Horizon to produce 14 accounting reports? 15 A. Correct. 16 Q. You say in your second witness statement -- it's 17 paragraph 36, no need to turn it up -- that Legacy 18 Horizon would reuse the data tree once it had been built 19 with new messages being added; is that right? 20 A. Correct. 21 Q. Therefore, an example, would you agree, of a data tree 22 build failure is where Legacy Horizon missed a node such 23 as a node containing payments in, resulting in the 24 accounts being inaccurate? 25 A. There were different sort of errors at different times 196 1 and, as time went on, more protection was put in to 2 check against things like that. So I think -- 3 Q. That would be an example, sorry, to -- 4 A. That was an example back in around 2000, and I think 5 checks and balances were then put in place to detect 6 such problems, should they occur in the future. 7 Q. Now, the Inquiry is aware that Fujitsu came to know 8 about data tree build failures and that they could lead 9 to discrepancies very early on in the life of Legacy 10 Horizon. I think you know that there were such data 11 tree build failures identified in KELs in 1999 and 2000? 12 A. I'm aware of that now. I think I only became aware of 13 that when I looked at the stuff in 2018. 14 Q. Can we look, please, at FUJ00056677 -- sorry, before we 15 get to that, I should just explore that last answer. 16 How is it that you only became aware of the KELs and the 17 PinICLs of 1999 and 2000 when you came to look at it 18 nearly two decades later? 19 A. Because I hadn't been involved in EPOSS at that time, as 20 we discussed this morning. So those were issues when 21 I was -- didn't have any responsibility or connection 22 with EPOSS in around the year 2000 or so. However, as 23 part of the GLO, then I was asked to look at some of the 24 PEAKs from that area and, in general, I relied on the 25 information that the SSC came up with at that time. But 197 1 I -- as a side effect of that, I became aware that there 2 had been those problems back in about the year 2000 that 3 I was not aware of until that time. 4 Q. When you became involved in data tree build failure 5 issues in between those times, ie '99/2000 versus 6 2018/2019, did you not access PinICLs and KELs -- 7 A. Yes. 8 Q. -- that we're going to look at? 9 A. Yes, I could have looked at the things at the time, it's 10 just I had no reason to, which is why I wasn't aware of 11 the problems until 2018, or so. 12 Q. This is a PinICL PC0033128, raised, I think, on 13 10 November 1999 -- can we see that -- 14 A. Yes. 15 Q. -- the date that it's opened, relating to losses, we can 16 see on the left-hand side, at branches in Dungannon; can 17 you see that? 18 A. I can see that, yes. 19 Q. I think, if you read the body of the PinICL: 20 "Outlet has a discrepancy [about six or seven lines 21 in] of £43,000 after balancing [stock units] and doing 22 office snapshot. 23 "... Steve Warwick development is investigating why 24 this misbalance occurred. 25 "Immediate impact of this weeks balance has been 198 1 addressed but [Post Office] are concerned that the cause 2 is still unknown and this will affect this and other 3 outlets." 4 If we go over the page, please, five lines in: 5 "Steve Warwick believes that this may be an isolated 6 incident ..." 7 Then about five lines further on: 8 "I will monitor progress of the incident which is 9 now with the EPOSS Development teams for investigation." 10 Five lines on: 11 "I have talked with development [referenced] this 12 problem. It is seen as a one-off. No fault can be 13 found ..." 14 Then over the page, three lines in: 15 "... no further occurrences have been reported from 16 Dungannon or similar ..." 17 Then nothing on that page. 18 Next page, please: nothing on that page. Next page, 19 please, the middle of the page: 20 "The problem is currently back with Development for 21 further investigation." 22 Development, who would that refer to? 23 A. That would be the EPOSS Development team, so I think 24 Steve Warwick's name was mentioned earlier and he was 25 the main designer and would almost certainly have been 199 1 the person who would have been looking after this 2 particular incident. 3 Q. Then over the page. Just over halfway down: 4 "The similar occurrence is currently an incident 5 I am investigating if the similarities are such that we 6 can add this into the problem." 7 Then over the page -- sorry, I should have read the 8 bottom of that last page: 9 "Discussed this issue at XDMF ..." 10 Is that some sort of conference? 11 A. I've no idea. I don't recognise that term. I'm 12 guessing the "MF" stands for management forum but I'm 13 just speculating. 14 Q. Something like that? 15 A. Yeah. 16 Q. It's thought that a similar incident has, over the page, 17 occurred at Yate Sodbury. So that's a different FAD 18 code given there, so a different branch, to the value of 19 £52,000, "the problem will remain open". 20 Then four lines on: 21 "A further occurrence has arisen at Appleby in 22 Westmoreland at [just under £10,000]. 23 "Chased call with Development and spoke with Martin 24 McConnell. [He] has made extensive investigations on 25 the issue [and is] unable to recreate the fault. 200 1 "Escalated to Chris Wannell (Development) who will 2 discuss options with Martin and Steve." 3 At the foot of the page: 4 "... a diagnostic fix was being prepared and was to 5 be submitted to the next Release Management Forum to 6 authorise release into the live estate." 7 Then, if we carried on reading -- I'm not going to 8 carry on reading -- it's reported that a fix was applied 9 in 2000 by CI14. 10 A. Yes, I believe so. 11 Q. In your statement, you say that the effects of locking 12 errors were more widespread when you were considering 13 the Callendar Square bug, yes? 14 A. Yes, but I'm not sure that was necessarily a locking 15 error that was the root cause for that. 16 Q. You say, in your statement, that that reflects the 17 nature of fourth line support, where not every PinICL on 18 the same or a similar issue was referred back to the 19 person who may have dealt with the same or similar 20 issues previously? 21 A. Correct. 22 Q. Was that a function of the way that third line support 23 was arranged? 24 A. Yes, but, I mean, I suspect what you're saying is this 25 particular PinICL refers to other occurrences and 201 1 I think the reason for that was it was felt they could 2 provide additional information to help diagnose the 3 problem. 4 Q. When slow running issues became noticed in 2006, was any 5 connection made between this series of issues that we've 6 just looked at in this PinICL and those slow running 7 issues? 8 A. No, because I believe that the issues that were found 9 back in 2000 had been resolved and sorted out at that 10 time and the issues that were occurring in 2006 was 11 a different problem, it was to do with the fact that, 12 when we moved from a weekly cash account, then, for 13 a week's worth of transactions you've got a certain 14 amount of data. When you move to a monthly branch 15 trading period, then you've got four or five times as 16 much transactions, and that's why performance again 17 became an issue in 2005, because of the fact that there 18 was just so much more data being processed in terms of 19 producing the data tree, as it then -- as it was then 20 operating. 21 Q. So Fujitsu's corporate statement to this Inquiry splits 22 the data build tree failures into three issues. Can 23 I see whether what you're saying essentially mirrors 24 that, that the classification is as follows: 25 Firstly, issue 1 happened in 1999 and 2000 and was 202 1 supposedly fixed as a result of the investigations that 2 we've just read on that PinICL? 3 A. That was my understanding, yes. 4 Q. Secondly, issue 2, or 2.1, arose as a result of testing 5 and didn't affect the live estate? 6 A. And I think that reflects the email you took me to of 7 a few minutes ago, yes. 8 Q. Then, thirdly, issue 2.2 describes a single incident 9 involving what is said by Fujitsu to be a transient 10 discrepancy that was itself fixed in 2006? 11 A. I believe that was a separate problem introduced by some 12 new software later on. I think in my witness statement 13 I described those as three different issues, and I was 14 disagreeing slightly with the way that the Fujitsu 15 corporate statement was laying that issue out. 16 Q. Let's look at your witness statement, it's the second 17 witness statement, the one ending in "0200" at page 11, 18 please. Page 11, paragraph 35, you say: 19 "All three categories of this problem [that's the 20 data tree build failure] concern the data server 21 component in Riposte but I believe each problem was 22 completely different from the other. The first category 23 seems to have occurred in 1999-2000. I do not believe 24 I was aware of this first category at the time." 25 Now, I have looked at the PinICLs for that first 203 1 category -- just for the transcript reference, I believe 2 they are FUJ00062016, FUJ00066601 and FUJ00086553, and 3 I can't see your name on any of them and, therefore, 4 I am not going to ask you questions about them, okay? 5 A. Thank you because I wouldn't know much about them. 6 Q. "The second category [you continue] occurred in 7 2005-2006. I believe that I was aware of this second 8 category at the time." 9 So let's look at that, the second category of data 10 tree build failure, by looking at PEAK PC0121925, which 11 is POL00028867. Look at the top of the page, you'll see 12 it should be raised on 13 June. Yes, if you just look 13 under "Progress Narrative"? 14 A. Yeah. 15 Q. Date raised, call open 13 June 2005. If we scroll down 16 a little bit: 17 "[Post Office] have raised the following incident: 18 "'A cash declaration was made in "stock balancing" 19 for the amount displayed on the snapshot. When the cash 20 variance was checked afterwards a gain of [£45-odd] was 21 displayed'." 22 A. Can I clarify there, please, that when we're talking 23 about Post Office has raised that, if you look slightly 24 above it talks about "E2E". So E2E was end-to-end 25 testing, so this was a test rig and this was a Post 204 1 Office testers on a test rig have identified this 2 problem. This wasn't a problem from a live branch. 3 Q. Got it. So is this Post Office staff embedded in the 4 test rig? 5 A. Correct. 6 Q. Okay. 7 A. So it was one of the stages of testing that was done, 8 was the Post Office staff would do their own independent 9 testing of new software before it went live. So this is 10 June 2005. So IMPACT S80 didn't actually go into live 11 estate until about August/September time. 12 Q. If we carry on reading, over the page, please, I think 13 we see, if we scroll on to 4 July, stop there -- thank 14 you -- 4 July, middle paragraph of Martin McConnell's 15 third entry. If that can be highlighted, it's 16 two-thirds of the way down the page: 17 "After spending more than a week on this I don't 18 know there is anything more that can be done because: 19 "1. The code is dependent on Escher's Notify 20 mechanism, which MAY itself beholdant [sic] of perhaps 21 some NT or other system failure. 22 "2. We cannot reproduce this at will with any 23 defined or prescribed pattern. 24 "3. We can recommend a course of action to allow 25 a clerk to get a fresh rebuild of the tree if he/she 205 1 finds anything odd, simply by logging off and on again 2 the dataserver tree will automatically refresh." 3 So what role did Martin McConnell perform? 4 A. He was one of the developers and, in fact, he was the 5 expert on the data tree build. You probably recognise 6 his name from 2000. 7 Q. So he is saying that he's not sure anything more can be 8 done and, in any event, a solution is to turn the 9 machine off and on again? 10 A. Yes, but I think we did actually do more than just that. 11 Q. If we go forward to page 3, please. If we scroll down, 12 please, and look at his entry at the foot of the page, 13 I'm not going to read all of the technical explanation 14 to start with but, just at the end of that paragraph, 15 he's referring to a second occurrence of the fault and 16 he is saying: 17 "... this second occurrence is a different category 18 and perhaps MORE disturbing. I am not able to tell 19 whether: 20 "1. A silent feature has occurred within Riposte. 21 "2. The failure has occurred within Riposte and 22 notified to us. 23 "3. Riposte is behaving okay but EPOSS has failed 24 within the loop." 25 So what was done? 206 1 A. I think we changed the configuration parameters to 2 increase the buffering, so that if the counter 3 application was a bit slow in processing the messages 4 that were being picked up, then there was a larger 5 buffer of messages that could be buffered up to allow 6 things to catch up. I'm probably not explaining that 7 very well. 8 Q. No, I think that's sufficient. More importantly, were 9 you able to tell if the problems that had been 10 identified in this test environment ever emerged and 11 affected the live estate? 12 A. I don't believe we ever saw any sign of a problem like 13 this on the live estate and I think we did find some 14 problems with the way that that particular test rig had 15 been set up at the time. 16 Q. Can we turn to your second witness statement at page 15, 17 please. You say on page 15, at the foot of the page, at 18 paragraph 51: 19 "... I do not know whether the problems identified 20 in the test environment ever emerged and affected live 21 data in actual branches, or whether the fix rolled out 22 for the second PEAK proved to be effective. In the 23 absence of further documents, I cannot say whether these 24 problems had an actual impact on branch accounts." 25 Was that the case at the time? 207 1 A. Yes, as I say, as part of the investigation into the 2 problem that we were looking at before, the testers 3 found another reproducible occurrence of the thing and 4 we found a genuine bug in the code that was then fixed, 5 I think it was in August from memory, and that fix went 6 in place where we found that the notification mechanism 7 had been accidentally switched off in the code and that 8 problem did get fixed, and there was some further 9 documentation about that. 10 Q. You say: 11 "Reflecting on the PEAKs now, and with the benefit 12 of hindsight, I believe there were probably other 13 processes on the test rig counter that slowed down 14 Riposte and made it very unpredictable." 15 You say: 16 "As [you] mentioned in [the] Test Report, it is 17 possible that these processes were caused by the testing 18 environment and would not have occurred in actual 19 branches." 20 A. That was my view, yes, at the time as well. 21 Q. Was that hypothesis tested in the live environment? 22 A. I'm not aware of seen any such issues happening the live 23 environment. 24 Q. That's an answer to a different question. 25 A. Okay. 208 1 Q. Was what had been observed in the test environment 2 tested in the live estate? 3 A. I'm not quite sure how you could test that in the live 4 estate, so -- 5 Q. Actively looked for? 6 A. It would -- the simple answer is I don't know. 7 Q. Would failures in the data tree build cause Legacy to 8 produce inaccurate accounting reports? 9 A. It could do in the short-term. However, if the -- if it 10 missed a transaction one time, then it -- there should 11 be a corresponding failure the other way the next time 12 the -- a report was produced. 13 Q. Should be or would be? 14 A. Certainly should be. I'm not -- I can't guarantee that 15 it would be. 16 Q. Would the failure to build the data tree have an effect 17 on the accuracy of ARQ data? 18 A. No. 19 Q. Why not? 20 A. Because the ARQ data was a record of the transactions as 21 recorded in the message store, not how they were built 22 up into a report. 23 Q. Thank you. Can we turn to the third issue or 2.2, as 24 Fujitsu call it in their statement. 25 A. Yeah. 209 1 Q. That can come down from the screen. Thank you. 2 I think that this is demonstrated through a PEAK. 3 If we can look at FUJ00086456. Thank you. This should 4 be PEAK PC0132133. It is. You can see that it's 5 raised, by looking in the narrative, on 10 February 6 2006. 7 A. Yes. 8 Q. If we see the summary: 9 "[Postmaster] states that she had a discrepancy that 10 seemed to become greater over the course of 20 minutes. 11 Then a few minutes later the discrepancy vanished and 12 normal figures remained normal. 13 "[She] noticed this on daily cash report previews. 14 "... Referred by NBSC. 15 "Discrepancy was at its highest £1,000 shortage. 16 But [postmaster] insists within minutes the cash balance 17 preview stated that the balance was fine and the 18 discrepancy was gone." 19 If we go forwards, please, over the page. Thank 20 you. Just stop there and scroll down to David Seddon's 21 entry at the foot of the page. He says he can see that: 22 "... on 4 February the [postmaster] made numerous 23 cash declarations [on a stock unit], the [postmaster] 24 was declaring all sorts of figures and getting different 25 [variations]. What does stick out though is the cash 210 1 declaration done on counter 2 ... the [postmaster] 2 declared [£1,200-odd] and the system declared 3 a [negative] value of [£1,400-odd]. Moments later the 4 [postmaster] logs on to counter 1 and declared 5 [£1,200-odd] cash once again. This time the system 6 calculated a [negative] of [£93-odd], which is roughly 7 what the postmaster expected, even though no 8 transactions had been done between these two 9 declarations." 10 Then skipping over a paragraph: 11 "It would appear that when working out the cash 12 [declarations] on counter 2 the system has used an old 13 'data tree' (the one it used at the earlier trial 14 balance) rather than creating a new one, so the 15 discrepancies were wrongly calculated. It wasn't until 16 the [postmaster] later moved to counter 1 that a new 17 'data tree' was produced and the discrepancies were 18 calculated correctly." 19 Can we summarise subsequently, rather than reading 20 it all, is it right that a fix was developed and 21 released later in 2006? 22 A. Correct, and I can't remember the exact -- why this 23 problem introduced, but I believe that the data tree was 24 actually being used for a new purpose as part of 25 a change that was made at around that time and, as 211 1 a result of this problem, then a different mechanism was 2 used, rather than reusing, effectively, a copy of the 3 data tree. And I think that gets referred to later in 4 this PEAK. 5 So the whole bit of software that was actually 6 calculating these cash levels was changed to not use the 7 data tree any more, because of the problems that was 8 identified in this PEAK. 9 Q. Thank you. Can we look at an email concerning this 10 problem, FUJ00086462. This is an email exchange 11 involving Anne Chambers, you, Graham Welsh and John 12 Burton. What did Graham Welsh do? 13 A. Graham Welsh was in Customer Services. I can't remember 14 his exact role but he was one of the people that used to 15 liaise with Post Office, and he was a fairly senior 16 manager at the time. 17 Q. John Burton? 18 A. He was a Programme Manager and, at that time, he was 19 probably my direct line manager. 20 Q. This exchange details problems in the scanning and 21 building of the data tree during balance? 22 A. I think this is relating not to the problem we've just 23 looked at and this isn't looking at a bug as such; it 24 was more to do with the fact that, because branches had 25 moved to the branch trading statement and therefore were 212 1 building up four or five weeks' worth of transactions, 2 it was taking an awfully long time to actually produce 3 reports. 4 So it was -- in huge branches, it was taking up to 5 an hour to do that and what Anne is reporting the fact 6 that she was getting complaints from postmasters as to 7 how long it was taking. 8 Now, we had always made it clear to Post Office that 9 one of the consequences of moving from a weekly cash 10 account to a monthly branch trading statement meant that 11 it was would take longer to balance, if you only 12 balanced once a month. So it was something that we'd 13 anticipated at the time and, as a result of this, 14 I think I did do some further investigations to see was 15 there anything we could do to improve the performance. 16 And I think that's what this email exchange is to do 17 with. 18 Q. I see. Can we look from page 2, then, please, and 19 scroll down, please, an email from Kimberley Yip to 20 Graham Welsh and Dave Hulbert, with a heading, "Horizon 21 System Performance". Second paragraph: 22 "I have been contacted again by the [Post Office] 23 Service Line to obtain an update on progress on the 24 current Horizon system performance issues. 25 "One particular branch has been escalated to me ... 213 1 and the last rollover timings have been sent to me by 2 Anne Chambers, see below: 3 "From 17.00 the branch started printing the daily 4 reports and this continued until 18.30. They then 5 declared stamps and cash, and pressed the balance report 6 at 18.37. The trial balance was not printed until 21.12 7 (ie over 2.5 hours later). Much of this time the system 8 was processing the month's transactions. There is a gap 9 between about 19.30 and 20.05 where it may have been 10 waiting for input from the [postmaster], but I can't be 11 certain. 12 "After the trial balance, the report was abandoned, 13 presumably because the [postmaster] needed to check and 14 resolve the discrepancies. At 21.27 cash and stamps 15 were redeclared (with some variation from the original), 16 and at 21.28 the balance report button was pressed 17 again. The second Trial Balance was printed at 22.58 18 (1.5 hours later) and the Final Balance at 23.04. 19 "I've looked at what was going on during the balance 20 report production. 21 "There was nothing out of the ordinary, apart from 22 the very large number of transactions being processed 23 (about 40,000). The number of transactions processed 24 per second was rather less than we sometimes see but not 25 significantly so, apart from the period 19.00 until 214 1 19.10 when the counter end-of-day processes were 2 running. 3 "Anne also provided me with some recommendations 4 which I have passed on to the branch, and I will ask FS 5 ..." 6 FS? 7 A. I assume Fujitsu Services but I'm not sure. 8 Q. "... to do a similar exercise to the one above 9 (ie provide timings) when the next TP rollover is 10 completed, 14 June, to see if there are any significant 11 improvements. I have been told about another branch so 12 I am hoping to do a similar exercise. In both cases the 13 rollover times do seem excessive and my worry is that 14 these are not isolated incidents. So in terms of the 15 time it is taking branches to complete the balance 16 process, can [it probably is Fujitsu] provide me with 17 details on what constitutes an acceptable length of 18 time, for example if it takes 4 hours then this is 19 reasonable, or if it's more than 5 hours then it needs 20 investigating. This will then give me a better 21 understanding of what I should be passing on to 22 [Fujitsu] or if I should be passing on the 23 recommendations to implement." 24 Then if we scroll up, please. 25 "Anne, can you please comment ..." 215 1 Scroll up, and scroll up again. Anne Chambers' 2 reply -- you're not copied in yet: 3 "I've looked at many branches now, and they range 4 from very slow to horrifically slow when rolling over 5 stock units. It does vary depending on the particular 6 process followed at each branch, and if you break it 7 down into various components each may appear to be 8 (just) within 4 times as long as the weekly rollover 9 used to be, but the impact on the [postmasters] is 10 horrible. 11 "There have been some piecemeal changes to try to 12 improve certain areas, but most of these have made 13 little improvement, and overall, may have been a waste 14 of effort. 15 "... there are two main problems: 16 "1. The balancing process repeatedly scans and 17 rebuilds the data tree. This was identified as 18 a problem at least 6 months ago, and improvements to 19 this are, I think, what Gareth is proposing. 20 "Counters are inadequate for the applications now 21 being run on them and do run generally slowly at times. 22 This hasn't really been fully investigated, and is 23 really difficult to quantify or prove that is 24 happening -- the only evidence is what the [postmaster] 25 reports. It is however adding to customer 216 1 dissatisfaction and could get worse even if we improve 2 balancing. 3 "I am not at all happy about fobbing postmasters off 4 and telling them that the system is working as designed 5 when it is plainly inadequate for the job. I am also 6 very unhappy that it has taken six months even to get to 7 the point of starting to consider whether [Post Office] 8 will pay for improvements. 9 "I too would like guidance on when 2nd and 3rd line 10 support should investigate further. Our current 11 response has to be 'yes we know balancing is very slow, 12 it is being investigated' -- what else can we say?" 13 Was it the case that subpostmasters were being 14 fobbed off? 15 A. I have to take Anne's word for that. I didn't have 16 direct contact with postmasters. 17 Q. And being told that the system was working as designed, 18 when, in fact, it was inadequate for the job at hand? 19 A. We had made it clear, as part of the changes to IMPACT, 20 that one of the consequences of moving from a one-week 21 cash account period to a four or five-week branch 22 trading period was that balancing would take 23 a significant -- ie four or five times as long, unless 24 they did intermediate balancing in the weeks in between. 25 Q. Would you agree with the view that she expresses here 217 1 that the system was inadequate for the job: plainly 2 inadequate for the job? 3 A. Clearly, it was causing problems for the postmasters but 4 what we were responding to is what Post Office had 5 actually asked us to do, and we had warned them that 6 there would be performance implications if they carried 7 on, made the changes as had been proposed, and suggested 8 that we could investigate improving the performance but 9 that would require additional cost to actually do that. 10 Q. You say in your witness statement that you wouldn't view 11 this category of problem as a recurrence of the first 12 category of the data tree build failure that we saw; is 13 that right? 14 A. Correct. This was just things going slowly, not 15 actually going wrong. They were just going very slowly. 16 And, clearly, I can understand the postmasters weren't 17 happy with that but they were actually coming up with 18 the right answer, but just very slowly. 19 Q. So where it says in her numbered paragraph 1 "The 20 balancing process repeatedly scans and rebuilds the data 21 tree", that's not connected to either the first, second 22 or third categories of data tree -- 23 A. No. 24 Q. -- build failure that we looked at earlier; is that 25 right? 218 1 A. Correct. 2 Q. If we scroll up, please. You're included on this chain. 3 Mr Burton replies: 4 "I have reviewed Gareth's [your] feasibility report 5 and costings this morning ... [Your] report is based on 6 a great deal of prototyping work that has been done over 7 the last few months -- of the order of 100 man-days. 8 The work looked at a number of options, and has homed in 9 on the one that gave the best improvements ... 10 "The report should go into [Post Office] next week. 11 It'll then be up to them if they want to pay us to do 12 the work. If they decide to go ahead, we're looking at 13 a likely delivery date of first calendar quarter in 14 2007. That would give around 2 years of useful life 15 before being overtaken by [Horizon Online]. 16 "I understand your frustration at having to deal 17 with irate postmasters and having to tell them that the 18 system is working to ... spec. We can only hope that 19 [the Post Office] do agree to funding this work, so that 20 you have something positive to say." 21 Was anything done? 22 A. I honestly can't remember. Until I saw this email, I'd 23 forgotten about the fact that I'd even done some further 24 investigations into it and I don't believe the report 25 has been made available to me to even know what I was 219 1 suggesting. 2 Q. On the third category of the data tree build failure, 3 when you were -- or indeed anyone else in Development 4 who was looking at that -- was examining it, would you 5 look into knowledge repositories within Fujitsu to see 6 whether there had been similar problems in the past? 7 A. Not necessarily. I think we relied on the fact that 8 people knew what had been looked at in the past. 9 Q. Was there a system in place at Fujitsu that allowed 10 technicians to ensure that, when they were looking at 11 a problem, they could investigate whether a related, 12 prior problem existed? 13 A. There probably was a simple way of doing that. I can't 14 remember one. 15 Q. The Inquiry has heard evidence that the PinICL system, 16 and the PEAK system that followed it and, indeed, the 17 KEL database, were searchable; is that right? 18 A. Yes, they were searchable. I'm not sure that I was very 19 good at knowing how to search it. If I wanted to search 20 for things, I would tend to talk to someone in the SSC, 21 "can you find information about this sort of thing?" 22 Q. So in 2006, if a technician had searched for "data tree 23 build" or "data tree build failure" -- 24 A. They probably could have found the PEAKs from 2000 if 25 they had applied the right sort of search criteria, yes. 220 1 Q. So they could have found evidence of earlier problems, 2 allowing a potentially fuller picture -- 3 A. Yes. 4 Q. -- to have emerged? 5 A. Yes. 6 Q. Was that kind of search ever done routinely by fourth 7 line support? 8 A. I don't know, is the simple answer. 9 Q. Did you ever do that kind of search? 10 A. I don't think I would have needed to, in the sort of 11 fourth line support that I needed to do. It was 12 something that maybe would have been done more by the 13 actual developers, rather than the designers. 14 Q. You tell us, as a final reflection in your second 15 witness statement, that you are: 16 "... struck by the lack of support that was afforded 17 to subpostmasters when they got into difficulties and 18 that there would have been a real benefit in having 19 a team whose function it was to have sight of all issues 20 across the different levels of support who could have 21 drawn together the PEAKs and had oversight of what 22 frontline support were fielding, and who understood the 23 practical ramifications of issues upon those who worked 24 in the branches and ensured that monitoring was 25 working." 221 1 Is that right? 2 A. That is how I see things now, looking back from 3 2023/2024. It was not something that had occurred to me 4 at the time. Perhaps it should have done but it is 5 certainly becoming very clear now, from what I've seen, 6 both with the GLO and with the work that's been done for 7 the Inquiry. 8 Q. Fourth line support was really the final stage in the 9 process on very difficult problems that couldn't be 10 fixed by the earlier lines of support, correct? 11 A. Correct. 12 Q. There went the problems that the SSC couldn't understand 13 or didn't have the skill to remedy? 14 A. Not necessarily. If a code change was needed, it had to 15 go to fourth line support because they were the only 16 people who actually had access to changing the code. 17 All third line support could do, and earlier lines of 18 support could do, was actually identify avoidance 19 actions or where people had been following incorrect 20 process, and that sort of thing. If there was actually 21 a bug in the code that needed fixing, then it had to go 22 to fourth line support. 23 Q. For problems to get to fourth line support, they would 24 have had to have got through the HSH or the NBSC, and 25 through to the SSC; is that right? 222 1 A. Only if they came from live because a lot of faults were 2 identified from test systems and things like that, and 3 some things were identified internally. So there were 4 a number of issues that were identified by SMC, who were 5 effectively second line support, that would then be able 6 to raise them through to be investigated by third line 7 support without involving the actual postmasters. 8 Q. Just like it would be reasonable to expect that, when 9 you were investigating a current problem, you might 10 search the PEAKs or the PinICLs or the KEL database for 11 the existence of past problems of a similar kind, if you 12 were giving evidence in court, you could use those 13 facilities to search for issues or problems of a kind 14 suggested by the subpostmaster, couldn't you? 15 A. Yes, I guess I could have done. 16 Q. But you didn't? 17 A. I didn't. 18 MR BEER: Sir, it is just coming up to 4.30. That would be 19 an appropriate moment to break. It's been a long day. 20 SIR WYN WILLIAMS: Yes. 21 MR BEER: We are reconvening tomorrow at 10.05. 22 SIR WYN WILLIAMS: All right. I'll see you all tomorrow at 23 10.05. 24 Mr Jenkins, I'm sure you wouldn't, but I should just 25 tell you that it's not appropriate for you to discuss 223 1 your evidence during the adjournments. All right? So 2 that obviously applies with greater significance when 3 you go home for the evening. 4 A. Yes. 5 SIR WYN WILLIAMS: Fine. 6 MR BEER: Thank you very much, sir. 7 (4.28 pm) 8 (The hearing adjourned until 10.05 am the following day) 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 224 I N D E X GARETH IDRIS JENKINS (sworn) ..................1 Questioned by MR BEER .........................1 225