1 Friday, 28 July 2023 2 (10.00 am) 3 MR BLAKE: Good morning, sir. Can you see and hear 4 me? 5 SIR WYN WILLIAMS: When I unmute myself I can 6 confirm both. 7 MR BLAKE: Thank you very much, sir. This morning 8 we're going to hear from Mr Jan Holmes, who is 9 appearing remotely from abroad. 10 SIR WYN WILLIAMS: Yes. 11 MR BLAKE: Can Mr Holmes be sworn, please? 12 JAN ROBERT HOLMES (affirmed) 13 Questioned by MR BLAKE 14 MR BLAKE: Thank you very much, can you give your 15 full name please. 16 A. Yes, it's Jan Robert Holmes. 17 Q. Thank you Mr Holmes. Thank you very much also 18 for returning today. You appeared during our 19 Phase 2 of this Inquiry and you gave a witness 20 statement in Phase 2. I'm not going to bring it 21 up on screen but, for the transcript, it's 22 WITN04600100. 23 You have since then produced two further 24 statements and I'm just going to take you to 25 those. The first of them is dated 27 January 1 1 2023. Do you have a copy of that in front of 2 you? 3 A. Yes, it's on my screen. 4 Q. Thank you. For the purpose of the transcript 5 it's WITN04600200. 6 If I could ask you to turn to the final 7 page, page 8, is that your signature? 8 A. Yes, it is. 9 Q. Is that statement true to the best of your 10 knowledge and belief? 11 A. Yes, it is. 12 Q. Then there is another statement, a third 13 statement, dated 16 May 2023, that's 14 WITN04600300. Can I ask you to turn to the 15 final page of that statement, please. 16 A. Yeah. 17 Q. That's page 14. Is that your signature at the 18 end? 19 A. Yes, it is. 20 Q. Is that statement true to the best of your 21 knowledge and belief? 22 A. Yes, it is. 23 Q. Mr Holmes, when you appeared in Phase 2 we went 24 through your career history, so I'm not going to 25 repeat that, save to say you were the Audit 2 1 Manager at ICL from 1997. There was a brief 2 period where you went out of that role in 3 2000/2001, but you returned in 2001 and I think 4 you stayed at ICL until 2008; is that correct? 5 A. Yes, that's correct. 6 Q. Thank you. I'm going to begin with 7 a whistlestop tour of some of the documents that 8 we looked at in Phase 2, just to refresh your 9 memory of some early incidents relating to 10 Horizon. Can we start by looking at 11 FUJ00080690. This is a document that will be 12 familiar to many people in this room, it's the 13 report of the EPOSS PinICL Taskforce, of 14 autumn -- which took place in autumn 1998. 15 I think you'll recall that in the top right-hand 16 corner it seems as though you refreshed your 17 memory in some way or had cause to look at the 18 document in May 2001 as well. 19 A. Yes. 20 Q. That's a document that you wrote. If we scroll 21 down we can see your name there. 22 A. Yes, with David McDonnell as well. 23 Q. With David McDonnell. Absolutely. Let's look 24 at page 7 of that report. As I say, the 25 contents will be familiar, so we won't stay on 3 1 this document for a long time but let's just 2 have a look at the first paragraph. It says 3 there in the report: 4 "It is clear that senior members of the 5 Taskforce are extremely concerned about the 6 quality of code in the EPOSS product. Earlier 7 this year the EPOSS code was re-engineered by 8 Escher and the expectation is that the work 9 carried out in Boston was to a high standard and 10 of good quality. Since then many hundreds of 11 PinICL fixes have been applied to the code and 12 the fear is that code decay will, assuming it 13 hasn't already, cause the product to become 14 unstable. This presents a situation where there 15 is no guarantee that a PinICL fix or additional 16 functionality can be made without adversely 17 affect another part of the system." 18 Then it goes on to say: 19 "... a more worrying concern from the 20 Programme's perspective should be reliance on 21 the EPOSS product for its current state as 22 a basis for planning and delivery." 23 If we scroll down to the next paragraph, it 24 says there: 25 "Lack of code reviews in the development and 4 1 fix process has resulted in poor workmanship and 2 bad code." 3 If we go on to page 17, there is a section 4 that I think was written significantly by 5 Mr McDonnell. If we scroll down, it's a section 6 on existing code. 7 A. Yes. 8 Q. Just to refresh your memory, it says there: 9 "Although parts of the EPOSS code are well 10 written, significant sections are a combination 11 of poor technical design, bad programming and 12 ill thought out bug fixes." 13 If we scroll over the page, you'll remember 14 those comments: 15 "Whoever wrote this code clearly has no 16 understanding of elementary mathematics or the 17 most basic rules of programming." 18 Mr McDonnell's subsequent evidence to the 19 Inquiry was that the EPOSS Development team was, 20 in his words, like the Wild West and was the 21 joke of the building. I can't recall, did you 22 see Mr McDonnell's evidence on that? 23 A. Yes, I did. 24 Q. Yes. Another document that I will refresh your 25 memory with is the CSR+ development audit 5 1 document, that's FUJ00079782. This is now in 2 October 1999. If we scroll down, we can see 3 that that again is a document written by 4 yourself. 5 A. Yeah. 6 Q. Can we turn to page 19 of that document, please. 7 There's a section there on the "Electronic Point 8 of Sale Service", the EPOSS. If we scroll over 9 the page, I'm just going to read that top part, 10 it says: 11 "The figures indicate that the problems 12 facing the EPOSS during the Taskforce period 13 have not diminished [we're now in late 1999]. 14 Of greater concern are the non-EPOSS PinICLs 15 within the group suggesting that there are still 16 serious quality problems in this vital, customer 17 facing element of the system. 18 "The EPOSS Solutions Report made specific 19 recommendations to consider the redesign and 20 rewrite of EPOSS, in part or in whole, to 21 address the then known shortcomings. In light 22 of the continued evidence of poor product 23 quality these recommendations should be 24 reconsidered." 25 Then you'll remember I took you on the last 6 1 occasion to WITN04600104, that's the schedule of 2 corrective actions. We're now in May 2000 and, 3 if we look at page 9, it addresses the 4 recommendation to redesign and rewrite the EPOSS 5 system. 6 If we scroll over the page to page 10, 7 you'll recall that entry on 10 May 2000 which 8 says: 9 "As discussed this should be closed. 10 Effectively as a management team we have 11 accepted the ongoing cost of maintenance rather 12 than the cost of a rewrite. Rewrites of the 13 product will only be considered if we need to 14 reopen the code to introduce significant changes 15 in functionality. We will continue to monitor 16 the code quality based on product defects as we 17 progress through the final passes of testing and 18 the introduction of the modified CI4 codeset 19 into live usage in the network. PJ, can we make 20 sure that it is specifically covered in our 21 reviews of the B&TC cycles?" 22 That recommendation was closed. Thank you 23 that can come down. 24 A. Yeah. 25 Q. So just to recap the developments over that 7 1 period, we have the PinICL Taskforce document, 2 of which you were the author in 1998; the CSR 3 development audit of which you were the author 4 in 1999; and the Schedule of Corrective Actions, 5 where you are involved in that process, and 6 I think you were the author of that document in 7 May 2000; is that right? 8 A. Yes, that's correct. 9 Q. Now, we're looking today at the Cleveleys case. 10 In February 2000, so in between the CSR+ audit 11 and the closing of that recommendation, Horizon 12 was installed in the Post Office of Mrs Julie 13 Wolstenholme and her contract was terminated in 14 December of that year, still in 2000. 15 Looking at the documents you've just seen, 16 it's fair, isn't it, to say that you were aware 17 of at least fairly notable issues with at least 18 some important part of the Horizon System that 19 affected things like balancing, in the period 20 leading up to and including Mrs Wolstenholme's 21 brief period as a subpostmistress. 22 A. Well, yeah, I mean, the dates would suggest that 23 but the linkage between the two is not something 24 that I would have made. 25 Also, just to say that the number of defects 8 1 that I'm reporting in the CSR audit report are 2 simply numbers. They're not an analysis of what 3 those defects were. So I can't say what areas 4 of EPOSS were actually affected by those bugs. 5 Q. But you have been told by, for example, 6 Mr McDonnell that the code is of poor quality? 7 A. Oh, yeah, yeah, yeah. 8 Q. I think there was also reference to code decay 9 and things like that with PinICL fixes? 10 A. Well, in that context, what I was meaning was 11 that the more you tinker with something, the 12 more likely you are to introduce a problem, and 13 that's the same with anything, whether it's 14 software, a car, decorating. You know, if you 15 fiddle about with stuff, it's likely to start 16 going wrong, you know, which is the "do it right 17 first time" principle. 18 So yeah, that was a concern but that was, if 19 you like, trying to raise a risk in people's 20 minds that this could happen if we carry on -- 21 carried on working in that manner. 22 Q. Absolutely. So you were aware that there were 23 imperfections in the system and that there were 24 attempts to fix those imperfections but attempts 25 to fix the imperfections themselves could cause 9 1 other problems? 2 A. Well, yes, that essentially is what I meant by 3 the "code decay" comment. 4 Q. Yes. Can we now look at FUJ00059075, please? 5 A. Can I just say something here? Again, I don't 6 know whether I've got the timings and the 7 baselines in my head correctly but CSR+ was 8 a new revised baseline and, at that time, 9 whatever was running at the outlets would have 10 been a different baseline anyway. So whether 11 they had the same defects and issues in them, 12 I don't know. I wouldn't know. 13 Q. You wouldn't know because you didn't make 14 enquiries about the ongoing impact of changes to 15 Horizon? 16 A. No, because in conducting the audit, I was 17 looking at the way the work was being done, 18 right, and I was using numbers to demonstrate 19 where there would seem to be still issues but 20 I can't equate CSR+ development audit back to 21 whatever baseline was in place at Cleveleys at 22 that time. 23 Q. So comments in late 1999 about "the figures 24 indicate that problems facing EPOSS during the 25 taskforce period have not diminished concerns 10 1 about non-EPOSS PinICLs", I mean, can they not 2 be broadly applied, that they were concerns you 3 had relating to the EPOSS product? 4 A. Well, the EPOSS taskforce was looking at the 5 baseline that was in place at the time, because 6 we were looking at live defects there. The CSR+ 7 activity, unless I'm vastly mistaken, the 8 development, the PinICLs were test PinICLs that 9 were produced during testing cycles. 10 Unless I've misunderstood my own report, you 11 know, I don't know if those PinICLs that I'm 12 counting there were live or whether they were 13 the results of test. 14 Q. So after 1998/1999 into 2000, you weren't 15 concerned about the Horizon System? 16 A. Well, only insofar as I'd identified all of 17 these PinICLs and defects that were still 18 arising a year after the Taskforce had tried 19 to -- not put a stop to it, that would be 20 virtually impossible, but to try to reduce the 21 number to, you know, a manageable figure. 22 Q. Well, let's move on and look at some specific 23 instances that then crop up in 2000 and 24 thereafter. On the screen at the moment we have 25 FUJ00059075, and that is a KEL, a Known Error 11 1 Log. This is November 2000, or is raised by 2 Anne Chambers in November 2000, so that is the 3 time of a particular incident in 4 Mrs Wolstenholme's case. 5 If we look at -- that is described as 6 follows, it says "Critical event on counter", 7 and then it gives the message, and it says, 8 "Sometimes a storm of these events occurs", and 9 refers to another Known Error Log. If we look 10 at the bottom, the final sentence under 11 "Problem", it refers to another KEL and it says: 12 "[It] is with development for problems 13 balancing while these events are occurring." 14 I'm going to take you to a couple more. If 15 we look at FUJ00059141, this is the same Known 16 Error Log but it says it's version 2. If we 17 look at the problem there, it says: 18 "Possibly caused by an outstanding lock on 19 the run table, which is cleared subsequently 20 when ClearDesk runs. The problem is known to 21 have affected at least 129 counters." 22 If we now look at FUJ00086680. This is 23 a PEAK. It's one that we've looked at 24 previously in this Inquiry. It's a "Master Call 25 for Phantom Transactions". It begins on 12 1 17 April 2001, and I'll just read you a few 2 entries within that PEAK. If we look about 3 halfway down, 14 April. It says, near the 4 bottom of this page: 5 "The system seems to lose transactions and 6 PM is concerned that for every transaction that 7 error he notices there is the probability that 8 there are ones he misses, leading to 9 discrepancies. The PM is at present finding the 10 whole scenario very stressful and is suffering 11 sleepless nights due to these problems. In the 12 light of what has gone on the PM is prepared to 13 break his contractual obligations with POCL and 14 refuse to pay any more discrepancies and will 15 take legal action if required." 16 If we scroll down that page, it says: 17 "As I was on the phone to the PM, he advised 18 that three First Class stamps that were on the 19 screen just 'dropped off'. PM had 3 First Class 20 stamps, and other stamps for 30p. When the 21 other stamps 30p went on, the First Class stamps 22 disappeared. They have since put the 3 First 23 Class stamps again. The first transaction (that 24 disappeared) was put on as 2 First Class 25 stamps", et cetera. 13 1 If we go to page 4, about halfway down, it 2 says, 1 May 2001: 3 "PM feels that the system is unreliable. PM 4 cannot trust this system." 5 If we go over the next page, 2 May: 6 "PM called in because the screen has crashed 7 during his balance -- he has got the blue stop 8 screen error." 9 If we go down towards the bottom of the 10 page, 4 May, it says: 11 "Ki Barnes has called in. I am unsure what 12 to do with this call now. Romec have been to 13 site and state that they have actually seen the 14 phantom transactions, so it is not just the PM's 15 word now. They have fitted suppressors to the 16 kit but the PM is still having problems. As yet 17 there has been no recurrence to the phantom 18 transactions but there still may be problems", 19 et cetera. 20 Thank you, that can come down. 21 Would you accept that post-rollout there 22 were issues affecting balancing or transactions 23 or counters of some subpostmasters in these 24 kinds of logs and logs that you would have had 25 access to? 14 1 A. Well, based on what I've seen just there, then 2 yes. That's clear, isn't it? 3 Q. Yes, combined, of course, with earlier concerns 4 about the EPOSS system. 5 I'm going to now take you to WITN04600202. 6 We're now on 21 August 2003, and we are looking 7 at the Cleveleys case. You, by that stage, had 8 been contacted by Jim Cruise about the Cleveleys 9 case, and -- 10 A. Yeah. 11 Q. -- your answer to Mr Cruise is as follows in 12 relation to what you call the easy stuff. You 13 say: 14 "1. We will have no record of any 15 transaction data from Cleveleys dated before 16 November 2000 in the central audit archive since 17 this was automatically deleted 18 months from 18 the date that it is written. So, if 30 November 19 2000 was the last active day for the Counter 20 that data would have been deleted on or about 21 30 May 2002." 22 Then you say, second: 23 "Similarly, there will be no Helpdesk logs 24 since these are also deleted after 18 months." 25 Now, that number 2, that was wrong, wasn't 15 1 it? 2 A. Yeah, I mean, what I'm referring to there is the 3 formal audit records that are maintained that 4 were then stored on the audit system and deleted 5 after 18 months. And, as you rightly say, 6 subsequently, an -- what I referred to and what 7 Jason Coyne referred to as an unregulated source 8 of data was found and located. But they were 9 not the formal audit data that would have been 10 submitted, had it been requested to a court 11 case. Personally, I wouldn't necessarily place 12 reliance on it, since it wasn't formally 13 captured in the audit record. 14 Q. So you wouldn't place reliance on the Helpdesk 15 logs that were, in fact, ultimately provided to 16 the expert, Mr Coyne? 17 A. Well, from -- I think from a legal perspective, 18 we -- I wouldn't know where they came from. You 19 know, they weren't captured at the point of 20 origin, at the time to origin, and they weren't 21 stored securely so they probably were the same 22 as what was captured on the audit record. 23 Q. Did you say weren't or were? 24 A. No, they probably were the same, but I can't 25 vouch for that. 16 1 Q. If number 2 was wrong, could you be sure that 2 number 1 also wasn't wrong? 3 A. Also wasn't wrong? Um, no number 1 was correct 4 because the transaction data would have been 5 deleted 18 months after it was written. So it 6 wouldn't be there. And 2 -- as I said, 2 refers 7 to the Helpdesk logs that were captured as part 8 of the audit archive. They would have been 9 deleted, as well, after 18 months. So -- 10 Q. But 2 was found? 11 A. No, no, no. 12 Q. Or an unregulated version of 2 was found? 13 A. An unregulated set of records that were the 14 Helpdesk logs were located after the event, 15 after they were deleted from the audit archive. 16 Q. Was an attempt made to look for the equivalent 17 for number 1, so some sort of unregulated record 18 of the transaction data? 19 A. Not to my knowledge, no. 20 Q. Might they have been available? 21 A. Not -- I-- no, I don't think they would have 22 been, based on my knowledge of the system and 23 how it worked, no, I don't think they would have 24 been. 25 Q. Why is it that you were able to find unregulated 17 1 Helpdesk logs but no unregulated transaction 2 data was produced? Was it looked for? 3 A. No, they emanate from different sources. So 4 I can only assume that the Helpdesk itself was 5 maintaining records but they weren't passed to 6 the audit archive system, and that was the part 7 I was commenting on there. They weren't 8 available in the audit archive. And, as I say, 9 subsequently, Helpdesk records were found, but 10 they were nothing to do with the audit archive. 11 Q. Can we look at POL00095375, please. We're now 12 on 5 February 2004 so quite some time after that 13 initial contact in August 2003. 14 A. Yeah. 15 Q. This is a letter to Colin Lenton-Smith from 16 Keith Baines at the Post Office. Did you see 17 this letter at the time? 18 A. Possibly. 19 Q. Had you had any contact with the Post Office 20 subsequent to that 21 August contact, in respect 21 of this case? 22 A. Well, I don't know. I mean, there are some 23 documents that have got dates and the points of 24 my involvement with the various people but 25 I know there was a delay, or quite a delay 18 1 between when all these problems were supposed to 2 have occurred -- or when they occurred in 2000, 3 and us getting involved through Jim Cruise's 4 request for help. And it was linked, I think, 5 to the production of Jason Coyne's expert 6 report. 7 Q. So if we look at the third paragraph that's 8 currently on our screen, it says: 9 "The County Court instructed the parties 10 jointly to commission a report from an expert 11 approved by the Court." 12 Pausing there. Did you understand that the 13 expert was jointly instructed? 14 A. Well, only insofar as it's in Keith Baines' 15 letter. 16 Q. Did you understand the implications of that? 17 Did you know what a jointly instructed expert 18 was? 19 A. No, because I just assumed, rather, that we 20 would be involved in that decision who to 21 appoint. But it would appear not. 22 Q. It says there the expert was approved by the 23 Court. 24 A. Yeah. But the parties being presumably 25 Mrs Wolstenholme and Post Office. We were not 19 1 part of any kind of appointment there. 2 Q. So did you not take it to have any real 3 significance that the expert had been jointly 4 commissioned? 5 A. No, I didn't. 6 Q. If we scroll down on that page, we'll see there: 7 "I'm sure you'll understand, Post Office is 8 concerned by these findings, not only in 9 relation to this particular case, but also 10 because of any precedent that this may set and 11 that may be used by the Post Office's agents to 12 support claims that the Horizon System is 13 causing errors in their branch accounts." 14 Was this the first occasion where you 15 realised the potential implications of the 16 Cleveleys case? 17 A. No. I mean, I think we got -- how can I explain 18 this? From Horizon's point of view, from my 19 point of view, helping the Post Office resolve 20 this was really just another piece of work 21 coming down the line. I didn't have any 22 personal buy-in to it. It was just a request 23 for help to provide information. 24 Q. So -- 25 A. Essentially, it was -- it started off by seeking 20 1 any help to counter some of the opinions in the 2 expert's report. 3 Q. So you didn't take it in any way personally? 4 A. No, no. It was work. 5 Q. You didn't feel any kind of vested effort to win 6 in this case? 7 A. No, I didn't have any skin in it at all, so why 8 would I? It was just a piece of work that we 9 were asked to do by the Post Office. We didn't 10 have any real background, other than the fact 11 that the postmaster was in dispute and had 12 confiscated the equipment and wouldn't return 13 it. But, other than that, it was just 14 responding to the request made to us by the Post 15 Office. 16 Q. Do you feel you maintained that neutral approach 17 throughout the conduct of the case? 18 A. That's an interesting one because, clearly, we 19 didn't much like some of the comments and the 20 attitude of the expert's report but that's his 21 report. So all we could do was try to counter 22 some of the claims with what we felt were 23 reasonable counterarguments. 24 Q. Is your evidence that you maintained your 25 professional approach throughout? 21 1 A. Yes, I think I did. 2 Q. Let's look at the actual witness -- the expert's 3 report, it's WITN00210101. Now, we've been 4 through the report, it's over the page. We've 5 been through the various opinions with other 6 witnesses. I think you've been provided with 7 this as part of your preparation, so I'm not 8 going to go into detail at this stage on the 9 report itself. I just want to take you to 10 page 5 of the report, and that is the CV of the 11 expert that appeared as part of the report. 12 I'll just read to you the first few paragraphs. 13 He says there: 14 "I confirm that I have made clear in my 15 report those facts that are within my own 16 knowledge and which I believe to be true, and 17 that opinions I have expressed represent my true 18 and complete professional opinion. 19 "I have no known connection with any of the 20 parties, witnesses or advisers involved in this 21 case. 22 "Under the requirements of the Civil 23 Procedure Rules 1999, as amended in January 2002 24 I confirm that I fully understand my duty to the 25 court and I have complied and will continue to 22 1 comply with that duty." 2 Were you aware at that time that an expert 3 owes duties to the court? 4 A. No, I hadn't seen that document at that time. 5 Q. Irrespective of -- well, this document is in 6 fact part of the opinion that you were provided 7 with. If we scroll back you'll see that this is 8 just part of the opinion. It's, effectively, 9 the back page of the opinion. But, irrespective 10 of the opinion itself, you were somebody who was 11 providing statements in court proceedings. Were 12 you aware that an expert owes duties to the 13 court? 14 A. No, no. 15 Q. We see there, near the bottom it says, "Law 16 Society 2003 Accredited Expert Witness", and it 17 gives his accredited expert witness number. 18 Can we now look at WITN04600302. So the 19 expert report is dated 21 January -- or 20 20 January, provided thereafter. By 11 February 21 2002, you are carrying out some investigations 22 in order to address what's written in that 23 report; is that correct? 24 A. Yeah, that's correct, yeah. 25 Q. So we have correspondence here between somebody 23 1 called Mark Jones -- who was Mark Jones? 2 A. I'm not sure I can remember. I think he was 3 probably somebody who worked in customer 4 services, but -- 5 Q. If we scroll down -- 6 A. -- I can't remember. 7 Q. -- we can see Mark Jones, MSU team. Who was the 8 MSU team? 9 A. I don't know. Management Support Unit, 10 possibly. 11 Q. Let's have a look at that email. It says: 12 "Richard asked me to provide some call 13 analysis from 2000. All the FADs here are 6 14 counter outlets -- Cleveleys comes out top (or 15 bottom) depending on your view." 16 Now, we see there that there are six 17 outlets. You have Cleveleys. SD02 is software 18 error, so that's the code that's been attributed 19 by the Helpdesk for software errors and, on the 20 analysis from Mr Jones that's come back soon 21 after your receipt of that report, it seems as 22 though Cleveleys has quite a high number, in 23 fact the highest -- comes out at the top or at 24 the bottom, depending on your viewpoint -- of 25 those six counter outlets. 24 1 I mean, this analysis is very much the point 2 that Mr Coyne was making in his report, isn't 3 it? 4 A. Well, I guess so, yes. 5 Q. So we're at February 2004. You knew at this 6 stage, as we've been over the troubled history 7 of the EPOSS product, you were aware of 8 complaints by subpostmasters or you had access 9 to those KELs and call logs that we have looked 10 over and now you've received from Mr Jones of 11 the Management Support Unit team figures that 12 are entirely consistent with the report of Jason 13 Coyne. 14 Wasn't this, February 2004, the point to say 15 there might actually be something in the 16 subpostmistress's complaint? 17 A. Yes, but remember we were just responding to the 18 Post Office request for help. It's not our 19 place. It was not our place, at least I wasn't 20 aware it was my place, to start challenging on 21 behalf of, you know, the postmaster or the 22 postmistress, this information. 23 Q. Because, as you've said, you were entirely 24 professional and neutral throughout? 25 A. Indeed, yes. Trying to be. 25 1 Q. I mean, you have the independent expert's report 2 and now you see that Cleveleys is very much 3 having problems with software errors. It was 4 only one branch. It had obviously made calls 5 about software errors. You didn't have the data 6 because it had been deleted. How could you be 7 confident that there was not a software issue in 8 the branch? 9 A. Well, I couldn't, could I? 10 Q. Can we please look at WITN04600203. This is 11 a document that you produced. I think we've 12 seen a few of these kinds of documents. Are 13 they documents that you kept yourself, notes of 14 what was happening at the time? 15 A. Yeah, this was an internal -- a document for 16 internal consumption, just as an initial 17 response to that report because, clearly, Post 18 Office wanted us to make some kind of formal 19 response so these were, if you like, my first 20 drop down thoughts of what was going on. 21 Q. If we scroll down, so you note there "Today, 22 (12 February)", so as we to take it that these 23 notes were written on 12 February? 24 A. That's what that would imply, yes. 25 Q. I think, if we look at the top, it says the 26 1 report was, in fact, received by you on 2 6 February? 3 A. Yeah. 4 Q. Can we please have a look over the page. You've 5 said: 6 "On P5. I have asked for an analysis of HSH 7 calls from 1 January 2000 to 20 November 2000 8 for Cleveleys and 20 other 6 Counter Outlets 9 chosen at random with full error code analysis. 10 I should have this today/tomorrow ..." 11 It seems by that stage, having received that 12 information from Mark Jones, which showed 13 Cleveleys at the top or the bottom of the list, 14 you then commissioned further analysis of 20 15 other six-counter outlets or 20 six-counter 16 outlets. 17 I'd like to look at that final paragraph as 18 well. You say at the end of that: 19 "Blue screens and system freezes have always 20 been a problem and the stock HSH reply has 21 always been to reboot." 22 Did you acknowledge at that time in February 23 that blue screens and system freezes had always 24 been a problem? 25 A. Well, isn't that what that sentence says? 27 1 Q. Absolutely. So are those your words and your 2 acknowledgement there, at that time, that blue 3 screens and system freezes had always been 4 a problem and the stock, as in the standard, the 5 automatic, HSH reply had always been to reboot? 6 A. That was my understanding at the time, yes. 7 Q. Can we go to FUJ00121485. This is an early view 8 that you have provided to Colin Lenton-Smith on 9 18 February 2004. We can look at the 10 attachment, that's FUJ00121486. So this is your 11 early view response. We've seen this document 12 already but if we scroll down to the second half 13 of the page, it says: 14 "On 6 February POA [that's Fujitsu's Post 15 Office Account] received a copy of the Expert's 16 report with a request from POL for an early 17 response. POL are concerned that the Expert's 18 opinion (that the system was at fault) might set 19 a precedent against future POL prosecutions." 20 So that's essentially that letter that we 21 saw from the Post Office; is that right? 22 A. Yeah, that was our understanding at the time. 23 Or my understanding at the time. 24 Q. If we scroll down to "'Reasonableness' of 25 calls", it says there: 28 1 "[Post Office Account] are able to review 2 an unregulated archive of records of the other 3 installed 6 Counter Outlets over a comparable 4 period. At the table below shows the output 5 from that analysis." 6 Now, sorry, could we just stick there for 7 one moment on the previous page. 8 So the unregulated archive are records of 9 the other installed six-counter outlets over 10 a comparable period. Do you remember what that 11 exactly meant? Because that's quite -- it seems 12 to be quite specifically worded, "the other 13 installed 6 counter outlets". 14 A. Yeah, now on an earlier document, I said 15 I wanted an analysis of 20 six-counter outlets 16 selected at random. So when this information 17 came back, if you scroll down, you will see that 18 there aren't 20 six-counter outlets in existence 19 at that time. 20 Q. Absolutely. Well, there aren't 20 in that 21 table. 22 A. No, but those are the six-counter outlets that 23 were identified, I believe. So there weren't 24 any dropped off the end. They were the 25 six-counter outlets in existence at that point 29 1 in time. 2 Q. Can we put that side by side with WITN04600302, 3 please. Thank you very much. So that's the 4 email you received from Mark Jones. 5 A. Yes. 6 Q. Let's look at the left-hand side. He is saying 7 all the FADs here are six-counter outlets? 8 A. Yeah. 9 Q. Worlds End, where's Worlds End on the right-hand 10 side? 11 A. No, I can't see it. 12 Q. Shoreham-by-Sea? 13 A. No, can't see it. 14 Q. Cookstown? 15 A. Can't see it. 16 Q. Halstead is there, in fact, on the right-hand 17 side, so there is one that matches, although, on 18 the left-hand side we have three SD02 codes but 19 that has increased to seven on the right-hand 20 side. Staines? 21 A. No, I can't explain the discrepancy. 22 Q. I mean, those previous six counter outlets that 23 had been identified seem to largely have been 24 excluded from this table, the ones that had low 25 numbers for software errors? 30 1 A. So it appears. 2 Q. Do you know why those original figures were 3 excluded from this table? 4 A. No, I've no idea. I mean, this second table on 5 the right was also provided to me. So, you 6 know, I'm just being fed information to try to 7 put it together to form some kind of coherent 8 response. To be fair, I didn't notice the 9 disparity between the two lists. Well -- 10 Q. Is it possible that some were removed that 11 didn't suit your case? 12 A. Not by me. 13 Q. We can take down the left-hand side now. Thank 14 you very much. In this table, the one that has 15 been provided that doesn't provide those other 16 counters with the lower figures, in that table 17 we still see Cleveleys with the second highest 18 figure for "S", software issues. 19 A. Yeah. 20 Q. Again, I don't want to repeat all the things 21 that I've already said about the history, the 22 logs that were available, the table that you 23 originally received, your knowledge of things 24 like blue screen freezes, blue screen system 25 freezes. Looking at this, software number 2, 31 1 the second worst in terms of numbers. Did this 2 not give you pause for thought at this stage? 3 A. No, because that wasn't what I was doing here. 4 I wasn't, sort of, doing some kind of 5 substantive test of the numbers and of errors. 6 I was merely collating information to respond to 7 the report. So whether that's an oversight on 8 my part, okay, but I wasn't there analysing the 9 figures themselves to understand what was going 10 on beneath them. I was merely collating them to 11 make a response. 12 Q. Cleveleys seems to have distinguished between 13 rollout and post-rollout, that's the difference 14 between 1 and 2. Was there the same analysis 15 for those other branches or was it just 16 Cleveleys that the two differences were 17 identified? 18 A. No, it looks just like Cleveleys. 19 Q. Even if we take the second, the lower figure, 20 because it doesn't include the Horizon rollout, 21 the subpostmistress still had cause to call the 22 Helpdesk 85 times in that period. 23 A. Mm-hm. 24 Q. It's not beyond fantasy that Mrs Wolstenholme 25 may have been experiencing software problems 32 1 with Horizon, is it? 2 A. Well, no, those figures suggest that that was 3 the case. 4 Q. If we scroll down, there's the comment about 5 Ms Elaine Tagg's statement. I wonder, actually, 6 if we could bring that on to screen. The 7 extract from Ms Tagg's statement in Mr Coyne's 8 report. That's WITN00210101. If that could be 9 bought alongside, that would be fantastic. So 10 it's the second, over the page on the left-hand 11 side, and it is that second -- the second half 12 of that page. Thank you. So on the right-hand 13 side, what you've said about the statement of 14 Ms Elaine Tagg is: 15 "Based on the analysis, and without 16 analysing each and every call record it would be 17 hard to dispute the opinion of the Expert." 18 On the left-hand side, it shows what the 19 expert was saying. He identified that Ms Elaine 20 Tagg had said that: 21 "'Mrs Wolstenholme persisted in telephoning 22 the Horizon System Helpdesk in relation to any 23 problems which she had with the system 24 generally, these problems related to the use and 25 general operation of the system and were not 33 1 technical problems ...'" 2 Then Mr Coyne's opinion was: 3 "This, in my opinion is not a true 4 representation on the evidence that have had 5 access to." 6 It seems as though you agreed with Mr Coyne 7 in that respect, at least at that stage. 8 A. Well, insofar as I said based on the analysis, 9 either table, but without analysing each and 10 every call record, you know, it's hard to 11 dispute. It is. 12 Q. Because I think you're saying on the right-hand 13 side that, even if you take out the 15 that are 14 classified as advice and guidance and the 16 15 that are to do with the rollout, you still have 16 70 calls. So Mrs Tagg's statement in that 17 respect can't possibly be right in terms of them 18 not being technical problems; is that right? 19 A. Yeah, well, I guess so. I mean clearly there 20 were some problems there that Mrs Wolstenholme 21 was experiencing. 22 Q. There is at least some truth in Mr Coyne's 23 opinion in that respect? 24 A. Yes, I believe that's the case, yeah. 25 Q. Can we stay with the document on the right-hand 34 1 side, please, and enlarge -- thank you. If we 2 could scroll over to the next page, page 3. We 3 then get to the "Operator advice to 'Reboot'", 4 and the second paragraph there, you say: 5 "In this context the opinion of the Expert, 6 that 'this instruction treats the effect and not 7 the cause' is correct." 8 Again, it seems, at least in some respects 9 there, you are agreeing with Mr Coyne? 10 A. Well, yes, he is right. I mean, rebooting the 11 system might be the instruction that's given to 12 the postmaster and it does deal with the effect 13 and not the cause. So there should be further 14 work undertaken to try to understand what the 15 cause was, determine the root cause and find 16 a way of removing it. 17 Q. In this document, so in this initial response, 18 one thing you don't go as far as to say is your 19 admission that blue screens and freezes have 20 always been a problem. I don't think that's 21 contained in this document, is it? 22 A. Possibly not, no. 23 Q. If we scroll down to the final paragraph in this 24 document, about discrepancies, it says: 25 "This argument has been put forward by 35 1 a number of PMs in the past when challenged and 2 prosecuted by POL for alleged fraudulent 3 behaviour and each time it has fallen when 4 confronted by transaction data that demonstrates 5 that the system was operating normally during 6 the disputed time period." 7 Was that the general view at Fujitsu? Is 8 that a view that was held by your colleagues: 9 that because it hadn't been proved in the past, 10 that in itself demonstrates that there isn't 11 a problem? 12 A. Yeah, that's fair enough. You could say that. 13 Q. Can we please now look at WITN04600304. We are 14 now on 17 February 2004, and this is an email 15 from Steve Parker. Can you help us with who 16 Steve Parker was? 17 A. No, I was looking at this myself the other day 18 and -- can you just scroll down to what was 19 presumably my request for help? 20 Q. Absolutely. 21 A. Right, well, that's not very helpful, is it? 22 Yeah. So, obviously, I sent him the expert 23 report and the call volume comparison. 24 Q. If we scroll up -- 25 A. Yeah, you can -- 36 1 Q. -- the words in -- I'm slightly colourblind, 2 I think I'd say purple -- 3 A. Blue. 4 Q. -- blue -- are from Steve Parker; is that right? 5 A. Yes, that's correct, yeah. 6 Q. He said: 7 "I have had a chance to read through this 8 now. Here are a few thoughts you might like to 9 develop." 10 It's the "Worrying discrepancies" section 11 that I'd like to look at. It says: 12 "Must be the major issue. Counter systems 13 cause discrepancies." 14 I think he's stating that that is the 15 argument that is put forward: 16 "Answer has to be 'no way': 17 "a) Almost all accounting errors in computer 18 systems are caused by user error. GIGO 19 principle." 20 Did you understand that to be "garbage in 21 garbage out" principle? 22 A. Yes. Yes, I did, yeah. 23 Q. Then it says systems are in place to resolve 24 discrepancies, and then we have: 25 "Yes, software errors can make such 37 1 mistakes. However, the systems in place ensure 2 that such errors are picked up and resolved. If 3 this problem was caused by a software error, the 4 same error would exist on all Horizon sites. 5 17,000 [postmasters] are not complaining of 6 misbalancing and discrepancies." 7 I mean, that is nonsense, isn't it? 8 A. I -- at the time, I thought that was rather glib 9 and I don't think -- I don't think any of this 10 was used in further correspondence between me 11 and the Post Office. 12 Q. I mean, it seems -- the impression that's given 13 is that it's getting a little desperate to look 14 for some sort of counterargument to what 15 Mr Coyne is saying; would you accept that? 16 A. Yeah, I mean, if I can remember who Steve Parker 17 was and what role he played, what position he 18 had, I might be able to put this more into 19 context as to why that reply came back looking 20 like it did. But I remember at the time, it 21 wasn't -- I wasn't overly impressed with it and, 22 you know, the comment at the end that, you know, 23 it would exist on all Horizon sites 17,000, PMs 24 are not complaining, I mean, that's -- well, I'm 25 sorry, that's just silly. 38 1 Q. Can we look at WITN04600305. This is a response 2 from Richard Brunskill to yourself. So his 3 response is in colour. Your original email is 4 in black. I'm going to start with your words. 5 You say there: 6 "The Cleveleys situation has us a bit on the 7 back foot and I'm having to bring in POL imposed 8 SLAs as a form of justification for HSH advice 9 being to reboot 9 times out of 10." 10 Back foot: I mean, that sounds a bit like 11 a competition, like you're getting 12 out-manoeuvred. Was that your concern at the 13 time? 14 A. No, out-manoeuvre is wrong. I mean we had to go 15 on the defensive and clearly HSH advice being to 16 reboot nine times out of ten, on the face of it, 17 looks unacceptable, so we have to understand why 18 that is. And the argument was used that the 19 purpose of the HSH is to get the postmasters up 20 and running as quickly as possible and not 21 interfere and not interrupt their business 22 cycle. 23 So sometimes the easy answer was to say 24 reboot and then you're up and running again. 25 The question that was raised by Mr Coyne was 39 1 what happened to that after they rebooted? Did 2 they do anything with it? Well, yes, they did 3 because the presence of KELs where these things 4 have been further investigated, suggests that 5 work was done after the event to try to better 6 understand what was going on. So it wasn't just 7 the case of "Yeah, you've got a problem, reboot; 8 next one, reboot; next one". It wasn't quite 9 that straightforward and simple. 10 Q. That's not actually the question I'm asking you. 11 I'm asking you about that first sentence and the 12 way that it's worded. You're saying you're on 13 the back foot and then you're saying, "I'm 14 having to bring POL imposed SLAs a form of 15 justification". Now, it might be suggested that 16 having to bring in the SLAs, it means you didn't 17 otherwise have a good answer to why the advice 18 was to reboot nine out of ten times. 19 A. Yeah, well, that's a fair comment. 20 Q. I mean, thinking back to your evidence earlier 21 and about how you were professionally and coolly 22 responding, only providing information, do you 23 think by this stage you've lost a little bit of 24 objectivity? 25 A. I don't think so. I mean, you know, I'm trying 40 1 to -- I say, justify. I mean if the advice is 2 being given by the HSH to do that, they must be 3 doing it that for a reason, and one of those 4 reasons is SLAs that are imposed by POL to get 5 the postmasters up and running a bit more 6 quickly. Having us on the bit of a back foot is 7 just a phrase, isn't it? I mean, we're trying 8 to find ways of -- or I'm trying to find ways of 9 justifying the advice but not necessarily just 10 limiting it to that advice. You don't just do 11 that and walk away. 12 Q. If we look at what Mr Brunskill has said, in 13 response to question 1, your question was: " 14 "Time to fault resolution from original call 15 and how was fault resolution defined -- getting 16 the Outlet back up or fully resolving the 17 problem?" 18 So your question about the SLAs, was the SLA 19 about getting the outlet up and running or was 20 it about actually resolving the problem? We 21 have there -- 22 A. (Unclear). 23 Q. -- Mr Brunskill's response. He says: 24 "Depends on the issue. Simple advice and 25 guidance at the time ... centred around 41 1 resolution in 5, 10, 30 or 45 minutes depending 2 on the query. Blue screens or system freezes 3 would generally have been coded as 'Software' 4 and claimed under break fix SLTs, which, 5 depending on which branches location, fall into 6 the local, intermediate or remote category with 7 anything from 4 to 24 hours to fix. I would 8 expect a software call to be resolved by either 9 a reboot, workaround or kit swap. The 10 underlying issue is not subject to SLA -- this 11 can be fixed as per release timetable, etc, the 12 issues under the SLA is 'get the branch up and 13 running again'. Hence the proliferation of 14 reboots to solve known problems and get things 15 going." 16 So what he's saying there is that the SLA is 17 not about fixing the problem. It's about 18 getting the branches up and running because it's 19 about timing and speed of resolution. Do you 20 accept that? 21 A. It's to do with timing to get the branch up and 22 running again. The problem resolution is 23 subject to other activities and other schedules. 24 Q. What he's saying there is that you can't rely on 25 the SLA about resolution because the SLA is 42 1 about getting up and running again, essentially 2 as quickly as possible. 3 A. That's correct, yeah. 4 Q. He says: 5 "Hence the proliferation of reboots to solve 6 known problems ..." 7 Proliferation, it's not a positive word. 8 I mean, it's often used in the context of 9 nuclear weapons, something you don't want to 10 happen. Is your reading here that he is not 11 being overly positive about the use of reboots 12 to fix problems? 13 A. No, I didn't read that into his words. I just 14 read it as a high number. 15 Q. Do you read it now, looking at it, 16 "proliferation of reboots"? I mean -- 17 A. No, I don't. I'm still just seeing as nine out 18 of ten reboots being the solution from the 19 Horizon System Helpdesk. I didn't see anything 20 deeper in it than that. 21 Q. Can we look at WITN04600306. This is another 22 response, here from David Cooke. 23 A. Yeah. 24 Q. Can I just ask you to have a read of that and 25 what I want to understand is that second 43 1 paragraph. He says: 2 "... any reboots that took place during 3 a session would have wiped all EPOSS 4 transactions." 5 Am I to read into this that there is 6 a suggestion that reboots could wipe EPOSS 7 transactions or not? 8 A. Well, that's what the words say, and David and 9 James were both very knowledgeable members of 10 the -- technical members of the team. So if -- 11 this is Dave Cooke, isn't it? 12 Q. Yes. 13 A. So if Dave says any reboots that took place 14 during the session would have wiped all EPOSS 15 transactions, I would take that as a truth. 16 Q. Is that potentially problematic? 17 A. I don't know because I don't know all the 18 technical details that go behind the reboot and 19 wiping old transactions and re-establishing 20 transactions once they've been rebooted. As 21 I said in my first -- in November, I'm not 22 a technical person. So you're getting into 23 levels of detail that I really can't answer. 24 Q. Well, I mean, these are emails to you from 25 people providing you with information in which 44 1 to respond to an expert -- a technical expert's 2 report. At the time, did you feel you didn't 3 have the technical expertise to respond to that 4 expert report? 5 A. No, because I did reply to it and -- 6 Q. Did you say to anybody "I just don't have the 7 expertise to be able to address these matters"? 8 A. No, I can reflect what people tell me. 9 Q. I mean, we're in the context of litigation here, 10 court proceedings, quite a serious matter, where 11 witness statements are being provided and 12 an expert has been jointly instructed, 13 a response is being provided to that expert in 14 the context of court proceedings. Do you feel 15 you didn't have the expertise to be doing the 16 job you were doing in that respect? 17 A. Well, I felt comfortable at the time but, now 18 you're challenging it, I'm not so sure I was. 19 Q. Can we look at FUJ00121512, please. We're here 20 now on 20 February 2004. This is the response 21 to Keith Baines from Colin Lenton-Smith, 22 attaching the appendix which is the response. 23 A. Yeah. 24 Q. If we turn over the page, please, thank you. 25 "Horizon System Helpdesk". 45 1 Sorry to trouble you again, but could we 2 bring onto screen WITN04600305 alongside this. 3 Thank you very much. 4 The first paragraph under "Horizon System 5 Helpdesk", it says: 6 "It operates under strict contractual 7 Service Level Agreements covering aspects such 8 as pick-up time, first time fix and time to 9 close." 10 Refreshing our memory from what Richard 11 Brunskill said, I mean the essence of what he 12 said was it's not really about the SLA. Do you 13 think there was too much focus, in light of 14 that, on the right-hand side in the response on 15 the SLA? 16 A. I'm not quite sure what you're getting at. The 17 SLA exists and has to be met in our relationship 18 with the Post Office. 19 Q. Your request to Richard Brunskill was about "I'm 20 having to bring in the SLA in order to defend 21 ourselves. I effectively want to say that 22 that's going to -- that that's also about 23 getting the outlet back up and running fully", 24 and Richards response to you is "It's not really 25 about getting it up and running fully; it's 46 1 about the speed at which we get up". It's not 2 about resolving the -- the SLA itself is not 3 about resolving the underlying dispute, 4 technical issue -- 5 A. Is that what you mean by "fully"? So you 6 fully -- if you fully resolve the call, you get 7 them up and running and you solve the problem. 8 In that case, no, the SLA isn't about fully 9 doing that. It's about getting the outlet back 10 up and running. 11 Q. Do you think it was still appropriate, after 12 having received Mr Brunskill's email, to be 13 focusing in your response on the SLA, which was 14 about quick fix? 15 A. In response to the statement of nine out of ten, 16 I can't think of any other way of doing it. You 17 know, if the reboot advice is there and it's 18 happening nine times out of ten or too many 19 times to be acceptable to Mr Coyne, based on his 20 knowledge and experience, we have to explain why 21 that is operating like that, with that frequency 22 of reboots, and the reason it was, was through 23 the SLAs requiring that we get the post offices 24 back up and running. 25 Q. What Mr Brunskill is saying is that the 47 1 underlying issue itself is not subject to the 2 SLA. 3 A. No, it's not. Because that will fall into 4 a different sort of line of activity, which 5 was -- I suppose putting all these problems into 6 a bucket, looking at PinICLs and PEAKs, and then 7 aligning them with releases where the work is 8 done to fix the problem, or putting a workaround 9 in place through a KEL. 10 Q. The final sentence of Mr Brunskill's first 11 paragraph, where he says "Hence the 12 proliferation of reboots to solve known problems 13 and get things going", does that in any way 14 feature in the response? 15 A. What in the response on the right? 16 Q. Yes. So the kind of -- you know, his mention of 17 proliferation of reboots, the mention of known 18 problems. Does the words "known problems" even 19 appear on the right-hand side? 20 A. No, because on the right-hand side we're saying 21 why we do the reboots: 22 "... quick response to their call and, to 23 the extent possible over the phone time, 24 a timely ..." 25 Sorry: 48 1 "... a quick response to their call and, to 2 the extent possible over the phone, a timely 3 return to normal business operations", which is 4 what the SLA was trying to do. 5 Q. If you had mentioned in that response that 6 people were being told to reboot to solve known 7 problems, to get things going, do you think that 8 that would have undermined your case? 9 A. No, because the desire to reboot to get the 10 business up and running still exists, whether 11 it's a known fault or an unknown fault -- new 12 fault. 13 Q. Is there a reference to known faults on the 14 right-hand side? 15 A. I don't know. I don't think so. 16 Q. Can we keep the right-hand side on the screen, 17 please, and turn to page number 4 of that 18 document. We have there, I think for the first 19 time, this paragraph: 20 "It is worth noting that Fujitsu Services is 21 not aware of similar complaints or claims being 22 made from other Outlets in the above list, some 23 of which have higher call profiles than 24 Cleveleys." 25 Now, that's quite a carefully worded 49 1 sentence. You are there referring only, it 2 seems, to the outlets in the above table; is 3 that right? 4 A. Yeah. Yes. 5 Q. Yes, because, of course, Fujitsu couldn't have 6 said that they are not aware of similar 7 complaints or claims being made by outlets in 8 the general, could they? You -- 9 A. (Unclear). 10 Q. You couldn't have broadened it to outlets 11 because that wouldn't have been true, because 12 there were similar complaints or claims being 13 made from other outlets; is that right? 14 A. Presumably, yes. 15 Q. Do you remember wording this quite carefully in 16 that regard? 17 A. No, because I don't think I worded that 18 sentence. This is a letter from Colin 19 Lenton-Smith, isn't it? 20 Q. The letter is sent by Colin Lenton-Smith, it 21 includes what is called a response to expert's 22 opinion. 23 A. Yes. 24 Q. Most of the words have appeared in your initial 25 thoughts document, not all, and we'll come to 50 1 "not all". But is it your evidence that, in 2 fact, this paragraph was inserted by somebody 3 else? 4 A. I honestly can't remember. I just don't know. 5 I know that Colin may well have tweaked some of 6 the content before he sent it off, because it 7 was going off in his name. Whether he put that 8 sentence in or not, I don't know. Whether I put 9 that sentence in, I can't remember. 10 Q. Does it sound like your kind of writing? Does 11 it sound like something that you would have 12 said? 13 A. I can't comment on that. I don't know. 14 Q. If we zoom out on this page, are you able to 15 assist us with where on this response the 16 acceptance of Mr Coyne's criticism of Elaine 17 Tagg's statement features? I mean, we read that 18 in your initial draft. Where is that now? 19 A. I don't know. It's not there. 20 Q. Did you remove it? Did somebody else remove it? 21 A. I can't remember and I don't know. 22 Q. I mean, you'll remember when we looked at it, it 23 was something that you felt that Mr Coyne was 24 right on. Do you remember removing comments 25 that were positive in favour of Mr Coyne? 51 1 A. No, no, I don't think -- no. I wouldn't have 2 done that. 3 Q. Well, who would have? 4 A. I've no idea. I mean, this is 24 years ago. 5 Q. Well, you said -- 6 A. Or -- 7 Q. -- you're sure that you wouldn't have removed it 8 but, equally, you have no recollection because 9 it was 24 years ago. 10 A. I mean, I'm not sure whether I would have 11 removed it or not. I can't remember if I did. 12 Q. Can we look at page 5, please. We have there 13 the "Conclusion": 14 "The report presented by the Expert is based 15 on a simple analysis of HSH records and not 16 a detailed understanding of how the Horizon 17 System works, or even the prime objectives of 18 the Horizon System Helpdesk." 19 The criticism there is that it's a simple 20 analysis. Do you think that Mr Coyne had been 21 provided with enough information at that stage 22 to provide more of an analysis than he did? 23 A. I don't know. You'd have to ask the Post Office 24 that because they provided him with whatever 25 information they did. 52 1 Q. Well, he didn't have transaction data, we know 2 that much, don't we, because that had been 3 destroyed? 4 A. Destroyed. 5 Q. The Helpdesk original records had also, it's 6 your evidence, been destroyed, but another 7 version had been found. 8 A. Indeed, yeah, yeah. 9 Q. But that's all that you recall having been 10 provided by Fujitsu in respect of this case? 11 A. To POL, yeah. 12 Q. Well, to POL, sorry. Yes. 13 A. Yeah, and because we had no direct contact with 14 Jason Coyne, everything went through POL because 15 it was their relationship. 16 Q. We see in later documents reference to inviting 17 Mr Coyne to attend Fujitsu but, at this stage, 18 are you aware of any invitation having been made 19 to Mr Coyne to attend Fujitsu or to obtain 20 further documentation? 21 A. No. Not at this first cycle. 22 Q. We see there it says: 23 "Consequently the opinions expressed in the 24 report, whilst not always incorrect, do not 25 present the whole story and are presented from 53 1 a single perspective." 2 Mr Coyne was a joint expert. The parties 3 had agreed to rely on a joint expert. Your 4 criticism is that he presents from a single 5 perspective. How many experts would have been 6 enough. 7 A. Sorry, I don't understand what you're getting at 8 there. Um -- 9 Q. Presented from -- what did you mean by 10 "presented from a single perspective"? Do you 11 mean he's only one man? 12 A. No, no, no, no, no. He had a view and he put it 13 forward. We didn't always agree with it. 14 Q. Well, "simple analysis", "single perspective". 15 He was a jointly appointed expert selected by 16 agreement between the Post Office and the 17 defendant in that case -- 18 A. Yeah. 19 Q. -- liaising with the Post Office's own 20 solicitors. Do you think that that's a fair 21 criticism? 22 A. I think it's just a statement. It's not 23 a criticism, as such. It's just, you know, it's 24 a statement that that was how we felt at the 25 time about his report and we put together what 54 1 we felt was a fairly well-presented rebuttal 2 and, as we'll see consequently, he just said it 3 didn't make any difference -- 4 Q. As -- 5 A. -- to his opinion. 6 Q. -- at 20 February 2004, is your evidence that 7 you were still maintaining your professional, 8 calm approach to this case? 9 A. Yes. 10 MR BLAKE: Sir, I think that might be an appropriate 11 time to take a 15-minute break. Could we come 12 back at 11.35? 13 SIR WYN WILLIAMS: Certainly. All right. 11.35. 14 MR BLAKE: Thank you very much. 15 (11.18 am) 16 (A short break) 17 (11.35 am) 18 MR BLAKE: Thank you very much, sir. 19 Mr Holmes, I'm just going to take you -- 20 before the break you referred to the expert's 21 response, where his view stayed the same. Let's 22 look at that. That's FUJ00121535. 23 So we have here, this is the response from 24 Mr Coyne to Weightman Vizards and, if we look at 25 the final page -- sorry, the second page even, 55 1 he says things like: 2 "Defective equipment 3 "I am confident in my statement ... 4 "Worrying discrepancies ..." 5 We see the final sentence is there: 6 "In short to answer the question posed in 7 your letter, no my opinion, currently, remains 8 as stated in my original note." 9 We went through this morning your history of 10 the EPOSS problems, 1998, 1999, 2000. We went 11 through some error logs from the contemporaneous 12 period that Mrs Wolstenholme was having her 13 issues, referring to software problems affecting 14 transactions. We saw that first table that was 15 sent to you by Mark Jones in February, showing 16 that Cleveleys was at the top of the list or the 17 bottom of the list, depending on your 18 perspective. We saw Richard Brunskill's 19 comments about the proliferation of reboots. We 20 saw also you own agreement with important parts 21 of Mr Coyne's original report. Now we have 22 a further response from the jointly appointed 23 expert, where he has maintained his view. 24 Was now, 2 March 2004, not the time to say 25 "Enough is enough, we give in, put up the white 56 1 flag"? 2 A. Sorry, is that a question? 3 Q. Yes. It started with the word "would". 4 A. Possibly, yes, but at the time we decided we'd 5 have one more pass around to see if we could 6 provide any kind of extra assistance that might 7 help him to change his mind. Things like, you 8 know, because we couldn't provide crashdumps on 9 failure, of the period in question, we could 10 invite him to Stevenage to see how they actually 11 worked and operated. But, I mean, he'd 12 obviously decided he'd turned his face against 13 the offer. It wasn't just a walk around, as 14 suggested, we were prepared to give him access 15 to systems so he could do further testing and 16 digging around that he wanted to do. So it 17 wasn't just a PR stunt to try to soften him up; 18 it was a serious attempt to offer him access to 19 parts of the system that normally we wouldn't 20 provide access to. 21 Q. Did you do your own vanes at that time of the 22 Known Error Logs, the PEAKs, the PinICLs, 23 looking at the time when Mrs Wolstenholme was 24 experiencing her problems, investigating other 25 branches to see if they were experiencing 57 1 similar problems -- 2 A. No. 3 Q. -- looking at the kinds of KELs that we looked 4 at this morning? 5 A. No. 6 Q. Can we look at FUJ00121549. This is a day after 7 that reply was received. This is from you to 8 Colin Lenton-Smith attaching your initial 9 thoughts on Jason Coyne's reply. 10 A. Mm-hm. 11 Q. Your response is at FUJ00121550. Thank you. 12 The third paragraph there, you say: 13 "I have spoken to Jim Cruise ... and we both 14 feel that there is probably another opportunity 15 to influence Jason Coyne's opinion by inviting 16 him to Bracknell and providing him with access 17 to data, records and people who can deal with 18 his observations directly." 19 Did you really feel at that stage, so far 20 down the line, having everything in mind that 21 I've just been mentioning, did you really think 22 that it was proper to be trying to influence the 23 independent, jointly appointed expert's opinion 24 at that late stage? 25 A. Well, "influence" is a word. I mean, can we 58 1 just put something into context here? At the 2 beginning of Horizon going in, this provision of 3 litigation support, prosecution support that we 4 were obliged to do, was absolutely in its 5 infancy and each case was unique, insofar as, if 6 it wasn't just providing them with audit data, 7 if it was other pieces of information, work or 8 whatever that they wanted, it was new every 9 time. 10 So we were learning all the time how to go 11 through this process with the Post Office. This 12 was the first time, to my knowledge, that we had 13 to interact, through the Post Office, with 14 an expert witness so when I say "influence" 15 I don't mean "Come on, Jason, change your mind", 16 it was more about giving him access to the 17 things that might help him change his mind or 18 change his opinion. 19 Q. Were you, by this stage, quite frustrated with 20 the situation? 21 A. This was disappointing. The response was 22 disappointing. Essentially, we put quite a lot 23 of work into preparing our response to his 24 report and it was pretty much dismissed as 25 having no effect on his opinion. So yes, 59 1 "disappointing" is a word, not frustrated, but 2 disappointing, and we wanted to try to find -- 3 see if there was a way we could -- "influence" 4 is probably not the right word, you know, offer 5 him further evidence that might help him change 6 his opinion. That was all. 7 Q. Is "disappointing" as high as you would put it? 8 Were you still cool and professional? 9 A. (The witness laughed). Yes, yes. We were. 10 Q. Can we look at FUJ00121557, please. I'll let 11 you have a moment with this particular email, 12 Mr Holmes? 13 A. Yeah, that's fine. 14 Q. 11 March 2004: 15 "Colin, I've done a bit more to this but if 16 I continue I fear I might call him a git or 17 something worse." 18 Still -- 19 A. Well, it's like the other email -- 20 Q. -- professional? 21 A. It's like the other email where I said I don't 22 want to piss the expert off. I mean, yeah, 23 disappointing, frustrating but, you know, it's 24 just business, isn't it? 25 Q. Is it? I mean, calling somebody a "git", I mean 60 1 the impression that is given here is that you're 2 taking it a little too personally. Do you 3 disagree with that? 4 A. I do. It's an internal email. 5 Q. You originally agreed with some of the comments 6 in Mr Coyne's report. We took you to those 7 earlier. 8 A. Yeah, that's -- 9 Q. Do you think taking this approach so far down 10 the line, March 2004, was appropriate? 11 A. What, you mean going back for another -- giving 12 him the opportunity to come and visit locations 13 and speak to people and dig into systems? 14 Q. No, I mean calling an independently, jointly 15 appointed expert in court proceedings, who has 16 produced a report and a further report, some of 17 which you in fact agreed with, calling him 18 a "git"? 19 A. Well, I've got nothing more to say to that. 20 Q. Can we therefore turn to FUJ00121561, please. 21 We're now at 12 March 2004. An email from 22 yourself to Colin Lenton-Smith. You've said: 23 "I've transferred the contentious statement 24 from the paper to the email because it's not in 25 our interests to piss the expert off. That said 61 1 it has to be pointed out to Jim that his report 2 is far from impartial and, in truth, we have 3 a problem because there is little we can do to 4 dispel some of his assertions other than say 5 'rubbish'." 6 Final sentence there: 7 "POL have to decide what they want to see 8 happening here. I understand the reputational 9 aspects of the situation but I fear that POA 10 [that's the Fujitsu Post Office Account] are on 11 the back foot." 12 There's again reference to "back foot". Did 13 you still continue to see it as some sort of 14 combat between yourselves and Mrs Wolstenholme? 15 A. No, because, as I've explained before, we 16 weren't involved directly with Mrs Wolstenholme. 17 We were involved with Post Office Limited, POL, 18 okay? So the arrangement between POL and 19 Mrs Wolstenholme, and through Jason Coyne, was 20 their relationship. We were just trying to help 21 POL solve their problem. 22 Q. Who were you on the back foot against? 23 A. Well, yeah, because whether or not POL managed 24 the expert, they did very little to refute or 25 challenge the report. They just passed it all 62 1 on to us to do. So if there was an issue of 2 reputation on the system, brought through by 3 Jason Coyne's report or his opinion, POL did 4 very little to dispute or challenge it and it 5 was all down to us. And if the system came 6 under challenge through that, then obviously 7 Post Office Account were potentially at risk of 8 a reputational impact and we are on the back 9 foot. There is nothing we can do about it. 10 Q. Did you see this case as an important case in 11 which to defend your reputation as a company? 12 A. Well, obviously, anything is -- you know, 13 anything to do with the reputation is 14 an important one to defend but, as I said 15 a couple of minutes earlier, these were early 16 days, this was a learning curve, this was the 17 first time thorough for this kind of work. We 18 used to provide audit data, that's fairly 19 straightforward, you know, take the request, 20 provide a CD and a witness statement as to how 21 it was stored and retrieved, et cetera. But 22 this, where it's a bespoke, I suppose, piece of 23 work, we're learning as we are going. 24 Q. As time went on, of course, Fujitsu got more 25 involved in cases and provided witness 63 1 statements in respect of quite a number of 2 cases. Were you involved in that? 3 A. No, no. This essentially, I think, was my last 4 sort of stab at this, because from about 2005 5 onwards, I started migrating to different 6 aspects of work away from litigation support. 7 And if you go back to the very original email 8 from Pete Sewell to me, the original request was 9 sent into security and he asked if I could take 10 it up, you know, take up the work because, 11 presumably, they were busy. I don't know. So 12 I did. 13 Q. So if this was all a bit difficult for you and 14 for your department, because you weren't used to 15 this kind of thing, do you think it was then -- 16 A. No, that's -- 17 Q. -- appropriate -- 18 A. That's not what I said. I didn't say it was 19 difficult. I just said it was novel and new. 20 Q. Do you think that Fujitsu and people within 21 Fujitsu were sufficiently qualified to continue 22 to be involved in court proceedings after this? 23 A. Well, interesting one. I mean, this started off 24 as being -- asking for help to refute a report 25 from an expert witness. Yes, and I felt that we 64 1 were able to do that, and we made our attempt 2 and failed. 3 Q. If we scroll down on this page, there's the 4 draft email to Jim Cruise that was written by 5 yourself. 6 A. Yeah. 7 Q. If we look at the penultimate paragraph, or just 8 above that actually, it says: 9 "... the next step is to make available to 10 him the people, data and resources at Post 11 Office Account and allow him to address his 12 doubts to the true experts and practitioners. 13 "In conclusion it has to be said that his 14 analysis of the situation is at best selective 15 and at worst simply wrong, and his conclusions 16 partial." 17 Do you regret calling Mr Coyne's conclusions 18 "partial"? 19 A. Well, that was my opinion at the time. 20 Q. I think you saw some of Mr Coyne's evidence 21 yesterday. Looking back, do you still continue 22 to view him as partial? 23 A. Yeah, well, you're applying hindsight now. This 24 was written in 2004. That was my view at the 25 time. Whether my view has changed now, based on 65 1 what I saw yesterday or the situation as it's 2 gone on, is -- it's what it is now. But that's 3 what it was then. 4 Q. What is it now? 5 A. I don't know. Maybe we got it wrong. 6 Q. I can read to you an extract from the Bates 7 litigation. It's the Horizon Issues judgment 8 number 6. I can bring it on to screen, if you 9 like, but I'll just read you an extract from 10 paragraph 800. This is Mr Justice Fraser, he 11 says: 12 "I consider Mr Coyne to have been a helpful 13 and constructive witness ..." 14 This is in the Bates litigation, not in the 15 Cleveleys case, of course. 16 "... and I find the suggestions made to him 17 that he was biased to the claimants and not 18 independent are criticisms that are not 19 justified. He and his small number of 20 assistants had done a great amount of 21 investigation into the very numerous PEAKs and 22 the smaller number of KELs and he had embarked 23 upon a careful and sensible exercise necessary 24 for him to reach conclusions on the Horizon 25 Issues as drafted and agreed by the parties and 66 1 approved by the court." 2 I wouldn't bring it up but, for the 3 transcript, it's POL00022840. 4 Considering that, considering the evidence 5 that you saw yesterday -- or the day before 6 yesterday, in fact, of Mr Coyne, do you regret 7 the kinds of allegations that are made in these 8 emails to the Post Office, or in your internal 9 emails, the words used, "git", for example, 10 looking back now on what you have heard and what 11 you heard in Phase 2 of this Inquiry? 12 A. Well, yeah, some of those comments may have been 13 misplaced and inappropriate. But, as I said, 14 that was the situation in 2004. We're now in 15 2023 with different information available. So, 16 you know, your opinion can change. 17 Q. But, of course, you did agree with some of his 18 original opinion, and you -- 19 A. Yes -- 20 Q. -- had background from your earlier involvement 21 in the EPOSS matters in those various issues 22 that were going on within Fujitsu? 23 A. Yes, that's correct. 24 Q. Can we look at FUJ00121724, please. Now, we're 25 going to come back to this email because it's 67 1 dated 20 August 2004, so I'm taking it slightly 2 out of sequence in time. But I just want to 3 take you to number 2 there, which references the 4 Shobnall Road, case. It says there that: 5 "Shobnall Road has come back. Bill has 6 apparently been asked to provide a Witness 7 Statement to the effect that nothing contained 8 in the HSH calls over the period in question 9 could have caused, or be described as, a system 10 malfunction. I'm attaching a brief analysis of 11 the HSH transcripts that I did in April. 12 Comments made by engineers that 'keyboards can 13 cause phantom transactions' do not help the Post 14 Office's position. I suspect that we cannot 15 make the statement required and when [Post 16 Office] read the transcripts in detail they may 17 well think that they could not submit them 18 anyway." 19 The reason I'm taking this to you now is 20 there is reference to you being involved in 21 analysing the transcripts in April from Shobnall 22 Road. So this is around the same time as those 23 emails I've taken you to, where you are accusing 24 Mr Coyne of being partial. You were, it seems, 25 reading the transcripts in the Shobnall Road 68 1 case. Do you remember doing that? 2 A. No, I don't. I don't. 3 Q. Well, I can take you to the analysis of the 4 transcripts. I think it's FUJ00121725. 5 Is this the analysis that you were doing in 6 April 2004 in the Shobnall Road case? 7 A. Well, it looks like it but I don't remember it. 8 Q. Okay, I'll read you a few extracts from it. 9 If we start with 6 January 2003, it says: 10 "[Postmaster]: failing to register 11 transactions. Freezing when scanning. Phantoms 12 when unfreezing. 7 January SSC could not 13 recreate the problem as described by [the 14 postmaster]. 2 [Known Error Logs] referenced. 15 HSH then asked about lighting as could cause 16 interference. PM also using an RF baby 17 machinery and local cab firm had been operating 18 and setting off burglar alarms. PM also stated 19 suffering from power surges. Call closed and 20 raising new call as environmental but no 21 evidence of new call." 22 If we scroll down to the substantive entry 23 on the 12 December 2002, it says: 24 "New screen and cable fitted ... and both 25 engineers saying problem is NOT monitor. 69 1 Phantom transactions and not registering real 2 transactions. Certain products not registering 3 at all. HSH advise 'keyboard can cause phantom 4 transactions as well but as she has been having 5 so many problems with figures disappearing then 6 the keyboard and base unit will be swapped'." 7 An entry on 6 December 2002: 8 "PM: screen not responding and keyboard will 9 not work. Also system freezing and then 10 releasing itself. Later HSH suggested rebooting 11 but did not resolve the problem. 12 "HSH records shows 'Replaced screen due to 13 phantom transactions/calibration problems." 14 This was your analysis in April 2004. At 15 the risk of repeating myself, is now not the 16 time to say maybe Mr Coyne's report is not so 17 wrong after all? 18 I mean, we have there a subpostmaster 19 experiencing issues with transactions, they are 20 referred to by Fujitsu as "phantom 21 transactions". We've seen the earlier Known 22 Error Log from years before that referring to 23 phantom transactions. You were no doubt aware 24 of the description of phantom transactions. Is 25 April 2004 the time to say, "Maybe we should 70 1 agree with what Mr Coyne has to say"? 2 A. With hindsight, possibly, yes. But we still -- 3 I still felt it was worth having another go. 4 Post Office didn't. They didn't offer the 5 response to Mr Coyne, so that was the end of it. 6 Q. Can we turn to FUJ00121637, please. We're now 7 moving to June 2004, 7 June. It's an email from 8 yourself to Colin Lenton-Smith. I'm going to 9 read to you a few extracts: 10 "POL are still taking advice as to how best 11 to deal with this and Mandy's view/belief was 12 the safest way to manage this is to throw money 13 at it and get a confidentiality agreement 14 signed. She is not happy with the 'Expert's' as 15 she considers it to be not well balanced and 16 wants, if possible, to keep it out of the public 17 domain. This is unlikely to happen if it goes 18 to Court." 19 The next paragraph, the final sentence: 20 "The liability question is removed and it's 21 then just about 'how much to go away and keep 22 your mouth shut'. 23 "One concern I have is that while they've 24 been dickering about waiting for guidance from 25 their agents, the trial date has been set and 71 1 it's now too late for them to enter a Witness 2 Statement that might further repudiate the 3 Expert's original report. This means that their 4 Council [I think that means counsel with an 'S'] 5 might have to have thorough briefing, by us, 6 before going to Court." 7 By 7 June 2004, do you think that you, by 8 that stage, had lost all objectivity in respect 9 of this matter? 10 A. What makes you say that? 11 Q. Well, for example, the history of EPOSS; error 12 logs that we saw; the table that was sent to 13 you; Richard Brunskill's comments; your 14 agreement -- 15 A. No, I know all that but what makes you say had 16 I lost all objectivity? 17 Q. Looking at the final paragraph, you are still 18 trying to further repudiate the expert's 19 original report and that is a report that, in 20 some respects at least, you originally agreed 21 with. Why are you persisting in June 2004 to 22 try to fight this matter? 23 A. I'm not. This is an internal mail to Colin 24 saying it's now too late for them, POL, to enter 25 a witness statement that might further repudiate 72 1 the expert's original report, if they chose to 2 do it. But if they didn't, and they didn't, 3 then it's a different end game, isn't it? 4 Q. The impression given by that email though is 5 that you're quite frustrated that you're not 6 going to be able to defend yourselves? 7 A. No, that's not so. That's your interpretation. 8 That's not what was meant by that. 9 Q. I'm giving you an opportunity to say whatever 10 you like on that. What's your interpretation of 11 this email? 12 A. Just that if they wanted to further repudiate, 13 which they decided not to, then it's getting 14 very late, the trial date has been set, for them 15 to answer a witness statement that might further 16 repudiate the expert's original report. That's 17 all. 18 Q. The final sentence: 19 "Do we need to involve Masons at this 20 stage?" 21 Masons being Fujitsu's lawyers, is it? 22 A. Yes, that's correct. 23 Q. Why were you wondering whether you needed to 24 involve your own lawyers at that stage? 25 A. It was -- well, simply because it was getting up 73 1 to a court situation and it was a question that 2 I was asking Colin as the Commercial Manager, do 3 we need to involve Masons at this stage? 4 Nothing more or less than that. 5 Q. Can we look at WITN04600309, please. Is this 6 your note? 7 A. Yeah, this is my daybook. 8 Q. Can you assist us with the date in the top 9 left-hand corner? Is this the same date, is 10 this 7 June 2004? The middle number could be 11 a 5 could be a 6. Perhaps we could zoom in on 12 that if possible. 13 I think that's a 6. Is that 7 June, the 14 same date? 15 A. It looks like a 6, yeah. 16 Q. The final entry on that page is: 17 "Fear that throwing money at the problem is 18 the only way to deal with it: 19 "[Either] admit [the] report. 20 "Concede [the] content is accurate (it is 21 but opinion is crap). 22 "Liability is removed 23 "Question 'how much to keep your mouth 24 shut!'" 25 What do you mean here in brackets, that the 74 1 content is accurate but "opinion is crap"? 2 A. Well, that's a contemporaneous record of the 3 conversation I had with Mandy Talbot on the 4 phone, so it must have come out of that 5 conversation. 6 Q. Do you think those were her words or your words? 7 A. I can't remember. 8 Q. Can we look at FUJ00121668, please. 30 July 9 2004, we have an email from yourself to Colin 10 Lenton-Smith. You've: 11 "... spoken to Keith Baines who alluded to 12 a number of other calls that he was going to 13 have to make on the case but didn't pass any 14 details on ... He said that Dave Smith would be 15 speaking to Ian on the subject -- it seems Dave 16 believes 'we' (not sure whether that's the Royal 17 we or just us) have conceded what 'we' should 18 not have done and POL are now in a difficult 19 position. Given our late involvement by POL 20 I trust he's not trying to park it all on us." 21 At this stage, July 2004, was there 22 something of a blame game going on in relation 23 to this case? 24 A. Well, no, it's just that, in that conversation 25 with Keith, obviously, you know, he said Dave 75 1 Smith and we saw that slide that Dave Smith 2 produced in the session this morning, with Jason 3 Coyne, who showed him the slide that Dave Smith 4 had produced. Again, I don't know who "we", is, 5 I don't know if it's "we" POA or "we" POL and 6 POA have conceded what we should not have done, 7 and I'm not sure what we did concede, and now 8 POL are in a difficult position. So, you know, 9 so my concern there was well, is he now trying 10 to park the blame on us or is he sort of saying 11 that, you know, we've got to find a way of 12 managing this? 13 And my comment about "given our late 14 involvement", bear in mind that we didn't get 15 involved in this until -- was it February 2004? 16 And the issues were all happening in 2000/2001. 17 You know, I just wanted to be certain that he 18 wasn't going to try and push this all onto us 19 and say we've done a bad job because I didn't 20 think we did. 21 Q. Was there pressure in that respect? Did you 22 feel pressure coming from POL that you may 23 not -- that you may be blamed in this case? 24 A. Well, no, I mean, the words there are the words 25 that came from that conversation with Keith. My 76 1 feeling, my understanding, my belief and my 2 concern but seeing as I didn't know who "we" 3 were, was it POA, was it POA and POL, or was it 4 POL -- yeah. So, you know, I was just concerned 5 that we were -- he was going to try and say it's 6 all our fault. 7 Q. Moving to August 2004, can we look at 8 WITN04600310, please. Is this also a note that 9 you made? 10 A. Yeah, it was a case conference call that was 11 held in August and I just making handwritten 12 notes of the progress of the meeting. 13 Q. Now "SL" is POL's counsel. So when there are 14 references to "SL", it seems that is to Stephan 15 Lewinski? 16 A. Again, incorrectly spelt, so I apologise for 17 that but yes. 18 Q. His advice is recorded there. If we look at the 19 final two sentences of his advice, it says: 20 "[Had] if goes to dispute likely to find 21 that computer system let [Julie Wolstenholme] 22 down." 23 Was it your understanding that counsel's 24 advice was that, if it went to court, the court 25 was likely to find that the computer had let the 77 1 subpostmistress down? 2 A. That's what you can read into that. That was 3 his opinion. 4 Q. Can we look page 3, please. We have halfway 5 down that page another opinion from the Post 6 Office's counsel. It's recorded as saying: 7 "Cannot say that there were no glitches. 8 Will be candid about that but what did we do to 9 help it." 10 So the focus seems to be not on the fact 11 that there were no errors but now it's turning 12 to the Helpdesk and assistance that was provided 13 to Mrs Wolstenholme. Do you recall that? 14 A. Well, that's what that looks like. 15 Q. At this meeting, did you feel able to talk 16 freely? 17 A. Yeah, I mean, there's a couple of entries with 18 my initials against it. 19 Q. Yes. I mean, if we look at the first page, it's 20 a discussion that involves yourself, POL 21 commercial, POL Legal, Weightman Vizards, 22 Masons -- so that's your own solicitors, is it? 23 A. Yeah, yeah. 24 Q. POL Commercial -- sorry, POL counsel there. 25 A. Yeah, so, I mean, essentially, it was a POL case 78 1 conference call and I was there, I think, 2 probably because of my involvement in it up to 3 that point. 4 Q. If you disagreed with something would you have 5 felt free to say so? 6 A. Yes, but I would have been out of my depth 7 talking to, you know, the legal guys. 8 Q. If we look at page 4, it says, I think it's 9 "Status of 2nd report". That's the second 10 report that you had written? 11 A. Yes. 12 Q. "Not forwarded to Mr Coyne." 13 A. Correct. 14 Q. "KB and I need to get our act together. KB 15 agreed." 16 Now, that second report was the one that 17 offered, I think, for Mr Coyne to attend 18 Fujitsu? 19 A. Yes, that's correct. 20 Q. Do you recall why it wasn't forwarded to 21 Mr Coyne? 22 A. No, no. I'm not sure I found out about it until 23 really quite late after the event. But it 24 wasn't forwarded to Coyne. I don't know why. 25 POL must have had a view that it wouldn't have 79 1 made any difference, so they ain't gonna do it. 2 Q. This was August 2004, where you were part of the 3 meeting with the Post Office and the legal team. 4 Was that a decision that you were part of? 5 A. Sorry, what decision: not to forward the report? 6 Q. Yes. Certainly it reads as though you were part 7 of the decision-making process in that respect? 8 A. No, that's how you're reading it but I wasn't 9 part of that decision-making process. That was 10 a POL decision on their own. 11 Q. Did you respond, when you were told that it 12 wasn't forwarded to Mr Coyne? 13 A. No, no, we were just disappointed. 14 Q. It says: 15 "[Keith Baines] and I need to get our act 16 together." 17 What did you mean by that? 18 A. Well, I think there was some outstanding 19 documents that we needed to get sorted out and 20 it was just -- it's just a phrase, isn't it? 21 "We need to get our act together". Keith 22 agreed. And there's an email following it, 23 where there's a list of items that I said 24 I would be producing, which I did, and I think 25 Keith had a similar list. 80 1 Q. Can we now look at WITN04600215, please. 2 16 August, we have an email from yourself to 3 Bill Mitchell. And you say: 4 "Bill, 5 "Just a quick note to let you know that 6 Mrs Wolstenholme finally accepted an increased 7 offer from the [Post Office] to settle her case. 8 Equipment will be returned as part of the Court 9 Order and she will have a confidentiality clause 10 associated with her settlement. However, she 11 was still going to call a host of other PMs as 12 part of her case so I guess the 'issue of poor 13 systems and inadequate levels of support' 14 argument could well be rolled out again." 15 In that meeting that you had with counsel, 16 the advice from counsel was that, if you get 17 into a dispute, it is likely to be found that 18 the system let Mrs Wolstenholme down. Also, 19 advice from counsel was that you cannot say that 20 there were no glitches. Did you still think 21 that the issue of poor systems and inadequate 22 levels of support argument was just a silly 23 argument that it was being rolled out? 24 A. No, I was just advising Bill that I think we 25 could see that one again. 81 1 Q. Did you, at that stage, think that there might 2 be something in that argument? 3 A. I can't say that I did, to be honest. I don't 4 think I put too much into it. 5 Q. I mean, the expression "rolled out" suggests 6 that -- I mean, it's a line that's rolled out 7 time and time again but not something with 8 substance in it. Is that a fair analysis of 9 what is said earlier or is that unfair? 10 A. No, again, I think that's your interpretation. 11 All I'm saying is that it's quite possible that 12 the issue of poor systems and inadequate levels 13 of support could well be rolled out in future 14 cases -- 15 Q. Did you a view -- 16 A. -- because it's -- 17 Q. -- as to those who were rolling it out in cases? 18 A. When you say "did I have a view", do you mean 19 did I have an opinion of the postmasters? 20 Q. Absolutely. Did you have an opinion as to the 21 type of person who rolled out that argument? 22 A. No, not at all. It's just an obvious thing to 23 do. If a case has been won using that argument, 24 then future cases could well use the same 25 argument in order to try and secure a victory. 82 1 Q. Can we look at FUJ00121724, please. This is the 2 document we looked at earlier and I said I'd 3 return to it, so we're now at 20 August and it's 4 to Colin Lenton-Smith and Bill Mitchell. We've 5 looked at the Shobnall Road issue but let's look 6 at Cleveleys now. It says: 7 "Although Cleveleys may appear to be closed 8 it could be construed that POL bought off 9 Mrs Wolstenholme rather than defend their 10 system. Even if a gagging order is placed on 11 the woman, she apparently has a gaggle of 12 postmasters lined up to support her case and 13 they will be well aware of what the final 14 outcome was." 15 Do you think, looking at that now, that that 16 was an appropriate stance to have taken? 17 A. Well, the words may be unfortunate but no, 18 I think it's to be expected, isn't it? 19 Q. Reference to a "gaggle of postmasters", I mean, 20 do you regret what appears to be effectively 21 a demonisation of subpostmasters in this regard? 22 A. Well, I think "demonisation" is a bit of 23 a strong word. I could have said, you know, 24 a "collection of postmasters", a "gaggle of 25 postmasters", I don't know what the collective 83 1 term is for them but she apparently was going to 2 circulate through a number of postmasters the 3 outcome of the court case and it was, you know, 4 I guess inevitably that the -- they would use it 5 to their benefit. 6 Q. What could possibly be wrong with subpostmasters 7 using success in a court case to their benefit? 8 A. Nothing. 9 Q. So why the critical words and why such 10 an attempt to prevent that from happening? 11 A. Sorry, where was I attempting to stop that 12 happening? 13 Q. Well, the continued effort to defend and respond 14 to Mr Coyne's independent report over months may 15 be said to be a continued effort to try to 16 prevent a negative outcome that could be shared 17 with a "gaggle of postmasters". Do you agree 18 with that? 19 A. No, I don't. No, I mean, the two things are 20 separates. I mean, in trying -- in refuting or 21 in arguing against the -- Jason Coyne's report, 22 that was one part of our support to the Post 23 Office. That's been and gone to now. This is 24 after the event. 25 Q. Do you not think that that reference there shows 84 1 concern on your part that Mrs Wolstenholme would 2 share what was a successful court case on her 3 part with other subpostmasters and that that 4 would cause you some sort of problem in the 5 future? 6 A. I don't think so. I mean, it just says what it 7 says. Even if a gagging order is placed -- 8 because don't forget, as far as I was aware, 9 I was talking to Mandy Talbot, she said that she 10 would want to get a gagging order. So I'm 11 saying I don't know what the outcome was but 12 even if it was in place, it's highly unlikely 13 that it would stop her or stop others trying to 14 find out what the outcome was and how that was 15 achieved. 16 Q. "Even if a gagging order is placed on the 17 woman", do you not think that the words used 18 were derogatory at that time, showed a sense of 19 frustration? 20 A. No, no. 21 Q. Is there anything that you would like to raise 22 with the chair before I finish? 23 A. Well, no. I did mention earlier about the fact 24 that this was a -- this was a unique situation 25 that we found ourselves in, that I found myself 85 1 in and we were having to learn how to do this 2 and to pick up how to do this as we went along. 3 It was the first time I'd been involved with 4 an expert witness as such. 5 POL had asked us to challenge the report, 6 which we did. All right, perhaps we took of the 7 challenge one step too far by offering him site 8 visits, and what have you, but there comes 9 a point when you want to see something reach 10 a natural conclusion and the second -- the offer 11 to host him and let him have access to people 12 and systems, was probably the last toss of the 13 die. The fact that Post Office didn't forward 14 that report to him was their choice and then we 15 take the next step which is going to court. 16 You know, as I said at the start, it's 17 a job. We had a job of work to do and we did 18 that work. 19 SIR WYN WILLIAMS: Did you ever discover why Post 20 Office decided against making the offer that you 21 thought that they should? 22 A. No, no sir, I didn't. 23 SIR WYN WILLIAMS: All right, thank you. 24 A. At one level, it didn't seem important. They 25 chose not to do it. It was their report. It 86 1 was their work. If they don't want to do 2 anything with it, then it's up to them, it's not 3 up to us. 4 SIR WYN WILLIAMS: No, I follow that. I just 5 wondered whether, because it was a specific 6 suggestion that had been made, no doubt made by 7 you but having no doubt considered it with 8 others, whether you would have liked to have 9 known the reason why they didn't do it. But 10 there we are. 11 A. Yeah, perhaps natural curiosity didn't get the 12 better of me that time. 13 SIR WYN WILLIAMS: Fine. Okay. 14 MR BLAKE: Sir, do you have any other questions, 15 otherwise we have some on behalf of Core 16 Participants? 17 SIR WYN WILLIAMS: No, they can fire away. 18 MR BLAKE: Thank you very much. 19 A. That's a bit of an unfortunate phrase! 20 SIR WYN WILLIAMS: There we are. Even chairs of 21 inquiries have to occasionally choose their 22 words carefully, Mr Holmes. 23 Questioned by MR JACOBS 24 MR JACOBS: Thank you. 25 Good afternoon, Mr Holmes, I represent 157 87 1 subpostmasters. I have couple of questions for 2 you. Could we go to document FUJ00121486, 3 please. We'll just wait for it to come up on 4 the screen. If we scroll down to just before 5 the paragraph beginning "The Expert's Opinion", 6 so that's right. So the paragraph that begins 7 "On 6th February". So it says: 8 "On 6th February POA received a copy of the 9 Expert's report with a request from POL for 10 an early response." 11 Then: 12 "POL are concerned that the Expert's opinion 13 (that the system was at fault) might set 14 a precedent against future POL prosecutions." 15 Now, this is a draft response that you sent 16 to Mr Lenton-Smith in February 2004, in relation 17 to a letter that he'd received from Mr Baines. 18 Do you recall that? 19 A. Yeah, yes, I do. 20 Q. Mr -- 21 A. Sorry, I'm looking to the side. I'm looking at 22 my laptop to find some documents, which I think, 23 or a daybook scan, where that concern was raised 24 in a conversation. 25 Q. Well, I'm going to ask you about that a bit 88 1 later. Mr Blake put the same point to you from 2 another document and you said that that was your 3 understanding at the time; is that right? 4 A. Yeah. 5 Q. If we could now go to your witness statement, at 6 paragraph 10 and your statement is WITN04600200. 7 A. Is this my third or second witness statement? 8 Q. I think it's your second witness statement. 9 It's your second witness statement, yes. 10 A. Right. Let me just get there. Yeah. 11 Q. Paragraph 10, please, which is -- I've got 12 page 76 but I think it's page 4 of 10? 13 A. Yes, that's correct, I've got it. 14 Q. We'll just scroll down to 10. So you say that 15 you were exasperated with the Post Office and 16 their handling of this issue and their handling 17 of the expert witness and lack of communication. 18 Then you say here: 19 "[Post Office Limited], specifically Jim 20 Cruise and Mandy Talbot, had both verbally 21 expressed concerns about precedent should the 22 Expert's report become common knowledge and had 23 asked for our assistance in challenging it." 24 Now, what I want to ask you is what did Jim 25 Cruise and Mandy Talbot verbally express about 89 1 precedent? You haven't said what it was that 2 they said. 3 A. Well, that would have been in phone 4 conversations. I'm just looking at -- just bear 5 with me a minute. Well, you've seen the day 6 book scan from the phone call with Mandy Talbot 7 in June. No, I mean these concerns would have 8 been expressed in phone conversations and also 9 in Keith Baines's letter. Obviously I've got 10 a record of Mandy Talbot's phone conversation 11 but not Jim's, where that concern was expressed. 12 So in a -- my recordkeeping is not as good as it 13 usually is. 14 Q. Well, of course, this was a long time ago but 15 you say in your statement that they both 16 verbally expressed concerns and I really wanted 17 to ask you whether you remember what it was that 18 they said, what specific concerns they expressed 19 about precedent. 20 A. Well, no, I suspect what I'm talking about there 21 is the fact that if the -- if the report -- 22 well, which it did, of course -- if the report 23 resulted in a positive outcome for the 24 postmaster, then I think they were concerned 25 that this issue of poor systems and poor support 90 1 would be, as I've said in another email, rolled 2 out time and again by postmasters who were 3 suffering the same sort of issue. 4 Q. You see, our clients' take on this evidence, and 5 I suggest that this must be right, is that Post 6 Office didn't want subpostmasters to use this 7 material in court because it might lead to them 8 being acquitted. That's really what this is 9 about, isn't it? 10 A. Well, I think that's what I've just said but 11 you've probably put it much better than I can. 12 Q. Thank you very much. 13 So our clients also say, and I wanted your 14 comment on this, that really what the Post 15 Office were doing was they were, in 2004, trying 16 to cover up evidence of defects in the Horizon 17 System. Do you accept that? 18 A. I can't really answer to that one. I don't -- 19 Q. Well -- 20 A. I don't know what their motive -- motivation was 21 to handling the case. I just know what we did 22 to try to assist them in what they were asking 23 us to do. 24 MR JACOBS: I'm just going to ask if I have any more 25 questions to ask. 91 1 No, that's it. Thank you very much. 2 MR BLAKE: Mr Moloney or Mr Henry. 3 Questions by MR MOLONEY 4 MR MOLONEY: I've just two things to ask you about, 5 please, Mr Holmes. 6 A. Yeah. 7 Q. Firstly, when you prepared your response to 8 Mr Coyne's report or indeed your response to his 9 reply, did you check PinICLs or KELs relating to 10 Cleveleys? 11 A. No. 12 Q. You're sure about that? 13 A. Yeah. I mean, I didn't get down to the 14 technical level of looking at individual 15 PinICLs, PEAKs or KELs. 16 Q. Right. 17 A. The discussion was a slightly higher level of 18 just, you know, the principle of whatever it 19 was -- I can't remember now, so far back. 20 Q. Could we, just to clarify that, just have a look 21 at WITN04600206. This is the first page. This 22 is "Fujitsu Services Post Office Account 23 Response to the Expert's Reply to Fujitsu 24 Services Submission"? 25 A. Yes, so this is our second report. 92 1 Q. Yes, and under the "Horizon System Helpdesk", if 2 we scroll down slightly to the second 3 subparagraph, we can see that: 4 "His statement that the HSH was not 5 interested in getting to the 'bottom of the 6 continual occurrence' is flawed given that one 7 of the calls that he specifically referenced was 8 closed using a Known Error Log. The presence of 9 a KEL clearly indicates that problems were 10 investigated and workarounds provided pending 11 a permanent fix through a system upgrade." 12 So does that in any way affect your answer 13 that you didn't look at PEAKs or KELs in 14 relation to Cleveleys or PinICLs? 15 A. No, what I suspect may have happened there and 16 I can only surmise that that was the case, is 17 that I may well have had conversations with 18 interested parties about the response that he'd 19 made and the question of HSH not interested in 20 getting to the bottom of the continual 21 occurrence was probably challenged by somebody, 22 one somebody I was talking to, and saying "Well, 23 no, that's not true because we have KELs", which 24 means that we do look at the problem to try to 25 come up with a quick solution for future use, 93 1 pending an update or a reissue of software at 2 a later release date. 3 Q. The second issue concerns that, as it were, 4 getting to the bottom of things and root cause 5 analysis. Could we look at -- 6 A. Yeah. 7 Q. -- document POL00089802. This, as you can see, 8 is an "Audit of Customer Service Support 9 Processes", and we can see that the originator 10 is yourself, Mr Holmes? 11 A. Yes, yes. 12 Q. It's 1 November 2001, and if we could please go 13 to paragraph 3.1, at page 3 -- thank you very 14 much -- which reads that: 15 "The overall opinion formed is that the 16 management of Incidents, Problems and Complaints 17 and Alerts ... has reached a level of maturity 18 where consistency now the norm ..." 19 But if we see that, further down: 20 "There are a number of relatively minor 21 issues that, while not impairing the current 22 management of incidents and problems could, if 23 accepted and addressed improve the performance 24 of this part of [Customer Service]. They are 25 ..." 94 1 Then at 2: 2 "Introduce formal Root Cause Analysis into 3 Problem and Complaints management as a matter of 4 course. This is already being addressed within 5 [Customer Service]." 6 Then, at page 7 of the document, we see 7 4.1.2, "Problem Root Cause Analysis": 8 "CS/PRD/021 also introduces the concept of 9 Root Cause Analysis for problems although again, 10 no guidance or examples are provided. No 11 evidence of completed [Root Cause Analyses] 12 could be identified. 13 "It is recommended that RCA guidance is 14 provided, either in CS/PRD/021 or in 15 a standalone procedure." 16 So by November 2001, around the time of or 17 just after, even, the Cleveleys relevant event, 18 there was sufficient concern to note that 19 there'd been no evidence of any RCAs completed 20 for any problems addressed so far. 21 A. Well, yes, that's what's in the report, so that 22 must have been what was found. 23 Q. You're the originator of this report, Mr Holmes? 24 A. Yeah, yeah. 25 Q. Was this not consistent with the analysis of 95 1 Mr Coyne? 2 A. Well, yes, I suppose it is. I mean, yeah. 3 Q. Was that reflected in your response to the 4 report of Mr Coyne? 5 A. I -- well, you can show me the response but 6 I guess what you're saying is that, although 7 we'd found this in an audit in 2001, it wasn't 8 reflected back in our response to him. 9 I suppose, in my defence, there is an awful lot 10 going on in different streams. So, yes, this 11 would have been found at the time of the audit 12 but it doesn't stick in the forefront of my mind 13 when I'm looking at other things going on. Now, 14 that may sound like a poor excuse, I know, but 15 that's the way it is. 16 Q. Shall we just look at your response, just to 17 assist you, which is FUJ00121504. Towards the 18 bottom of page 1, if we could, please. There we 19 see in terms of the Helpdesk: 20 "The HSH represents the 1st line of support 21 to postmasters. It operates under strict 22 Service Level Agreements covering aspects such 23 as pick-up time, first time fix and time to 24 close. These measures are imposed by Post 25 Office Limited and are designed to ensure that 96 1 PMs receive a quick response to their call and, 2 to the extent possible over the phone, a timely 3 return to normal business operations. 4 Depending on the nature of the call the HSH, 5 operator would work with the PM to solve the 6 problem and return the Outlet to normal 7 operation as soon as possible, in line with the 8 prevailing SLAs. If this could not be achieved 9 the call would be escalated up the support 10 channel to 2nd, 3rd or 4th line depending on the 11 severity of the problem. Again, the primary 12 objective is to return the Outlet to normal 13 operation as soon as possible and rebooting the 14 counter often meets that objective. This does 15 not mean that the problem was closed at that 16 point in time, as a detailed scrutiny of overall 17 problem management in the Post Office Account 18 would reveal." 19 Then we see into "Transaction Handling on 20 Reboot". 21 That's a fairly generic statement about the 22 purpose of HSH, isn't it, Mr Holmes? 23 A. Yes, it is, yeah. 24 Q. Did you take any steps to investigate whether 25 there'd been any root cause analysis in respect 97 1 of problems at Cleveleys? 2 A. No. 3 Q. Why not? 4 A. Because that wasn't really the objective of what 5 I was trying to do with this response. 6 Q. Did you have skin in the game here, as you said, 7 Mr Holmes? 8 A. Yeah, I'd been asked to do something and I was 9 doing it to, at that time, the best of -- what 10 I felt was the best of my ability. Of course, 11 21 years later, it's very easy to criticise and 12 come back and say "Well, you didn't do that very 13 well, did you?" And the answer is, no, 14 I didn't, I can see that and I could have done 15 more but, at that time, I felt that was what was 16 required. 17 MR MOLONEY: Thank you, Mr Holmes. 18 Questioned by MR HENRY 19 MR HENRY: Good afternoon, Mr Holmes. Edward Henry, 20 representing a number of subpostmasters, 21 including Ms Tracy Felstead. 22 Mr Holmes, could I just ask you, and it's in 23 relation to a question that you have been asked 24 by my learned friend just now, could I ask you 25 to go to FUJ00075674, please. Can you see that 98 1 clearly, Mr Holmes? 2 A. Yes, I can now. Yeah. 3 Q. You can. What does it say? It's a PEAK 4 incident management system. So what does that 5 mean? 6 A. Well, PEAK was the system where all incidents 7 were logged. 8 Q. You were, you say, not involved at that level? 9 A. I was on this one because this was an issue that 10 affected our ability to produce audit data 11 request -- request sort of thingy-bob -- what's 12 the word -- SQL type query statements because 13 there was no Attribute Grammar catalogue 14 available for the TMS journal. So we weren't 15 able to look at a request coming in from outside 16 and say "Well, in order to satisfy that we need 17 to get this piece of information, this piece of 18 information, that piece of information, and 19 stick it together as a query type statement". 20 So that was just identifying the absence of 21 a piece of technical documentation. 22 Q. Did this feature, this problem, that you have 23 just referred to, in the original report that 24 you wrote with Mr McDonnell? 25 A. No, I don't think it did. I don't think it did. 99 1 Q. But would you agree that it would again be 2 symptomatic of a financial accounting system 3 that was unfit for purpose? 4 A. No. No, this -- all this was, was just saying 5 we didn't have any record descriptions that 6 would allow us to search through the records to 7 pull audit data off. Nothing to do with what 8 the system itself was doing. 9 Q. But it's very, very important, isn't it, in 10 order to analyse root cause analysis of problems 11 in the system and, also, the recording and 12 retention of data must be fundamental to 13 a financial accounting system that works? 14 A. Well, I agree with you, and the TMS journals 15 were there in the audit archive and were 16 complete and valid and accurate. What we didn't 17 have was the means to get into them to pull the 18 information out that was required on request. 19 That was all. I'm trying to think of 20 a parallel. 21 Q. Well, I suppose the parallel could be, it could 22 be like, yeah, you've got all of the stuff, like 23 the hieroglyphs but until you've got the Rosetta 24 Stone, you don't know what the hieroglyphs are 25 because you can't access them. 100 1 A. Absolutely right. 2 Q. Right. So let's just go to the first entry 3 there, 1 July: 4 "The ability to interrogate TMS journals is 5 an integral element of the Audit Solution that 6 we supply to POCL. The interrogations are 7 achieved using R-Query and constructing SQL type 8 query statements. The structure of the TMS 9 records is complex and uses Attribute Grammar to 10 establish identities for the record attributes. 11 There is currently no definitive catalogue of 12 the Attribute Grammar used in the Horizon 13 solution ['solution' perhaps should have been 14 put in inverted commas] which makes the 15 construction of an R-Query statement to meet 16 a customer's business enquiry difficult, time 17 consuming and very 'hit and miss'." 18 Were those your words? 19 A. Yes. 20 Q. "Hit and miss". 21 You are still in control of this issue in 22 September 2001, are you not? 23 A. Yes. 24 Q. Because we go to the 12 September 2001 at 11.05. 25 The call record has been assigned to the team 101 1 member Jan Holmes, correct? So you're still in 2 control of this in September 2001, correct? 3 A. Where does it say that? 4 Q. Go to 12 September 2001, which is at page 3 of 4 5 in the internal numbering, 11.05. 6 A. Yeah, I can see it. 7 Q. You're still in control of this 21 June 2004? 8 A. Well, it's been passed back to me because as 9 I was the originator. So it's been sent back to 10 me and it says: 11 "This has been identified as a requirement 12 for Network Banking and has been included in the 13 NWB Performance Improvements document. It has 14 not been transferred to the SDS but the PinICL 15 remains open pending the delivery of such 16 a catalogue as part of BI3." 17 So what was happening there was the 18 production of this catalogue had been put on the 19 back burner, and -- until a later release of 20 Horizon was made. 21 Q. But it's still -- if we go to the last entry but 22 one, 21 June 2004, there's an entry: 23 "The call record has been assigned to the 24 Team Member: Jan Holmes." 25 A. Yes, that's right. 102 1 Q. That's 21 June. 2 A. Yeah, because on 19 January, there's 3 a document in PVCS, with a reference at version 4 0.4, that is an up-to-date Attribute Grammar 5 catalogue. So in other words, they had produced 6 the document that was missing when the PinICL 7 was first raised. So that was passed back to me 8 as the originator to close it, which I did. 9 Q. But this was a serious problem and it wasn't 10 closed until 2004? 11 A. Why do you say it's a serious problem, Mr Henry? 12 The absence of a catalogue that describes the 13 records for audit to build SQL queries was 14 an inconvenience and a problem and a nuisance 15 but it wasn't a serious error. 16 Q. There is currently no -- you agree with me it 17 was like the hieroglyphs before the discovering 18 of the Rosetta Stone: 19 "There is currently no definitive catalogue 20 of the Attribute Grammar used in the Horizon 21 solution which makes the construction of 22 an R-Query statement to meet a customer's 23 business enquiry difficult ..." 24 Your words, "difficult". 25 A. Yeah. 103 1 Q. "... time consuming and very 'hit and miss'." 2 That's a serious problem, Mr Holmes. 3 A. In your opinion, possibly. But it was just one 4 of a number of things that we were dealing with. 5 Q. That's a serious -- 6 A. We could -- 7 Q. -- problem, Mr Holmes, even if it is one of 8 number of things you may be dealing with 9 because, as you accepted on the last occasion 10 you gave evidence before the Inquiry, the Post 11 Office was entirely dependent on Horizon data 12 for its prosecution of subpostmasters. You 13 agreed with that? 14 A. Yes, that's right. 15 Q. Right, okay. 16 A. So all the absence of the Attribute Grammar 17 catalogue did was just made our life slightly 18 more difficult when we were pulling the audit 19 data. The audit data was there. It just -- 20 Q. The audit data has to be picked out impartially, 21 independently, comprehensively and fully. It 22 can't be picked out on a hit and miss basis. 23 Surely you agree with that? 24 A. Exactly, that's why the PinICL was raised. 25 Q. Yes. Right. 104 1 Now, that is not closed until 2004, is it? 2 A. That's correct, yeah. 3 Q. Right. Ms Felstead was convicted in 2002. So 4 I now move on to her case. You were asked about 5 this on the last occasion, and I was asking 6 about Tracy Felstead when I was questioning you 7 on the last occasion. Would you like to go to 8 the transcript of that? Shall we go to it? 9 It's INQ00001019. 10 Can I just do a quick whistlestop tour 11 through this transcript. Could we go internal 12 numbering, please, to 5 of 83, page 20 of the 13 transcript, line 7. This is when Mr Beer is 14 questioning Mr McDonnell, and it's page 20. 15 Yes. Thank you. Line 7. This was about the 16 EPOS System being the joke of the building. 17 Then he said: 18 "Yes, I think everybody knew, specifically 19 the test team who, when I spoke to those guys, 20 they would make it very clear that the quality 21 of the code that was being deliver was to such 22 a bad, poor level that they're wasting their 23 time testing it, because they knew that it was 24 just broken. They were going to end up raising 25 lots of PinICLs from it. So they'd give a very 105 1 frank and very honest opinion about the ability 2 of some of the guys, not all of them -- some of 3 them were good -- in the team, and the quality 4 of the product that that team was producing. It 5 was a standing joke in the building." 6 Mr Beer said: 7 "You say in your statement it was known up 8 to the highest level, including Fujitsu Japan, 9 because they sent over three coders to perform 10 an audit." 11 Mr McDonnell confirms that he can't remember 12 what date that actually was. 13 Could we go to page 31, please. At page 31, 14 I'm sorry I don't see the numbering on this. 15 I'm so sorry. I was meaning page 31 of the 16 actual transcript. It's internal page 8. 17 That's entirely my fault. Internal page 8 of 18 83. Thank you. So is that page 8 of 83. Thank 19 you very much. Page 31, lines 1 to 25, and this 20 is: 21 "... he says there was a concern which he 22 considered to be the greater of the two concerns 23 which relate to the impact of continual changes 24 to existing code to fix problems and/or to 25 insert new functionality into the code. Do you 106 1 agree with Mr Holmes that that was a concern?" 2 Because he was being taken to your concerns, 3 Mr Holmes, and the answer was: 4 "I do and, in fact, within this document 5 there's a very good example of that when, during 6 the Taskforce, which was supposed to be all 7 about getting the quality under control, they 8 took away of sum of the resource to force in 9 extra functionality for, I think it was 10 balancing and something else. There's three 11 parts to it. It's referred to in the document 12 somewhere. But it was a sizeable piece of 13 development work which was being developed on 14 the fly and shoe-horned into the code right in 15 the middle of the Taskforce initiative, where we 16 were trying to stabilise the product, and that's 17 a typical example of not understanding the 18 problem of where we were at the time and 19 continuing with the same bad behaviour, in my 20 view. 21 "Question: Those two concerns that 22 Mr Holmes mentions and which you agree with, in 23 your view, would they have had any impact on the 24 integrity of the system, how it operated or how 25 it was operated by subpostmasters? 107 1 "Answer: Yes, it would. It would result in 2 functional errors, bugs, spurious behaviour." 3 Internal -- forgive me, I just go to the top 4 of page 32: 5 "Was that the view held by you and others at 6 ICL Pathway at the time? 7 "Answer: Yes, it was. I think it was 8 a belief that was pervasive throughout the 9 building." 10 Of course, Mr Holmes, you being the 11 co-author of the report, it would have been 12 a view that you trenchantly shared together with 13 your co-author, wouldn't it? 14 A. I would have shared it with him, yes. 15 Q. Trenchantly, I'm sure. You're not backward in 16 coming forward, Mr Holmes. You're the co-author 17 of that report. 18 A. Well, all right, if you want to use that term, 19 feel free. 20 Q. Well, Mr Holmes, it's not a secret that when you 21 wished to express an opinion, you express 22 an opinion with a certain degree of confidence 23 and, shall we say, vigour? 24 A. No, "trenchantly" is not a term that I would 25 normally use. That's all. 108 1 Q. I see. Confidence, then. Unequivocal in your 2 view? 3 A. No, I agreed with the report. It was a joint 4 report. So I was, you know, happy to concur 5 with him. 6 Q. Page 11, internal, please. 7 MR BLAKE: Sir, I'm sorry to interrupt, Mr Henry. 8 In terms of timing, I just want to make sure 9 because we're certainly not going to use up this 10 afternoon's session. The question is, whether 11 we take a short break now or at some appropriate 12 point before Mr Henry continues or whether 13 Mr Henry considers he will be finished shortly. 14 If not, there's no problem at all. 15 SIR WYN WILLIAMS: Well, Mr Blake, I was beginning 16 to wonder. I'm not really keen, Mr Henry, on 17 going back to the points that you're focusing on 18 at the moment, unless they are really laying the 19 ground for some important point in relation to 20 this morning's evidence. 21 MR HENRY: Well, they are, sir, but I can take them 22 a lot more quickly. 23 SIR WYN WILLIAMS: Yes, I would prefer to complete 24 the evidence without a further break, if I can 25 put it in that way and I think we ought to be 109 1 able to do it. 2 MR HENRY: So be it, sir. 3 THE WITNESS: Okay, hang on, sorry? Can I say 4 something? I would like a break. 5 SIR WYN WILLIAMS: Do you mean a short break? 6 THE WITNESS: Yeah, just a natural break to go to 7 the loo, that's all. 8 SIR WYN WILLIAMS: Of course. Of course, you must 9 have that. Right. 10 MR HENRY: Could we have a break now, then, sir, to 11 accommodate Mr Holmes, perhaps? 12 SIR WYN WILLIAMS: Yes. We will break off for what 13 I will call a loo break. We'll bring us back at 14 1.00. 15 MR HENRY: Sir, could I ask for a little bit more 16 time than that? Just a little bit more. 17 Because I did have some material and, in 18 response to your request for, as it were, 19 speeding things along, I would just like to have 20 a little bit more time to make things more 21 concise. 22 SIR WYN WILLIAMS: Right. Well, what I'm going to 23 do is say this: we're going to break until 1.10. 24 MR HENRY: Thank you, sir. 25 SIR WYN WILLIAMS: But then the questioning will 110 1 cease no later than 1.30. 2 MR HENRY: So be it, sir. 3 SIR WYN WILLIAMS: Fine. 4 (12.53 pm) 5 (A short break) 6 (1.11 pm) 7 MR BLAKE: Thank you, sir. We can see and hear you. 8 Can you see and hear us? 9 SIR WYN WILLIAMS: Yes, I can, thank you. 10 MR HENRY: May I begin, sir? 11 SIR WYN WILLIAMS: Of course, yes. 12 MR HENRY: Thank you, sir. 13 Mr Holmes, the report that you co-authored 14 with Mr McDonnell, I asked you on a previous 15 occasion why you retrieved it on 14 May 2001. 16 Do you recall me asking you that question? 17 A. Yes, I believe you did, yes. 18 Q. Sorry, sir. I can't hear you. 19 A. I'm unmuted. Yeah, I think you did. 20 Q. I'm afraid we've lost the sound. 21 A. No, I am unmuted. 22 SIR WYN WILLIAMS: I can hear Mr Holmes, so it's 23 a problem in the hall, by the sound of it. 24 MR HENRY: Yes. 25 A. Can you hear me now? 111 1 Q. Oh, I can hear you now, Mr Holmes. Thank you. 2 Did you give an answer, because we didn't hear 3 it? 4 A. I gave an answer to the question you just asked 5 which was, yes, I do recall you asking the 6 question in November. 7 Q. Yes, and your answer was that it was a sort of 8 an administrative catch-up? 9 A. To the best of my memory, yes. 10 Q. Yes. Now, can I just ask you please, at the 11 time, had you been asked to assist in the 12 prosecution of Ms Tracy Felstead to the best of 13 your knowledge and belief? 14 A. No, which -- what was the name of the outlet she 15 was associated with? 16 Q. Camberwell Green. 17 A. Right. Now, I do believe I did a witness 18 statement for Camberwell Green but with no 19 knowledge of who was involved. 20 Q. You certainly did make a witness statement for 21 Camberwell Green and I'm sure that's right 22 because you do not mention Ms Felstead's name. 23 You were asked to consider data between 24 12 October 2000 and 1 March 2001. 25 A. When you say "consider", do you mean review it, 112 1 analyse it or extract it? 2 Q. Shall we go to your witness statement -- 3 A. Yeah, let's do that. 4 Q. -- that will probably be the best thing to do. 5 A. I can't get my hands on it immediately here. So 6 yeah. 7 Q. Don't worry because we can have it put up on the 8 screen. 9 A. Okay. 10 Q. It's WITN04600217. I wonder if you could put 11 that up. 12 Now you, sir, are not a technical person. 13 A. No. 14 Q. You have said that on more than one occasion. 15 A. Yeah. 16 Q. Right. This is not signed, but you have 17 mentioned that you were asked to make a witness 18 statement. Did you eventually sign a witness 19 statement? 20 A. I don't know. 21 Q. I suppose that would be the natural presumption 22 of going to the trouble of writing a draft 23 witness statement, a draft witness statement, of 24 course, which has been amended, as we can see. 25 A. Well, I was going to say, this is still in draft 113 1 form, isn't it? 2 Q. Yes. I mean, were you in the habit, as you 3 were, for example, in the Cleveleys case, of 4 making witness statements? 5 A. I made a few over the years, yes. 6 Q. In fact, I think in some of the documents that 7 you very kindly produced, you were remarking 8 about the fact that you were being asked to make 9 a few about various locations at one point and 10 that -- I'm not sure of the exact words -- but 11 that it was a bit of an imposition; do you 12 recall? 13 A. Well, they take time to produce, but it's part 14 and parcel of the job, isn't it? 15 Q. Yes, it is. You presumably read the declaration 16 that it was going to be, once signed, true to 17 the best of your knowledge and belief, and you 18 would make it knowing that if it were to be 19 tendered in evidence you would be liable to 20 prosecution if you had wilfully stated in it 21 anything which you knew to be false or did not 22 believe to be true. So, obviously, you realised 23 it was a matter of some solemnity and 24 importance. 25 A. I'm just concerned that I don't have a final 114 1 copy of this, in the sense that all I've got is 2 a copy that you've got, which is in a draft 3 form, and I can't -- unless I can see one that 4 I've signed, I don't know what to say. 5 Q. Well, I'm not going to be asking you about any 6 of the bits that have been amended. So all I am 7 going to be asking you about is your direct 8 knowledge of the contents of this statement. 9 A. Okay. 10 Q. Now -- 11 SIR WYN WILLIAMS: Before we go any further, 12 Mr Henry, will we be able to establish today 13 whether or not Mr Holmes actually made a signed 14 witness statement? 15 MR HENRY: I'm afraid, given the effluxion of time, 16 sir, unless there's something that I have not 17 yet detected, I am afraid I can't answer that 18 question. 19 SIR WYN WILLIAMS: So, I mean, obviously I will 20 defer to your view as to whether you pursue this 21 at the moment but I'm not quite sure what I will 22 get in terms of my investigation from 23 an examination of a draft which may never have 24 become more than a draft, if I can put it in 25 that way. 115 1 MR HENRY: Could I go to some subsidiary documents, 2 then, if I may. Could we go -- we'll come back 3 to this, but I want to be quite fair to you, 4 Mr Holmes. Could we go to WITN04600216. This 5 is a "Pathway Change Control Notice (CCN) -- 6 Sheet". It's a document that you produced, 7 Mr Holmes. It came from your garage or wherever 8 you were keeping these documents. 9 The date is 10 January 2002 and it was again 10 under the umbrella of the Camberwell Green Post 11 Office case. Could we go, please, to page 4 of 12 5. Do you remember on the last occasion, 13 Mr Holmes -- and I can give the precise 14 reference if you would wish to go to it, but let 15 me ask the question first -- that I put to you 16 that at Ms Felstead's trial, a request had been 17 made by ICL Fujitsu for £20,000 for unused 18 material, and do you see the figure there? 19 A. Yes, I do. 20 Q. It's remarkably similar, isn't it, to the 21 £20,000 in relation to a Camberwell Green case. 22 Do you agree? 23 A. Well, yeah, the numbers are roughly the same but 24 I have no knowledge of this at all. 25 Q. But these are your documents that you produced 116 1 to the Inquiry. You obviously kept them for 2 a reason. 3 A. I didn't produce the CCN. That's not my 4 document that I'm looking at now. 5 Q. Right. I'm very sorry, I thought these were 6 documents that were actually produced by you as 7 a result of being requested to produce them by 8 the Inquiry. 9 A. Can you scroll this one on screen up to the top, 10 and let's have a look at what it is. Right this 11 is a Change Control Note which is raised by us 12 in response to a change request that is raised 13 by the Post Office, which we then raise a change 14 proposal that effectively establishes the work 15 within Post Office Account. And the CCN is the 16 Change Control Note that goes back to the Post 17 Office to say, "This is what we're going to do 18 and this is how much it's going to cost". 19 I have nothing to do with that. In fact, 20 you can see the CCN was raised by Graham Hooper, 21 who was a Security Manager at that time. 22 Q. I see. Nevertheless, in relation to Camberwell 23 Green, because this is what we understand and 24 that's the description of the document as to its 25 provenance, in relation to the Camberwell Green 117 1 case, it's remarkably close, is it not, to the 2 £20,000 that I put out to you on the last 3 occasion? 4 A. Yeah, but I -- you can't expect me to comment on 5 something that I had no part in the production 6 of. Yes, the numbers are nearly the same but so 7 what? 8 Q. Could I just again -- I appreciate that you do 9 not refer to Ms Felstead by name in your witness 10 statement but could we go back to that draft 11 witness statement, please, and that's 12 WITN04600217. Again, just the introduction: 13 "I have been employed by ICL Pathway for 14 5 years ... employed as the Quality and Audit 15 manager responsible for Quality and Audit 16 Management and the User Authority for the Audit 17 Solution from where the Audit Data is sourced. 18 I have working knowledge of the computer system 19 known as Horizon, which is the computer supplied 20 by ICL Pathway Limited and used by Post Office 21 Limited in Post Office Outlets. I am authorised 22 by ICL Pathway Limited to undertake extractions 23 and analyses of audit that held on the Horizon 24 System." 25 Is all of that completely accurate? 118 1 A. Yes. 2 Q. Could we go to, please, if we may, to the 3 conclusion of the witness statement, please, or 4 page 4 of 5. Forgive me, page 4 of 5. Could 5 I ask you, please, to scroll down. I do 6 apologise. 7 You're dealing here with Horizon System 8 Helpdesk calls, Tivoli event logs, non-polling 9 reports. So far as you are aware, was this 10 information that was being supplied to you -- 11 you not being a technical person -- that this 12 was information being supplied to you by 13 somebody else? 14 A. I don't know, or these might have been audit 15 data that we'd extracted and then had a look at. 16 So I can't tell you, if the Horizon System 17 Helpdesk calls data, the 21 calls, were 18 extracted from the audit solution and then 19 I looked at it or whether it was given to me by 20 somebody else. 21 Q. I mean, obviously, a witness statement in 22 a civil case can often be an amalgam of 23 information that is relayed to one by other 24 people and that it is sometimes even constructed 25 by the solicitor on behalf of the witness, but 119 1 this was a statement that was intended for 2 submission in a criminal case. Do you follow? 3 A. Well, actually, if you go back to the previous 4 page, I'm putting my hand up there and saying: 5 "On various dates and at various times 6 between 14 January and 25 January I analysed 7 audit data for transactions and activities 8 undertaken by Camberwell Green Post Office 9 during the period 12 October 2000 to 1 March 10 2001 (the Material Period). 11 Q. Yes. 12 A. I have to assume that what follows, Horizon 13 System Helpdesk, Tivoli event logs, non-polling 14 reports, were pulled from the audit archive. 15 I mean, I can't remember, 21 years on. 16 Q. Of course. So, in other words, you were, 17 although not a technical person, you were 18 analysing the data. Was anybody there to assist 19 you, given the fact that you weren't a technical 20 person, in making an assessment of the data? 21 A. I can't remember whether there was or not, to be 22 honest with you. 23 Q. I see. Then could we go to the next page, 24 please, and the "Conclusion"? 25 A. Can I just stop you there, Mr Henry? 120 1 Q. Of course, of course. 2 A. I'm looking at a version of this document, which 3 is version 1, which has the changes in the same 4 way that you've got on yours, but I've also got 5 another -- oh, I must be imagining it. 6 I thought there was something there about time. 7 No, right. Forget it, sorry. I beg your 8 pardon. 9 Q. Don't worry, Mr Holmes. 10 Your "Conclusion": 11 "There are no reasonable grounds for 12 believing that the information stored on the 13 Horizon system would be inaccurate because of 14 improper use of computer terminal. During the 15 Material Time the Horizon system was operating 16 properly at the Camberwell Green Post Office 17 Outlet or if not, any respect in which it was 18 not operating properly or was out of operation 19 was not such as to affect the production of 20 audit records or accuracy of their contents." 21 A. Mm-hm. 22 Q. Now, sir, are those your words or were they 23 effectively a rubric that you had been provided 24 with? 25 A. I can't recall. I cannot recall. 121 1 Q. Do you think they fairly reflect the knowledge, 2 your personal knowledge, that learned counsel to 3 the Inquiry was exploring with you this morning, 4 about bugs, errors and defects which were 5 apparent at this time? 6 A. Sorry, can you repeat that? 7 Q. Do you think that this fairly reflects the 8 knowledge, your personal knowledge, of bugs, 9 errors and defects that were apparent at this 10 time? 11 A. Well, I guess it must do because those did exist 12 at that time, we all know that now. We didn't 13 necessarily know it at the time. 14 Q. Well, you were aware, were you not, of 15 difficulties in retrieving data and difficulties 16 with, as it were, obtaining the information that 17 was required for these prosecutions, correct? 18 A. Yeah, well, that wasn't because the data didn't 19 exist or it was invalid or corrupt. It was 20 because we didn't have an Attribute Grammar 21 catalogue to help us build the SQL statements. 22 Q. You were aware of balancing errors that had been 23 drawn to your attention at the time? 24 A. If you say so. 25 Q. You were aware that the EPOSS code was of poor 122 1 quality, unstable and vulnerable to code decay. 2 You were aware of all these things? 3 A. I accept that one, yeah. 4 Q. Yeah. So, again, I ask you, do you think that 5 this is a fair reflection of that which was 6 known at that stage by you but also Fujitsu? 7 A. Yes. I do. 8 Q. I suggest to you, Mr Holmes, that that can't be 9 right. 10 A. Well, all right, fine. If you believe that to 11 be incorrect, perhaps you can enlighten me as to 12 how you arrive at that -- 13 SIR WYN WILLIAMS: Hang on now, hang on. 14 First of all, it's 1.30 and, secondly, this 15 session was supposed to be about Cleveleys and 16 the relevance -- sorry, the evidence relating to 17 Cleveleys. We're now going off on a completely 18 different tangent in circumstances about which 19 I am uncomfortable, when it is not even known 20 whether or not an actual witness statement was 21 made. 22 MR HENRY: So be it, sir. The reason why I have 23 been asking this is because it appears to be in 24 lock step with Cleveleys in, as it were, 25 a defence of the Horizon System, and Ms Felstead 123 1 was suspended in February 2001, so -- 2 SIR WYN WILLIAMS: I understand how it could be 3 extremely relevant to Ms Felstead's case but not 4 the Cleveleys case and I want to be clear in my 5 mind as to where we are day by day, so to speak. 6 MR HENRY: I see, sir. I'm very, very grateful to 7 Counsel to the Inquiry, who understood from our 8 representations that we were trying to, as it 9 were, draw a parallel with Cleveleys, in that 10 this being a defence of the Horizon System at or 11 about the same time. But that is my apology, in 12 its old-fashioned sense, to the Inquiry. 13 SIR WYN WILLIAMS: That's fine. That's fine. And 14 if there comes a point in time when there is 15 reasonable grounds to suspect -- I think I put 16 it as low as that -- that a witness statement 17 was made in Ms Felstead's case, I'm not saying 18 I won't return to that paragraph, Mr Henry. But 19 not at the moment. 20 MR HENRY: So be it, sir. 21 Mr Holmes, thank you very much. 22 SIR WYN WILLIAMS: Thank you, Mr Holmes, for giving 23 evidence for a second time and for giving in 24 total now three witness statements to the 25 Inquiry. 124 1 That concludes the Inquiry's business for 2 today. We're now going to embark upon 3 a vacation period, where no doubt many of the 4 people involved in the Inquiry, if not all of 5 them, will be taking a well-earned break. 6 I wish everyone a good holiday, if they're 7 having one and I will see you all on 8 5 September. Thank you. 9 THE WITNESS: Thank you, Sir Wyn. 10 (1.34 pm) 11 (The hearing adjourned until 5 September 2023) 12 13 14 15 16 17 18 19 20 21 22 23 24 25 125 1 I N D E X 2 JAN ROBERT HOLMES (affirmed) ..................1 3 Questioned by MR BLAKE ........................1 4 Questioned by MR JACOBS ......................87 5 Questions by MR MOLONEY ......................92 6 Questioned by MR HENRY .......................98 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 126