Monday, March 18, 2019

Miscarriages of justice

Blue screen of potential prosecution
At the very heart of this story, beyond the group litigation, is the possibility that some people were prosecuted by the Post Office when they were wholly innocent.

The Post Office has successfully criminalised dozens of Subpostmasters purely on computer evidence generated by its Horizon IT system (illustrated above).

On Monday, an almost mythical document was revealed in court by the claimants' QC, Patrick Green. It was during his cross-examination of the Post Office director Angela van den Bogerd.

It is the Rose report, dated 12 June 2013.

Helen Rose is a fraud analyst called in at great expense by a Subpostmaster in Lepton who was being accused by the Post Office of a (possibly fraudulent) user error. He was pretty sure it was a Horizon IT error. Ms Rose was known to the Post Office, and if she hadn't worked for them as an independent expert before, was certainly familiar with Horizon.

The sum at stake - £76.09. Yup. That's all.

The matter at stake - the Subpostmasters' integrity.

What happened?

Ms Rose's report revealed that really crucial information, which Post Office prosecutors relied on when taking Subpostmasters to court, might not be accurate.

As Mr Green explained:
"A transaction took place at Lepton sub-post office [in 2012?] for a British Telecom bill payment for £76.09 ; this was paid for by a Lloyds TSB cash withdrawal for £80 and change a given for £3.91.  At 10.37 on the same day the British Telecom bill payment was reversed out to cash settlement."
This is was suspicious. Certainly the Post Office thought a bill payment being reversed out to cash was odd. A transaction correction came down and the Subpostmaster was forced to settle the discrepancy out of his own pocket. Things escalated. He brought in Ms Rose: "as he believed his reputation was in doubt."

The Credence data - software which logs every keystroke on Horizon - made it clear the transaction reversal had been carried out by someone using the Subpostmaster's login. Credence data is very useful evidence.

Ms Rose also requested the ARQ logs - back end data from Horizon which will often give investigators a fuller picture of what was going on at a branch. Whilst waiting for the ARQ data, the redoubtable Ms Rose contacts a senior Fujitsu engineer, Gareth Jenkins (who was the Post Office's IT expert at Seema Misra's trial).

Mr Jenkins, being a helpful sort of fellow, goes and has a look at the live data (which informs the ARQ logs) about this particular issue. He noted that although Credence (which logged the keystrokes) made it clear that someone on the Subpostmaster's login had reversed the cash withdrawal, the live data showed it was a recovery reversal initiated by Horizon after a power failure. According to Mr Jenkins:
"It isn't clear what failed, but if it was a comms error, then the system would have printed a disconnected session receipt and the clerk should have given the customer £80 and told him his bill was unpaid.  The fact that there is no indication of such a receipt in the events table suggests the counter may have been rebooted and so perhaps may have crashed in which case the clerk may not have been told exactly what to do."
Mr Jenkins adds:
"The reversal was due to recovery so this was not an explicit reversal by the clerk. This scenario is fairly rare so it is certainly quite easy for the clerk to have made a mistake and either he or the customer could be in pocket/out-of-pocket (depending on exactly what happened!)."
Then he says: "The system is behaving as it should."

Ms Rose replied: "I can clearly see the recovery reversal on the Fujitsu logs received, but would this have been clear had we not previously discussed this issue?"

Mr Jenkins' answer: "Note that the standard ARQ spreadsheet may not make it easy to confirm that the reversal was part of recovery, but the underlying logs used to extract them can show it."

Woah.

To add another dimension to this drama - we are hearing it discussed it in court in the context of Angela van den Bogerd's evidence. Mrs van den Bogerd says (in her witness statement in paragraph 154) that the claimants' IT expert for the Horizon trial, Jason Coyne, has taken Ms Rose's report out of context. Ms van den Bogerd says Mr Coyne has "mistakenly claim[ed] that the relevant reversal was issued in error by Horizon, not the Subpostmaster."

Mr Green responds:
"Well, we can see from what Gareth Jenkins has actually said in the actual document you are looking at, that that's wrong, can't we. The reversal was not done by the Subpostmaster, it was done by the system."
Mrs van den Bogerd's witness statement was written before she was found to have misled the court in the previous, Common Issues trial. I have no idea if her witness statement was another attempt to mislead the court, but when challenged on what she said she immediately performed what Private Eye call a "reverse ferret". She told the court:

"It was done by the system absolutely, yes."

Mr Green QC asks for confirmation the ferret is being reversed: "It was done by the system and not by the Subpostmaster?"

Mrs van den Bogerd, perhaps not wishing to mislead the court again, directly contradicts her own witness statement: "Yes, it was done by the system, yes."

Mr Green rubs it in a bit: "So it's fair to say that what we have at paragraph 154 is wrong, isn't it?"

Mrs van den Bogerd confirms her witness statement is wrong:
"It wasn't issued in error, it was actually issued by Horizon. So I am obviously not be making myself clear, but yes, there's no question that that was done as part of that recovery and it was system generated."
The incomplete picture

But let's not get sidetracked. Let's go back to the Rose report. Ms Rose has a question for Mr Jenkins:
"I can see where this transaction is and now understand the reason behind it.  My main concern is that we use the basic ARQ logs for evidence in court and if we don't know what extra reports to ask for then in some circumstances we would not be giving a true picture."
She continues:
"I know you are aware of all the Horizon integrity issues ... [none of which Mrs van den Bogerd could remember when asked in court]
... and I want to ensure that the ARQ logs are used
 and understood fully by our operational team who have to
 work with this data both in interviews and in court."

Mr Jenkins agrees:
"I understand your concerns. It would be relatively simple to add an extra column into the existing ARQ report spreadsheet, that would make it clear whether the reversal basket was generated by recovery or not. I think this would address your concern.  I'm not sure what the formal process is for changing the report layout."
In court, on Monday 18 March 2019, Mrs van den Bogerd was asked if this relatively simple change, first flagged nearly six years ago, had been effected.

Mrs van den Bogerd: "I don't believe it has been acted on."
Mr Green QC: "You don't believe it has been?"
Mrs van den Bogerd: "I don't believe so."

I don't think I need to spell out the implications.

Read the Day 5 transcript here.
Read the other revelations about serious Horizon errors which came out on Day 5 here.

*******************

Please feel free to forward the link to this blog post. The more people who read it, the more people find out about what is the biggest trial going through the UK courts right now.

Please also consider a donation to keep the postofficetrial.com website going by choosing an amount and clicking the button at the bottom of this post. Contributors who donate £20 or more will be added to this secret email list. Thanks!


Choose an amount

Horizon trial: day 5 - Angela van den Bogerd witness statement

Angela van den Bogerd was the first witness to be called for the Post Office during the Horizon trial. She was cross-examined on 18 March 2019 and 19 March 2019. You can read Mrs van den Bogerd's witness statement on scribd here or embedded below.

Horizon trial day 5: Into the Horizone

Angela van den Bogerd arriving at the High Court this morning
Angela van den Bogerd was the in the witness box all day. She is the Post Office's business improvement director. On Friday the judge ruled that during the Common Issues trial, Mrs van den Bogerd had deliberately tried to mislead the court on more than one occasion. He added:

"Unless I state to the contrary, I would only accept the evidence of Mrs Van den Bogerd and Mr Beal [a senior Post Office colleague] in controversial areas of fact in issue in this Common Issue [sic] trial if these are clearly and uncontrovertibly corroborated by contemporaneous documents."

So we were all ears today. And there were all sorts of revelations.  

The MoneyGram duplicate transactions error

There's no doubt Horizon appears to work well for most Subpostmasters most of the time. But that's no comfort to a Subpostmaster on the end of a Horizon error. The Post Office's case is that yes Horizon has errors and yes a small number of them might affect branch accounts, but because of the systems in place and information largely only available to the Post Office, no claimant Subpostmasters have ever had to give money (much less be accused of stealing it) to the Post Office because of a computer error. It would only ever be a user error or fraud. The claimants dispute this strongly, but they have to prove it, in this trial, beyond the balance of probabilities to the judge.

Patrick Green QC
This afternoon, Patrick Green, QC for the claimants, produced a loaded gun and then for some reason, didn't fire it.

The case is that of Mr Aakash Patny, a claimant who gave evidence last week. Mr Patny had a number of problems with Horizon, all hotly disputed by the Post Office, and one concerned a MoneyGram reversal which Mr Patny performed on 23 February 2016 when a customer wished to send £3,100 abroad, but could not do so when his debit card was declined. What Mr Patny did was cancel the transaction, but in fact the Post Office claims he should have reversed it by calling the Horizon helpline and being taken through the steps to reverse it. 

At the end of the day, Mr Patny found he was £6,200 down. He called the helpline and they reversed his £3,100, but he was held liable for the remaining £3,100, which he disputes. 

Mr Patny's cross-examination on Wednesday last week was a little confusing. I think it was because the Post Office barrister Simon Henderson didn't seem ready for Mr Patny's explanations of what happened - and, if I might be so bold, seemed a little unclear as to what actually had happened.

Mr Henderson insisted that Mr Patny had followed the wrong procedure in not calling the Helpline to reverse the transaction at the time and he produced a document describing the correct procedure which he asked Mr Patny to agree with. 

Mr Henderson then said that Mr Patny called the helpline but did not perform the reversal. Mr Patny said he did - he called the helpline after something called "polling time" which is 7pm, when Horizon rolls over into the next working day. He called after 7pm because until that point he thought he had successfully cancelled the MoneyGram transaction, but then after balancing, he noticed he was £6,200 down. Mr Patny thought this must be the declined transaction which had somehow magically doubled. Mr Patny then said that the Helpline agreed he had a £3,100 reversal which needed performing, but he was then held liable for the other £3,100.

After a brief weird interlude during which Mr Henderson accused Mr Patny of committing a criminal offence over a ludicrous stamp declaration (which Mr Patny said was another Horizon error) we returned to the MoneyGram issue. Mr Henderson said the extra £3,100 was actually £3,694.88 and it was down to a variety of discrepancies caused by user error (quite why he thought the £3,100 couldn't be hidden within that larger number I don't know, but Mr Patny didn't challenge him on it). 

Mr Patny was having none of it. He said his area sales manager Mark Irwin came to have a look agreed it was probably a double transaction and actually lobbied the Post Office on his behalf to stop his branch being held liable.

And that was how we left it on Wednesday.

So let's go to today. Patrick Green established with Angela van den Bogerd that the very month Aakash Patny had a problem with his alleged double phantom MoneyGram transaction, there was a problem in the system causing double phantom errors.

The error was caused by a software update from MoneyGram in January 2016 causing "system latency". In MoneyGram's quarterly business review, the error is reviewed and readers are advised 

"Duplicate transactions are created in [MoneyGram] systems as a result of Post Office time outs. A joint MG/POL team is working to ascertain impact on settlement and implementation of an appropriate plan of action."

I thought Mr Green was going to suggest that here we had the evidence that a software error affecting Post Office systems was causing duplicate transactions identical to the one Mr Patny was being held liable for. But he didn't instead he said:

   Q.  It wasn't wildly well advertised to subpostmasters that
       this problem was being dealt with at the time?
   A.  No.  People were aware there was an issue, clearly, in
       branch and then --
   Q.  People in the branch had an issue, but they were
       probably being blamed for user error, weren't they?
   A.  Well, they knew there was an issue and they were ringing
       into NBSC at the time.
   Q.  Sorry?
   A.  They knew there was an issue and they were ringing into
       NBSC at the time.
   Q.  Well they were claiming there was an issue but they
       might be met with "This is user error"?

This to me looked like it was building up to the first clear cut copper-bottomed evidence of a Horizon error which had left a branch Subpostmaster out of pocket, but Mr Green didn't ask Mrs van den Bogerd if she thought that was a possibility. He didn't pull the trigger. I wonder why.

Shonky hardware, shonky software

In the morning session, there were two belters. In fact, during one of them, Karl Flinders from Computer Weekly leaned over to me and whispered "he's just put a 35 yarder in the top right hand corner", which, for those not familiar with sporting idioms, is an Association Football reference meaning "he's done quite well there". 

First of all Mrs van den Bogerd is taken to paragraph 13 of her witness statement in which she says she has been made aware of "phantom sales" in Horizon in around 2000, but, crucially, says Mrs van den Bogerd, "they should not cause a discrepancy in a branch's accounts". She then goes on to list how certain mysterious out of hours transactions take place which are either down to user error, standard Horizon processes or fraud.

Mr Green picks up on this and says that if Mrs van den Bogerd was told about phantom sales by a Subpostmaster, she'd assume it was user error. Mrs van den Bogerd agrees.

Mr Green takes Mrs van den Bogerd to something called a master PEAK created in 2001. This is a document "where there are a number of iterations of a problem what Fujitsu do is they pull together a number of PEAKs into a master PEAK to collect the cases of different people who have had the same problem."

It is about a Horizon issue which has been closed without the permission of the Subpostmaster, who is understandably irate. He is £1500 down, he has had to make good the amount and he is now at risk of suspension for refusing to make ongoing discrepancies good. The PEAK notes the Subpostmaster "wants a face-to-face meeting with someone in
 authority from [Horizon/Post Office] to discuss the issues. [The Subpostmaster] feels very strongly about this and says he is willing
 to take POCL [Post Office Counters Ltd, as was] to a tribunal/court because of the stress he has suffered because of the problems."

So - stressed out Subpostmaster, lots of money at stake and he's hacked off because the complaint was erroneously closed. Mr Green notes that the Subpostmaster had been told it was probably user error. 

A month later some Royal Mail engineers go out to have a look (Post Office was still part of Royal Mail in 2001). They see the phantom transactions for themselves.

"Now, pausing there," says Mr Green, "that is independent site visit
 corroboration of the problem by Royal Mail's own engineers at the branch, isn't it?
"

Mrs van den Bogerd agrees. 

Eventually after lots and lots of to-ing and fro-ing with several Subpostmasters all appearing to have the same problem, it turns out it was a fault with the touch screens which were creating the phantom entries.

Mrs van den Bogerd agrees this is clearly a hardware error affecting multiple branches.

But we see from the PEAK that the official resolution is to refuse to acknowledge the problem as a Horizon error as user error could not be precluded from it. So the stressed out Subpostmaster presumably got a new touch screen, but he didn't get his money back.

Mr Green then leaped forward 10 years to a software error reported in another PEAK dated February 2011 which showed negative stock items which the Post Office branch in question didn't even sell.
"The office went into the declaration and confirmed
 it showed the PO saving stamps and other stock items,
 some showing as minus stock amounts, that were not
 correct to the branch.... it somehow seems that the system has somehow picked
 up this declaration and this is the cause of the
 discrepancies appearing on the system.  The SPMR was
 told to declare the correct stock figures but is
 reluctant to do this as this will cause discrepancies
 when she next balances that are not relevant to herself."

Serious stuff.

Mrs van den Bogerd agreed this was an example of a software error on Horizon.

Let's end with an internal Post Office email exchange!

The first email is dated 9 Dec 2013, and is from the Helpline team to the branch support team (which Mrs van den Bogerd was part of). The helpline team email starts: 

"Branch reporting that he has found sensitive issue with Horizon when the system put a phantom cheque on the cheque line in July 2013.  Claims to have evidence to support his claim.... Although he himself did not suffer a loss, thinks that Horizon is flawed.  Did not ask to be contacted about this. Just wanted to say that he had this information and threatened to go to MP as a result."

Nigel Allen from the branch support team sends it up to Andy Winn saying: "Given the current media and in particular the BBC's attention on Horizon, do you think it is worthwhile looking into this 'alleged flaw' with Horizon that this SPMR has highlighted to preempt any enquiries from his MP?"

The response comes back from Mr Winn:

"There is nothing I can investigate given the level of detail provided unless he only accepted one cheque in July 13.  Even then I don't have the level of detail needed and would need Fujitsu support.  Without a date/value we can't really raise a request. I don't really understand what the purpose of the call is.  Does he want it investigated or not?  My instinct is that we have enough on with people asking us to look at things. I can't figure out how if a phantom cheque appeared on Horizon he could avoid a loss unless another phantom transaction took it away again!!!!"

Interesting. Mrs van den Bogerd is asked what she thinks of this email exchange. Would it not be useful to be a little bit curious about a potential Horizon error? Maybe ask for the evidence? Mrs van den Bogerd agrees this is a "totally inadequate response." Mr Green says " And the point about that is that shows, red in tooth and claw, that if a phantom cheque comes up into your account you're going to suffer a loss, aren't you?"

Mrs van den Bogerd agrees that is what Mr Winn is saying.

Mrs van den Bogerd continues to give evidence tomorrow.

*****************************
Please feel free to forward the link to this blog post. The more people who read it, the more people find out about what is the biggest trial going through the UK courts right now.

Please also consider a donation to keep the postofficetrial.com website going by choosing an amount and clicking the button at the bottom of this post. Contributors who donate £20 or more will be added to this secret email list. Thanks!


Choose an amount

Horizon Trial: Day 5 transcript

This is the unperfected transcript of the fifth day of the Horizon trial, held on 18 March 2019 in court 26 of the High Court's Rolls Building. It is the first day of cross-examination of Angela van den Bogerd, a witness found to have deliberately misled the court during her evidence at the Common Issues trial.

                                          Monday, 18 March 2019
   (10.30 am)
                           Housekeeping
   MR DE GARR ROBINSON:  My Lord, good morning.  It is my turn
       to be leading evidence.  There are a few housekeeping
       matters for me to raise with your Lordship and there
       might be some your Lordship would like to raise with me.
           My points for your Lordship.  On Day 2 Mr Latif and
       Mr Tank gave evidence.  Your Lord will recall that
       Mr Latif referred to some memo views that he received in
       January 2018.  My Lord, those have now been found and
       disclosed.
           My Lord, secondly in relation to Mr Latif there was
       a reference to an audit that was done at his branch, the
       Caddington branch.  That audit report has now been
       disclosed and indeed it is now in the trial bundles.
   MR JUSTICE FRASER:  Are any of the memo views in the trial
       bundles?
   MR DE GARR ROBINSON:  Not yet, my Lord.  They were found and
       disclosed over the weekend.
   MR JUSTICE FRASER:  Are they going in, do you know?
   MR DE GARR ROBINSON:  Yes, my expectation is that they will
       go in unless --
   MR JUSTICE FRASER:  I in particular would like to look for
       and look at at least one or two of them, so ...
   MR DE GARR ROBINSON:  My Lord, yes.
           My Lord, Mr Tank, your Lordship will recall that he
       was shown a Horizon Online quick reference guide and
       suggested he had never seen a document like that before,
       the version he was used to was much shorter.  The
       reference guide which was in force at the time of the
       events in question had already been disclosed but it is
       now in the bundle, my Lord, at {F/691.1}.
   MR JUSTICE FRASER:  And that's the one that was in force at
       the time Mr Tank --
   MR DE GARR ROBINSON:  Yes.  It is dated 11 August -- I say
       dated, when it was disclosed it was disclosed with the
       date 11 August 2010.
   MR JUSTICE FRASER:  And that I think related to a passage of
       evidence where I observed that the court had seen
       a different version of the reference guide.
   MR DE GARR ROBINSON:  Yes, I think your Lordship had seen
       perhaps a shorter one in the common issues trial.  This
       is the long version.  There is no real, as far as I can
       see, substantive difference between the one that was put
       to Mr Latif and the one that's now in the bundles.
   MR JUSTICE FRASER:  Does that deal with all your
       housekeeping?
   MR DE GARR ROBINSON:  Those are mine, my Lord.
   MR JUSTICE FRASER:  Mr Green, do you have any?
   MR GREEN:  My Lord, there is the Royal Mail application in
       the background but I don't want to trouble your Lordship
       with it now.
   MR JUSTICE FRASER:  That is something on my list and I would
       like to raise it now, not in respect of that but in
       respect of the correspondence at the end of last week.
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  Mr De Garr Robinson, obviously you know
       about this because you communicated it to the court as
       soon as you found out.
   MR DE GARR ROBINSON:  My Lord, yes.
   MR JUSTICE FRASER:  My understanding and do correct me if
       I'm wrong but I'm pretty sure that this is correct, is
       that what the court was told on Thursday about the
       Royal Mail have the E&Y pre-2011 audit reports and being
       unwilling to disclose them was incorrect, is that
       a correct understanding?
   MR DE GARR ROBINSON:  My Lord, yes, subject to this
       important clarification.  My Lord, there had been --
       there's a confusion on my side of the court.  Audit
       related documents had been requested from Royal Mail and
       Royal Mail had indicated that they were reluctant to
       provide anything without the protection of a court
       order.  Then the confusion was that it was thought that
       the audit related documents that had been requested and
       had met with that response were the actual audit reports
       prior to 2011 and that was the mistake and that was not
       the case.  There had not actually been a discussion with
       Royal Mail about those specific documents, which is why
       it's only right that my instructing solicitors have
       apologised to Freeths and I have apologised to
       your Lordship.
   MR JUSTICE FRASER:  Yes.  The situation vis-à-vis the
       Royal Mail seems to me to fall into two categories, one
       of which we can come on to deal with later, the other
       one of which I want to deal with now.
           The part we can come on to deal with later relates
       to the third party disclosure application that I made an
       order in respect of on Thursday and you have now
       confirmed my understanding.  There are some documents
       that the Royal Mail has relating to audit in respect of
       which an order may prove to be necessary.  That
       tributary can meander along or not in line with the
       proper procedure in the CPR which was initiated at the
       end of last week.  I assume that that application was
       served in accordance with my order.
   MR GREEN:  My Lord, yes.  I think the deadline for service
       is today.
   MR JUSTICE FRASER:  The second point is the actual pre-2011
       Ernst & Young reports.  My updated understanding of that
       derives solely from what I was told last week, the
       letter of correction from Freeths on 15 March and what
       you have told me and it seems to me that the position is
       as follows, that the Post Office still has not disclosed
       those reports because it does not believe it has got
       them, is that correct?
   MR DE GARR ROBINSON:  Yes, it hasn't been able to find them.
       It has looked for them and not been able to find them,
       my Lord.
   MR JUSTICE FRASER:  I am therefore going to order a witness
       statement please from your solicitors and I'm just going
       to identify what that witness statement has to deal
       with, because at the moment it seems know there is a bit
       too much generalism in these letters of explanation that
       say "The Post Office believed" or "The Post Office asked
       for", so I'm going to condescend to particulars.  I'm
       just going to read out what the witness statement -- and
       we will come on to the time and date that this has to be
       served at the end of this little passage.
           I would like a witness statement from the
       Post Office, and I imagine that will be from Mr Parsons,
       which deals with the following: 1, the steps taken by
       reference to the specific individuals who have
       instructed or directed those steps, in order to obtain
       copies of the pre-2011 Ernst & Young reports; 2, if
       those steps have led to a conclusion that the
       Post Office does not have those reports in its custody,
       possession or control, an explanation as to how that can
       be the case; 3, identification of what requests, if any,
       have been made to Ernst & Young before today's date --
   MR DE GARR ROBINSON:  Does your Lordship mean Royal Mail?
   MR JUSTICE FRASER:  Well, I have just said identification of
       what requests have been made by the Post Office --
   MR DE GARR ROBINSON:  To?
   MR JUSTICE FRASER:  To Ernst & Young, who produced them,
       before today to obtain copies of those reports.
           4, exactly the same in respect of Royal Mail, in
       other words identification of what requests have been
       made to Royal Mail.  5, if no such requests have been
       made either to Ernst & Young and/or Royal Mail, an
       explanation of how that comes to be the case.  And then
       6, which is effectively the catch-all, which is a copy
       please of an immediate request that will be made today
       in writing to Ernst & Young and to the Royal Mail for
       copies of these documents.
           Now, just pausing there, the explanation for that,
       Mr De Garr Robinson -- and I know I don't have to give
       the explanation for your purposes, but just to be
       clear -- I can't conceive of a situation whereby copies
       of those reports are not available in this litigation.
       If for some reason due to the corporate reorganisation
       in 2011 it is the case that they are unavailable, well,
       then steps are going to have to be taken to produce
       them.
           I would like that order drawing up today please.
       Would you like to tell me of a suitable time and date by
       which that witness statement has to be -- or ought to be
       provided?
   MR DE GARR ROBINSON:  My Lord, could I glance back and take
       instructions?
   MR JUSTICE FRASER:  Of course you can.
           (Pause).
   MR DE GARR ROBINSON:  Would your Lordship be happy with
       10.30 on Thursday morning?
   MR JUSTICE FRASER:  I am happy, subject to the following
       observation: depending on what the outcome is, it may be
       necessary to recall witnesses in respect of this,
       but for the moment 10.30 on which ..?
   MR DE GARR ROBINSON:  Thursday morning.
   MR JUSTICE FRASER:  I'm going to make it 9.30 on Thursday
       morning so that if there's anything arising out of it it
       can be dealt with on Thursday morning.  So that is the
       21st.
           Right, so that's that.  There are a range of other
       housekeeping type matters but they can all wait until
       the end of today.
           Right, so ...
   MR DE GARR ROBINSON:  My Lord, I call Angela van den Bogerd.
               MS ANGELA VAN DEN BOGERD (affirmed)
   MR JUSTICE FRASER:  Do have a seat, Mrs Van den Bogerd,
       please.
           Examination-in-chief by MR DE GARR ROBINSON
   MR DE GARR ROBINSON:  Mrs van den Bogerd, there will be file
       that you have in front of you.  Could I ask you to go to
       tab 5 of that file please and you will see on the first
       page of the document in front of you is a witness
       statement that is said to be made by you.  Is that your
       name and address on page 1 of that statement?  {E2/5/1}
   A.  Yes, it's mine.
   Q.  If I could ask you to go to the last page of the same
       tab, page 44 {E2/5/44}, is that your signature?
   A.  It is, yes.
   Q.  And still in the same bundle I would like you to go
       please to what I hope is tab 16 {E4/16/1}.
   A.  Tab 16?
   Q.  Yes.
   A.  Empty.
   Q.  It is empty.
   MR JUSTICE FRASER:  It might be at the front of the tab.
       That's where mine was.
   MR DE GARR ROBINSON:  If I could ask you to go back to
       tab 5, Mrs van den Bogerd.  I'm misleading you horribly,
       I do apologise.  Is there a series of corrections to
       your statement at the front?
   A.  Yes.
   Q.  Have you seen those corrections before?
   A.  Yes, I have.
   Q.  Subject to those corrections, is this witness statement
       true to the best of your knowledge, recollection and
       belief?
   A.  It is.
   Q.  You will be asked some questions now, if you could wait
       there.
                  Cross-examination by MR GREEN
   MR GREEN:  I think the statement on the screen is the one
       from the column issues trial, so probably the one we
       want up is {E2/5/1} please.
           Mrs van den Bogerd, this is a statement that you
       made in November 2018, isn't it?
   A.  It is, yes.
   Q.  And if you look at page 44 {E2/5/44}, it is signed on
       16 November 2018?
   A.  That's correct.
   Q.  Can we go to page 2 of that statement please {E2/5/2}.
       Now, at paragraph 5 you will see there you have
       explained a common theme in your responses in this
       witness statement and you say that the common theme:
           "... is that the evidence put forward by the
       claimants does not concern a bug, defect or error in
       Horizon (ie a Horizon generated shortfall as defined in
       paragraph 41.6 of Post Office's generic defence)."
           And then you say this:
           "They often appear to assume that alleged losses,
       shortfalls or other problems must have been caused by
       Horizon."
           Do you see that?
   A.  Yes, I do.
   Q.  So that's a part of the common theme of your witness
       statement, isn't it?
   A.  It is, yes.
   Q.  And then you say:
           "Given the unspecific nature of many of the
       allegations made and that some relate to things that are
       said to have occurred years ago, it is not now possible
       in this statement for me (and may not be possible for
       anyone else) to definitively state what may or may not
       have happened in each case."
           And then you say:
           "Instead, my evidence aims to give the court
       a balanced view of the range of possible causes behind
       each allegation ..."
           Now, was that your aim in November 2018, to give
       the court a balanced view?
   A.  It was, yes.
   Q.  Do you feel you have succeeded?
   A.  Well, I made some references and set out alternative
       explanations to causes of the losses.
   Q.  So you feel you did broadly succeed, do you?
   A.  Broadly.
   Q.  And then you explain the reason for that.  You say:
           "... because I believe that, assuming the basic
       facts alleged by the claimants are true, there is
       a plausible and much more likely explanation than the
       problem being caused by Horizon."
           And you say:
           "The most common alternative explanation is that
       there may have been an accounting mistake or a user
       error by the subpostmaster, his assistants or
       Post Office."
           Correct?
   A.  Correct.
   Q.  There is one particular example we will come to where
       you say it is a data entry error by Post Office, but it
       is fair to say, isn't it, that broadly your witness
       statement is actually pointing to the likelihood of
       errors made by subpostmasters or their assistants?
   A.  Yes.
   Q.  Now, you have made some corrections which we've got in
       the document that you were referred to and can we just
       look at the types of corrections you have made.  There
       are some corrections which relate to matters that you
       probably realised some time ago.  If we look at
       paragraph 128 of your witness statement on page
       {E2/5/30}.  Just to give you context, this is in the
       passage that begins on page 28 dealing with Pam Stubbs'
       case and if we go back to page 30 and look at
       paragraph 128 you see there it says:
           "The FSC customer account (ie the record of all
       branch discrepancies, TCs, credits and debits on the
       subpostmaster's account) confirms that Mrs Stubbs chose
       to settle ..."
           Is what you originally said:
           "... this shortfall centrally and so this does not
       appear to be a problem with Horizon as I would expect
       Mrs Stubbs to dispute this shortfall if Horizon was
       thought to be the cause."
           Yes?
   A.  That's correct.
   Q.  Now, taking it in stages, let's just look carefully if
       you can first of all at two aspects of what you have
       chosen to correct now.  If we look at -- I'm sorry we're
       going to have to go back between the documents because
       we haven't got it all in one place, but if we look at
       {E2/16/4} and the second box down relates to
       paragraph 128 of your witness statement, doesn't it?
   A.  Yes.
   Q.  And that's the paragraph we have just been looking at
       and what you now say is that Mrs Stubbs settled this
       shortfall centrally, full stop.
   A.  Yes.
   Q.  And you have deleted:
           "... and so this does not appear to be a problem
       with Horizon as I would expect Mrs Stubbs to dispute
       this shortfall if Horizon was thought to be the cause."
           Yes?
   A.  Yes.
   Q.  And you must have realised that -- at the latest you
       must have realised that that correction was necessary
       after the common issues trial and you had heard her
       evidence?
   A.  Yes.
   Q.  And the reason it was necessary -- two important points
       of correction there.  The first one is that the words
       "chose to settle" are very misleading, aren't they, in
       the context of accepting the transaction correction?
   A.  Yes.  Because she did actually dispute it via the
       contracts (inaudible), which I didn't know.
   Q.  Well, yes, let's take it in stages.  There are two
       different points.  The phrase "chose to settle
       centrally", the choice is pay or settle centrally
       effectively, isn't it?
   A.  Yes.
   Q.  And there's no dispute button, as we have already
       established in the previous trial.  So there's no
       dispute button on Horizon and that's why you have
       amended what you previously said, that she chose to
       central centrally, you have taken the word "chose" out
       because it is not a real choice for the subpostmaster,
       is it?
   A.  So the reason I took it out here is because when I wrote
       this I said she chose to settle it because she didn't
       dispute it at the time and that's why I wrote it as
       I did.  What I subsequently found is that she did
       dispute, albeit later, so therefore it was incorrect, as
       I had stated.
   Q.  Right, well, I will take it quite quickly because we
       have already dealt with this in the previous trial but
       the short point is however much you dispute something,
       the maximum you can do on Horizon to dispute it is to
       press "settle centrally"?
   A.  Yes.
   Q.  Now, that's an example of a correction you have made
       from something you would have known a long time ago.
       There's another one at paragraph 145 of your witness
       statement on page {E2/5/33}.  This relates to
       Mr Abdulla's case.  You see paragraph 145 there?
   A.  Yes, I do.
   Q.  This is talking about transaction corrections in the sum
       of £1,092 for two consecutive months and you will
       remember that in fact what we found was there were in
       fact three successive entries of 1,092 on the
       spreadsheet, weren't there?
   A.  Yes.
   Q.  Do you remember?  So what you have done is you have
       corrected the word "over" in the fourth line down just
       after "relevant" to "under" and you have nonetheless
       decided to maintain Post Office's position that although
       there were three identical items, month after month, for
       £1,092, it's only the third one that was in error?
   A.  Yes.
   Q.  So you don't accept that the second one was in error?
   A.  No.
   Q.  So the second one is a coincidence, two 1,092.  The
       third coincidence is a coincidence too far, is that
       right, because that's the one you corrected?
   A.  So the first two were the same figure for two months.
   Q.  Two months running.
   A.  Yes.  And the third one was issued mistakenly and that's
       why the correcting TC was issued to correct that, from
       the evidence then there's two on there that should
       stand.
   Q.  I see, and so that's a correction of the word but not
       the outcome in terms of where you would lay the blame
       for that?
   A.  Yes, that's correct.
   Q.  And then the other sort of correction we've got is
       corrections of evidence that you have only recently
       heard -- you were in court, weren't you, for the
       evidence of Angela Burke?
   A.  I was, but actually for Angela Burke I had already made
       that correction to my solicitors before I heard that
       evidence.
   Q.  So hold on a second, you had already decided to make the
       correction you have made in relation to Angela Burke
       before she gave evidence?
   A.  Yes.
   Q.  Okay.  But let's just have a look at that, if we may.
       So it is paragraph 104 of your witness statement, which
       is {E2/5/25} and there you are dealing with the fact
       that there were two other items left in the stack.
   A.  Correct.
   Q.  And it says, four lines up from the bottom:
           "Mrs Burke did not do this [ie clear the stack] and
       bundled together two customers' transactions into one
       basket ... from Horizon's perspective this would have
       looked like a set of transactions relevant to a single
       customer."
           Yes?
   A.  Yes and that's still correct.
   Q.  And the addition that you have made, if I can just read
       it out, it is -- thank you very much, very efficient.
       We can see the addition that you have made on
       paragraph 104 at the bottom {E2/16/3}.
   A.  Yes.
   Q.  "However this had no bearing on the failed recovery of
       the £150 cash withdrawal."
   A.  That's correct.
   Q.  Now, did you think it was odd then that Mrs Burke was
       extensively cross-examined on the basis of your evidence
       that you had already decided required modification?
   A.  So my understanding of the way that Mrs Burke was
       cross-examined was that that was not actually put to her
       in those words but that Mr Draper did say to her that he
       understood that that didn't have a bearing on the 150,
       that's my recollection.
   Q.  Let's take it in stages.  We've got the reference to the
       transcripts, but very briefly, the point that was being
       made to Mrs Burke was if you leave transactions in the
       stack for longer you are leaving a longer period during
       which an error might occur in relation to that basket.
   A.  And that is --
   Q.  That's the thrust of it.
   A.  And that is correct, yes.
   Q.  But the point was completely obvious from her own
       witness statement on a fair reading of it, that it was
       the last transaction that had gone astray so it had no
       bearing whatever, did it?
   A.  In this situation it didn't because it was a failed
       recovery.
   Q.  And you accept that your original witness statement as
       drafted did not make clear that you were not suggesting
       that any error by her had any effect on it?
   A.  That's correct.  When I put my witness statement
       together I did say that I hadn't done all the detail
       I would ordinarily have done, I didn't have as much time
       to do that and I have looked at other evidence in
       relation to Mrs Burke and what was very clear to me is
       that Mrs Burke had done absolutely nothing wrong in that
       situation.
   Q.  You didn't say that, did you?  The reason I'm asking
       this a little bit carefully is because you introduce
       your witness statement and focus on the likely reason
       being user error and that's the tenor of what we're then
       expecting to read about?
   A.  Yes.
   Q.  So when you don't point out she has done absolutely
       nothing wrong, the natural way of reading your witness
       statement is that she had done something wrong in some
       way?
   A.  I agree, which is why --
   Q.  And that's why you have corrected it now?
   A.  Absolutely, yes.
   Q.  But it was completely clear from her witness statement,
       wasn't it, that it was the last transaction that failed?
   A.  Yes.
   Q.  And the reason it was clear was she actually exhibited
       a copy of the receipt so when you looked at her witness
       statement there was an exhibit with a copy of the
       receipt and you could see it was the last transaction?
   A.  Yes.
   Q.  Then if we just look at the change made to paragraph 72
       please, back one page on that document on Opus
       {E2/16/2}, this is in relation to Mr Patny and his
       difficulties with MoneyGram:
           "Mr Patny is correct that on 23 February 2016 he
       processed a MoneyGram transaction for £3,100 and the
       customer's debit card payment for the transaction was
       declined by the customer's bank.  At this point the
       transaction was committed and could not be removed from
       the stack, therefore Mr Patny had to settle to cash.
       Process is that the MoneyGram transaction should have
       been cancelled on Horizon followed by a reversal of the
       transaction.  The data shows that Mr Patny cancelled the
       transaction, however did not complete the reversal.
       This would result in the £3,100 loss."
           Now, what you said at the time was a little more
       doubtful, you just recite what he is saying he did
       rather than accepting it.
   A.  Mm-hm.
   Q.  And --
   MR JUSTICE FRASER:  Can I just ask you to pause there just
       for a technical reason.  I have been working off the
       hard copy of Mrs van den Bogerd's corrections, which
       seems to be different to the one that's on the common
       screen.  Are there two different versions?
           Mrs van den Bogerd, the clip of documents which
       Mr De Garr Robinson took you to, is that three sheets in
       length?
   A.  It is three sheets I have in front of me.
   MR JUSTICE FRASER:  It is three sheets?
   A.  Yes.
   MR JUSTICE FRASER:  Can we just scroll to the end of
       {E2/16/4} and see how many pages long that is.
   MR GREEN:  My Lord, I'm using the one that is the common one
       for all witnesses.  We have one for each witness and --
   MR JUSTICE FRASER:  That's a perfect explanation.  As long
       as there is only one version of the document.
   MR GREEN:  There is only one version.
           Because what you actually said in your witness
       statement was that you were assuming the facts that the
       claimants alleged were true.
   A.  Yes.
   Q.  And it is only after hearing Mr Patny's evidence and the
       fact that the helpline did tell him to settle to cash
       that you have changed this, isn't it?
   A.  No, so again my corrections were made in advance of the
       claimants giving their evidence last week.  So I put
       this in to my solicitors before the trial started.
   Q.  Right, so we have only found out about it after they
       have been cross-examined?
   A.  Yes.
   Q.  I see.
   A.  I did put it in before.
   Q.  I understand, thank you very much.
           Can we move on to the first section in your witness
       statement please where you respond briefly to points
       made by Mr Henderson.  Now, you were familiar with
       Mr Henderson's role with Second Sight?
   A.  I was.
   Q.  You were the programme director of the branch support
       programme?
   A.  Correct.
   Q.  You were involved in preparing reports in response to
       Second Sight?
   A.  Correct.
   Q.  You were involved in decisions about what information
       was provided to Second Sight?
   A.  Correct.
   Q.  And at paragraph 9 of your witness statement on
       {E2/5/3}, you say:
           "At paragraph 2.4, Mr Henderson says that he
       analysed some transaction data provided to him.  I do
       not know what transaction data he is referring to or
       what he means by him being able to 'reverse engineer'
       this data."
           Now, paragraph 2.4 says that he was provided the
       information by Mr Jenkins, Gareth Jenkins, in his
       witness statement?
   A.  Mm-hm.
   Q.  You know who Mr Gareth Jenkins is, don't you?
   A.  I know who he is, yes.
   Q.  Have you ever met him?
   A.  No.
   Q.  Because there are quite a lot of references in the
       Fujitsu witnesses' witness statements in particular to
       Mr Jenkins, aren't there?
   A.  Yes, that's correct.
   Q.  And particularly in Mr Godeseth's witness statement?
   A.  Yes, I understand.
   Q.  And you in turn refer to Mr Godeseth's witness statement
       in your witness statement?
   A.  I do.
   Q.  Did you make any requests for clarity from Mr Jenkins
       about what that transaction data was if it was provided
       by Fujitsu?
   A.  No, I didn't.
   Q.  Because presumably Post Office would have had to agree
       for Fujitsu to provide it if it was to Second Sight?
   A.  So I wasn't involved in all of that as part of the
       programme.  I was aware that Mr Henderson had met with
       Gareth Jenkins, but I wasn't fully aware of all the
       detail of that at the time.
   Q.  Okay.  Let's move on to transactions outside working
       hours.  That's on {E2/5/3}, paragraph 12 of your witness
       statement and you would agree on the face of it, if
       Mr Henderson was concerned about transactions outside
       working hours that that's a reasonable thing for him to
       be at least interested and curious about, isn't it?
   A.  Yes.
   Q.  On the face of it?
   A.  Yes.
   Q.  And paragraph 13 of your witness statement gives what
       you explain are a number of plausible and legitimate
       explanations which you set out below, do you see that?
   A.  Yes, I do.
   Q.  First line.  {E2/5/4} and you say:
           "For the sake of clarity, I categorically confirm
       that I am not aware of any improper conduct by
       Post Office or Fujitsu like this, or of any reason why
       Post Office or Fujitsu would engage in such conduct.
       I am informed by Post Office's solicitors that in the
       course of investigating this matter ..."
           So that's for this trial, isn't it?
   A.  Yes.
   Q.  "... Fujitsu have advised that 'phantom sales' were
       reported in around 2000 which appeared to be caused by
       hardware issues."
           Now, pausing there, were you aware prior to being
       told that in relation to these proceedings that there
       had been phantom sales problems on the Horizon system?
   A.  Not that I can recall.
   Q.  So if a postmaster had said that to you at the time,
       prior to this litigation, you would have assumed it was
       actually more likely to be user error, is that fair?
   A.  I would have and then an issue would have been raised
       and would have been investigated from the Horizon logs
       to see what was there.
   Q.  And it is fair that you explain you have had quite a lot
       of experience looking into discrepancies with
       subpostmasters?
   A.  I have.
   Q.  So it's against the background of that experience that
       you nonetheless weren't aware of even the existence of
       a problem with phantom sales having happened?
   A.  Yes, that's correct.
   MR JUSTICE FRASER:  Can you just give me an approximate
       either month or year when you became aware of phantom
       sales?
   A.  I don't recall, my Lord, the approximate year.
       I mean -- so when I got involved with the mediation
       scheme that's probably when we started to hear claims of
       phantom transactions and that's -- when we investigated
       some of those claims and we could see no evidence of
       that throughout that, so that would have been a few
       years back.  I can't remember exactly when.
   Q.  Just to make sure I correctly heard your answer, you
       investigated and you could see no evidence of that?
   A.  So I can't remember exactly -- so when we were doing the
       mediation scheme then we had people saying to us they
       were phantom sales.  Some of the mediation ones that
       I looked into is, as I set out in here, when we have
       looked into that it was transactions being done after
       the system had automatically logged somebody off, so if
       they had been dormant for 59 minutes it would actually
       complete what was in the basket and produce the receipt
       and one of the instances I can remember was that was
       after they had closed the branch and that they thought
       that was a phantom transaction and we could explain that
       from the evidence on the transaction logs to how that
       happened.
   Q.  Just going back, just looking at paragraph 13 again,
       four lines down on the right-hand side:
           "I am informed by Post Office's solicitors that in
       the course of investigating this matter, Fujitsu have
       advised that 'phantom sales' were reported in around
       2000 ..."
           So that information appears to have come to you for
       the first time from Post Office's solicitors, is that
       right?
   A.  Yes, I wasn't aware that we had had phantom sales in
       2000.  I wasn't aware of that.
   Q.  So up until that point you would have assumed it was
       user error?
   A.  Yes, I would have.
   Q.  Let's have a look, if we may please, at the phrase that
       you have used, "plausible and legitimate explanations".
       If you focus on paragraph 13 for a second, you see that
       after the bit about phantom sales having been reported
       "which appeared to be caused by hardware issues" it
       says:
           "This is dealt with in the witness statement of
       Fujitsu's Steve Parker, but I understand the key point
       to be that such matters, provided they relate to stock
       sales, should not cause a discrepancy in a branch's
       accounts."
           Now, pausing there, are you saying that a phantom
       stock sale is a plausible and legitimate explanation for
       a phantom transaction?  What are you saying exactly
       there?
   A.  So if stock -- so say stamps had been put into the
       basket and it wasn't by the user, then that's what I'm
       referring to in terms of a phantom sale, so it would be
       put in as if it was being sold and hadn't been sold.
   Q.  But what if the stamps were self-declaring without input
       from the user?  That's a bit of a problem, isn't it, if
       you're the subpostmaster?
   A.  Yes.
   Q.  If a ridiculous value of stamps was self-declaring, it's
       not the stock you've got, that's a big problem for the
       subpostmaster, isn't it?
   A.  It would be in that -- putting into the stack, yes.
   Q.  Okay.  Let's have a look if we may at the phantom sales
       master PEAK.  Now, you know what a master PEAK is, don't
       you?
   A.  The main one.
   Q.  It is the main -- where there are a number of iterations
       of a problem what Fujitsu do is they pull together
       a number of PEAKs into a master PEAK to collect the
       cases of different people who have had the same problem?
   A.  Yes.
   Q.  And let's look please at {F/97}.  This is the master
       PEAK for phantom sales, PEAK number 0065021, dated
       17 April 2001.  That's when it is actually initiated in
       the main box and the target date in the PEAK you will
       see is the 13 June 2001, do you see that?
   A.  I do.
   Q.  Have you seen this PEAK before?
   A.  Not that I recall.
   Q.  So --
   A.  I certainly didn't see this back in 2001.
   Q.  So there are a number of different FAD codes for
       different SPMs in this PEAK, but can we just look -- is
       there any way we could slightly increase the size of the
       text?
   A.  Thank you.
   Q.  Mrs van den Bogerd, can you see that all right?
   A.  I can, yes.
   Q.  That's perfect, thank you very much.
           So if we just look at 14/04/01 at the top there, do
       you see:
           "New complaint call as previous ... closed WITHOUT
       permission from PM."
           Do you see that?
   A.  I see that.
   Q.  And the reason that "WITHOUT" is in capitals is because
       the call shouldn't be closed without the agreement of
       the subpostmaster; that's right, isn't it?
   A.  Yes.
   Q.  And the next one done:
           "PM wishing to complaint about ongoing system
       problems, see call 0104 [et cetera] ... for details."
           Then:
           "PM had previous complaint open ... that PM was
       under impression (correctly) that it could only be
       closed with his permission.  It would appear Ki Barnes
       gave authorisation to close that call.  PM VERY unhappy
       about this."
           Do you see that?
   A.  I do.
   Q.  Then the next one down, 12.58:
           "Information: PM extremely unhappy about the
       problems with his counters.  He says he has had to pay
       out over £1,500 in losses that are due to these
       problems.  He has informed POCL ..."
           Post Office Counters Limited, which was then the
       name of the Post Office, yes?
   A.  Yes.
   Q.  "He has informed POCL they can suspend him because he is
       refusing to make good any further losses."
           Yes?
   A.  Yes.
   Q.  "PM wants a face-to-face meeting with someone in
       authority from Pathway/POCL to discuss the issues.
       PM feels very strongly about this and says he is willing
       to take POCL to a tribunal/court because of the stress
       he has suffered because of the problems."
           Yes?
   A.  Yes.
   Q.  Then halfway down the page you can see "JULIAN HALL", in
       capitals, and:
           "Information: THIS CALL IS ONLY TO BE CLOSED WITH
       THE EXPRESS PERMISSION OF JULIAN HALL."
           Who is this particular subpostmaster who seems to be
       one of the main complaints in this PEAK, yes?  And
       that's the correct procedure, it shouldn't be closed
       without his position, yes?
   A.  That is the correct procedure, yes.
   Q.  Then 14/04/01:
           "PM has lost all confidence in system and Ki Barnes
       as he feels she has misled him over previous calls."
           Then underneath that:
           "The system seems to lose transaction and PM is
       concerned that for every transaction error he notices
       there is the probability that there are ones he misses,
       leading to discrepancies.  The PM is at present finding
       the whole scenario very stressful and is suffering
       sleepness nights due to these problems.  In the light of
       what has gone on the PM is prepared to break his
       contractual obligations with POCL and refuse to pay any
       more discrepancies and will take legal action if
       required."
           So it is clear, isn't it, here that before we get
       into the detail about what the root cause is, which we
       will come to in a moment, the PM has had a call closed
       wrongly that should not have been closed.
   A.  Yes.
   Q.  Is extremely distressed about it, yes?
   A.  Yes.
   Q.  And there are significant sums of money involved?
   A.  Yes.
   Q.  Is that fair?  And you can understand why it's a fair
       complaint to make about Ki Barnes that she has
       authorised the closure of a complaint that should not
       have been closed; that's fair?
   A.  That is fair.
   Q.  Right, let's look at what we find under 17/04/01 at
       9.48:
           "Contacted: I have left a message on Ki Barnes
       voicemail as the PM is now complaining about her.  I was
       speaking to her about the last complaint call and we
       both feel that this PM is complaining unjustly.  She has
       been in contact with him and I feel he is complaining
       because the feedback has been advising it is user error,
       whereas the PM thinks it is software."
           Now, pausing there, let's leave to one side for
       a moment the PM supposedly complaining unjustly, it is
       clear from this that the feedback the subpostmaster is
       getting is that this is user error, yes?
   A.  From -- yes, yes.
   Q.  And the PM thinks it is software?
   A.  Yes.
   Q.  So this is a good example of the competition which you
       have referred to in your witness statement --
   A.  Yes.
   Q.  -- of the subpostmaster saying it is software and
       effectively what appears to be the feedback he is
       getting that it is user error?
   A.  Yes.
   Q.  The fact that the feedback says it is user error, do you
       think that might have contributed to the closing of the
       call by Ki Barnes, or is that not in your experience?
       You may not be able to answer.
   A.  No, it would depend on whether other investigations had
       taken place.
   Q.  Okay.  Let's have a look.  If we go down to the bottom
       of this page, 10.17:
           "As I was on the phone to the PM, he advised that
       three first class stamps that were on the screen just
       'dropped off'.  PM had three first class stamps ..."
   A.  Sorry, I have lost that.
   Q.  I'm so sorry, it is just at the bottom there.  Just
       under "Contacted".
   MR JUSTICE FRASER:  Hold on one second.  We have just jumped
       down to the second part but it is the same page.
       I shared your slight concern we had gone somewhere else.
       Can you now see the entry?
   A.  I can see that now, thank you.
   MR GREEN:  So you see "Contacted" there.
   A.  Yes.
   Q.  "As I was on the phone to the PM, he advised that three
       first class stamps that were on the screen just 'dropped
       off'."
           That's while he is actually on the phone he is
       reporting something that appears to happen either during
       the telephone call or had just happened:
           "PM had 3 first class stamps and other stamps for
       30p.  When the other stamps 30, went on, the first class
       stamps disappeared.  They have since put the 3 first
       class stamps again.  The first transaction (that
       disappeared) was put on as 2 first class stamps and
       1 normal first class stamp."
           And the transaction IDs follow.  So he is clearly
       having a problem with the stamps.
           If we go on to the next page of this PEAK {F/97/2}
       you will see there:
           "PM advises that the transactions taken after this
       one ... which basically skips out transaction
       [number 77].  We were unable to find out the transaction
       ref of the 1 first class stamp that was taken with the
       2 ... can SSC please investigate why he has had
       disappearing transactions again and perhaps
       a recommendation as to what action can be taken."
           Now, there is quite a lot of content in this and I'm
       just going to take you to a couple of points if I may.
       Let's go forward to page 4 of this PEAK please {F/97/4}
       and about halfway down, just more about two-thirds of
       the way down, about an inch below the Dell on the side
       of the screen can you see 01/05/01, 9.36?
   A.  Yes.
   Q.  This is actually a different FAD number.  It says:
           "PM wants to speak to someone face-to-face and is
       fed up with things getting passed back and forth to and
       from different departments and nothing ever happening.
       PM is willing to travel if he has to in order to speak
       to someone face-to-face."
           So we have had Mr Hall wanting to speak to someone
       face-to-face and we appear to have a different
       subpostmaster wanting to speak face-to-face.  You will
       have encountered, won't you, in your experience when
       these problems happen subpostmasters just eventually get
       to a point where they get fed up with dealing with the
       helpline and they just want to talk to someone
       face-to-face to get it sorted out?
   A.  Yes.
   Q.  And that doesn't always happen, does it, when they want
       to and some of them get quite frustrated?
   A.  And I can understand that frustration, yes.
   Q.  Let's look at page 5 please --
   MR JUSTICE FRASER:  Just before you -- are you moving off
       that one?
   MR GREEN:  I am.
   MR JUSTICE FRASER:  A little bit further down,
       Mrs van den Bogerd, there's an entry -- I just wondered
       if you understood what it is and if you don't, please
       don't worry.  There's an entry 1 May 2001 at 10.56, do
       you see that?
   A.  Yes.
   MR JUSTICE FRASER:  It says:
           "Information: ROMEC are contacting the site to let
       them know that they will be attending site ... to fit
       suppressors and double sheet flyleads, in order to help
       the enviromental fault."
           Do you know what ROMEC do?
   A.  ROMEC were Royal Mail engineers --
   MR JUSTICE FRASER:  Royal Mail engineers?
   A.  In-house electricians and engineers.
   MR JUSTICE FRASER:  All right, thank you very much.
       Mr Green.
   MR GREEN:  My Lord, I was just about to pick up --
   MR JUSTICE FRASER:  That's fine.
   MR GREEN:  If we look at page 5 of this PEAK {F/97/5}, if we
       go down to the bottom do you see 3 May 2001?
   A.  Which time?
   Q.  15.34.
   A.  Yes.
   Q.  So:
           "ROMEC have been to site and have done all they can
       do.  There is no more UKSS2 can do for this site."
           What is UK SS2?
   A.  I'm sorry, I don't know.
   Q.  Not sure.  And then underneath that, 04/05:
           "Information: Ki Barnes has called in.  I am unsure
       as to what to do with this call now.  ROMEC [which is
       Royal Mail engineers] have been to site and state that
       they have actually seen the phantom transactions, so it
       is not just the PM's word now.  They have fitted
       suppressers to the kit ..."
           Which is the reference his Lordship took you to:
           "... but the PM is still having problems.  As yet
       there has been no recurrence to the phantom transaction
       but there still may be problems."
           Now, pausing there, that is independent site visit
       corroboration of the problem by Royal Mail's own
       engineers at the branch, isn't it?
   A.  Yes.
   Q.  That's clearly not user error any more?
   A.  No.
   Q.  Let's look at page 7 {F/97/7}, and if we go down to
       19 June 2001 at 3.17, do you see that?
   A.  Yes.
   Q.  "I now have pressing evidence to suggest that unwanted
       peripheral input is occurring, the likely source being
       the screen.
           "This has been seen at Old Isleworth ... and
       Wawne ... with OI being the best site; when the PM has
       been asked to leave the screen on overnight I have
       observed system activity corresponding to screen presses
       happening with no corresponding evidence of either
       routine system activity or human interference ..."
           So on the face of it there's unwanted peripheral
       input occurring into the system?
   A.  Yes.
   Q.  Can we go to page 9 {F/97/9}, top of page 9,
       7 August 2001 please, do you see that?
   A.  Yes.
   Q.  Halfway across "Repeat call":
           "PM Mr Julian Hall back online - has just been
       speaking to Becky - insists that he would like this call
       updating with this information.  When the
       transaction ... was entered the cheque for £75 did not
       appear in the stack, but then turned up on the cheques
       listing.  This information was not mentioned when he
       initially spoke to Becky and would like it including in
       the call."
           Yes?  So he still seems to be having difficulties
       and he is trying to make sure that he is giving full and
       precise details of what's going on, isn't he?
   A.  Yes.
   Q.  And then if we look down at 24 September 2001 -- this is
       just slightly outside the main stream of the items
       I have been asking to you look at -- we can see this is
       a different SPM.  You can see there it says:
           "PM called as her system said that it was printing
       a report for 20 minutes and she wasn't even printing
       a report.  She tried to settle a transaction to cash and
       that came on screen.  Advised PM to reboot and told her
       I would update the call.  PM happy."
           Yes?
   A.  Yes.
   Q.  So there are some things which appear to be actual
       phantom transactions and then there are other things
       which are phantom events that we might find in the
       events log?
   A.  Yes.
   Q.  Like printing a report, when she is there; it is not
       printing a report for 20 minutes at all, yes?
   A.  Yes.
   Q.  And then there's a repeat call about a monitor and then
       if we look at the bottom of the page, 12 November, this
       is basically the conclusion to all this:
           "Phantom transactions have not been proven in
       circumstances which preclude user error.  In all cases
       where these have occurred a user error related cause can
       be attributed to the phenomenon."
           And over the page {F/97/10}, "No fault in product".
           It doesn't suggest, does it, a great willingness to
       acknowledge what appear to be independently observed
       faults with aspects of the Horizon system, does it?
   A.  Not from this, no.
   Q.  Now, if we go please to {F/773}.  This is PEAK 0208335
       and if we just quickly look in the summary it will give
       us -- this is February 2011:
           "Branches will be forced to declare stock when they
       don't want to.  Apparent reappearance of withdrawn stock
       may cause spurious discrepancies."
           Do you see in the summary box across the middle?
   A.  Yes.
   Q.  "Avoidance action taken so far is only appropriate in
       the short-term while very few branches are affected;
       over the next few months it could be affecting around
       10 branches per week.  NBSC aware of problem."
           Do you see that?
   A.  Yes.
   Q.  Just if we go over the page please to {F/773/2}, if you
       look at the top -- just go halfway across the line at
       the top:
           "The office went into the declaration and confirmed
       it showed the PO saving stamps and other stock items,
       some showing as minus stock amounts, that were not
       correct to the branch."
           So the problem with negative stock appearing in --
       this is in 2011:
           "It somehow seems that the system has somehow picked
       up this declaration and this is the cause of the
       discrepancies appearing on the system.  The SPMR was
       told to declare the correct stock figures but is
       reluctant to do this as this will cause discrepancies
       when she next balances that are not relevant to herself
       at the moment."
           So that's a matter that would concern any
       subpostmaster, isn't it, because they are effectively
       having to declare something that they don't agree with?
   A.  Yes.
   Q.  And enter into the account on the Horizon system
       effectively an account which accepts a discrepancy that
       they completely disagree with?
   A.  From reading that, yes.
   Q.  Yes.  So had you seen this PEAK before?
   A.  No.
   MR JUSTICE FRASER:  Before today, or before your witness
       statement?
   A.  No, I don't recall seeing this at all.
   MR GREEN:  So is it right that if you had heard that story
       from a postmaster in any of the many conversations you
       have had about difficulties they faced, you would have
       assumed it was user error, wouldn't you?
   A.  Without -- yes, so without them ... with the situation
       being referred to the Horizon service desk and Fujitsu
       looking at it, without that happening then yes I would
       have, if there was nothing to support ... but what
       I would have done is looked at what's happened in branch
       to determine whether there was a plausible explanation
       in branch and if not if there's anything -- I mean this
       is quite specific around -- I have seen this happen at
       this time, I've got the reference -- the number of the
       session ID, so this is quite specific.
   Q.  In a sense this subpostmaster was quite lucky that what
       was being thrown up was stock that wasn't in use any
       more because it's a lot easier for them to say "Well,
       hold on a second, I don't even have that --"
   A.  Yes.
   Q.  -- "that item shouldn't even be there".  So that's
       really luck that that was the problem for this
       subpostmaster, isn't it?
   A.  In that it was obvious, yes.
   Q.  Much easier for them?
   A.  Yes.
   Q.  But if what's thrown up is a stamps discrepancy for
       first class stamps that they do stock, it's much harder
       for the subpostmaster?
   A.  Yes, it would be.
   Q.  Yes.  And what we see there is that whilst the previous
       PEAK appears to be a hardware problem of phantom
       transactions, this one appears more like a software
       problem, does not it?
   A.  From this, yes, it does.
   Q.  And we can see at page 7 of this PEAK {F/773/7}, we can
       see that this is closed on 10 September, final
       "Administrative response" is the category for that PEAK.
       It doesn't really give a hint of the trouble that's been
       caused, does it, that category?
   A.  No.
   Q.  Mr Parker very fairly accepts in his witness statement
       that the categories assigned were somewhat subjective.
   A.  Mm-hm.
   Q.  And were not ever actually audited to ensure
       consistency.
   A.  Yes.
   Q.  I think that's a pretty shining example of that,
       isn't it?
   A.  Yes.
   Q.  Now, can we go now to {F/1286.2/2}.  This is
       a 9 December email and it is from the NBSC admin team to
       the branch support team.  Now, in 2014 you were head of
       the branch support team, weren't you?
   A.  Yes.
   Q.  And what we see there is:
           "Branch reporting that he has found sensitive issue
       with Horizon when the system put a phantom cheque on the
       cheque line in July 2013.  Claims to have evidence to
       support his claim."
           Do you see that?
   A.  Yes.
   Q.  That would be pretty interesting to see, wouldn't it,
       the evidence, if someone claims to have it?
   A.  Mm-hm.
   Q.  "Although he himself did not suffer a loss, thinks that
       Horizon is flawed.  Did not ask to be contacted about
       this.  Just wanted to say that he had this information
       and threatened to go to MP as a result."
           Do you see that?
   A.  Yes.
   Q.  Now, can we go back to page 1 please and just follow
       this through {F/1286.2/1}.  This is sent up by
       Nigel Allen to Andrew Winn, Andy as he is known, yes?
       And you know Andrew Winn, don't you?
   A.  Yes.
   Q.  And you have worked with him?
   A.  Not for very long; not directly.
   Q.  Okay.  And what we see there is in the middle of the
       page, Nigel Allen is saying:
           "Given the current media and in particular the BBC's
       attention on Horizon, do you think it is worthwhile
       looking into this 'alleged flaw' with Horizon that this
       SPMR has highlighted to preempt any enquiries from his
       MP?"
           And what comes back is:
           "Hi Nigel,
           "There is nothing I can investigate given the level
       of detail provided unless he only accepted one cheque
       in July 13.  Even then I don't have the level of detail
       needed and would need Fujitsu support.  Without
       a date/value we can't really raise a request.
           "I don't really understand what the purpose of the
       call is.  Does he want it investigated or not?  My
       instinct is that we have enough on with people asking us
       to look at things.
           "I can't figure out how if a phantom cheque appeared
       on Horizon he could avoid a loss unless another phantom
       transaction took it away again!!!!"
           Yes?  So let's just unpack that, if we may.  The
       first point is that information which might be relevant
       if you had an interest in knowing the truth about the
       reliability of Horizon has been offered --
   A.  Yes.
   Q.  -- by a subpostmaster who has rung up, and the
       subpostmaster has said they've got evidence, as we saw,
       yes?
   A.  Yes.
   Q.  And if one had any interest whatsoever in knowing the
       truth about the reliability of Horizon, you would say
       "Well, can you please get him to send the evidence",
       wouldn't you?
   A.  Yes, you would.
   Q.  And that's not what happened?
   A.  No.
   MR JUSTICE FRASER:  Do you consider that that reaction from
       Mr Winn is an adequate reaction --
   A.  No.
   MR JUSTICE FRASER:  -- or an inadequate reaction?
   A.  That's a totally inadequate reaction.  The fact that we
       had the details from the branch on there -- you know,
       the name of the branch, the FAD code, it would have been
       very easy for Mr Winn to have contacted to get further
       information and I would have expected him to have done
       so.
   MR GREEN:  There is one final point on this email, if I may,
       just quickly.  The last bit of Mr Winn's email where it
       says:
           "I can't figure out how if a phantom cheque appeared
       on Horizon he could avoid a loss unless another phantom
       transaction took it away again!!!!"
           And the point about that is that shows, red in tooth
       and claw, that if a phantom cheque comes up into your
       account you're going to suffer a loss, aren't you?
       That's what he's saying?
   A.  That's what he is saying there, yes.
   Q.  And that's right, isn't it?
   A.  Well, it depends -- you know, it would come in as
       a payment for something, so if he has had -- I'm not
       sure actually and this is why it would have needed to
       have been investigated because if a phantom cheque came
       in then it would come in as a method of payment.
   Q.  Well, if a phantom cheque -- if you're supposed to have
       received money that you haven't actually had, you are
       not going to have the money in your branch, are you?
       You're going to be short of something if on the system
       you are supposed to have had a payment for something and
       you haven't actually received it?
   A.  He is saying he has had a phantom cheque appear on
       Horizon.
   Q.  Yes.
   A.  So to me that means he's got a cheque there that -- oh,
       I see, he's got a cheque there that he hasn't actually
       got.
   Q.  Yes.  Horizon is showing a cheque he hasn't actually
       had?
   A.  I think that is the classic example of why that needed
       to be investigated.
   Q.  Yes, because -- let's put it very neutrally -- at the
       very lowest, on the face of it there's a risk that he
       may have had a loss --
   A.  Yes, absolutely.
   Q.  -- in that type of situation; is that fair?
   A.  Absolutely.  And even though he has said he didn't have
       a loss it could have meant that he made an error
       somewhere else and he could have had a loss after all.
   Q.  Quite.  Yes.
           So this doesn't seem to have been drawn to your
       attention at all, as far as we can see?
   A.  I don't remember this at all.
   Q.  No.  Were you aware of that sort of problem being
       discussed?
   A.  No.  I mean if I had become aware of that I would have
       requested an investigation to be done on that,
       absolutely.
   Q.  Because people have had quite a lot of problems with
       cheques over the years, haven't they?  Subpostmasters,
       one of the things that --
   A.  Yes, there have been errors with cheques.
   Q.  Yes.  Let's have a look, if we may please, at the next
       topic in your witness statement, transactions not
       associated with SPM user IDs, which we see on page 6 of
       your witness statement {E2/5/6}.  If we go to
       paragraph 18 {E2/5/6} and this is you responding to
       Mr Henderson's suggestion that there may be transactions
       associated with the user ID that the subpostmaster
       didn't conduct.
   A.  Mm-hm.
   Q.  Leaving aside for a moment the husband or wife who have
       moved over while the other one has logged on and used
       the log in or something like that, just leave that aside
       for a minute, but one possibility is that an auditor has
       accepted a TC on behalf of a subpostmaster, isn't it?
   A.  An auditor whilst in branch?
   Q.  Yes.
   A.  Yes, that is a possibility.
   Q.  Yes, because if we look at {F/499}, this note is
       actually in Mr Abdulla's case:
           "Call from auditor on 08.04.09 - postmaster not
       present at audit so this TC was accepted and settled
       centrally by auditor."
   A.  Okay.
   Q.  Is it your understanding that that's how things should
       proceed in the absence of a subpostmaster at a branch?
   A.  Not normally.  If a postmaster isn't available when
       there's an audit then it is usually whoever they have
       left in charge that we would ask to process anything
       that would need to be processed on the day.
   Q.  Okay.  Let's look at paragraph 18.4 {E2/5/6} where you
       go on to deal with what you describe as a:
           "... further very rare scenario in relation to
       Legacy Horizon only, involving the insertion of
       a transaction at the counter by the SSC."
   A.  Yes.
   Q.  "In this instance Horizon would associate the
       transaction with the user ID of the individual logged on
       at that counter.  If nobody was logged on at the time
       the transaction was inserted then the user ID would be
       missing.  These transactions would be clearly
       identifiable in the audit trail as having been inserted
       by SSC."
           Now, taking it in stages, in the records where the
       log in ID was shown, if the subpostmaster is logged on
       at the time, the subpostmaster's own ID would show,
       wouldn't it?
   A.  Yes.
   Q.  And you made this witness statement in November 2018.
       This was something that you were aware of when you made
       this witness statement, isn't it --
   A.  Yes.
   Q.  -- the possibility of inserting transactions?
   A.  The possibility but I've never actually seen this
       happen.  The possibility of it, yes.
   Q.  How long have you known about that possibility?
   A.  This is something I have not -- because I have not
       experienced it myself, I have not known of it that long
       actually.
   Q.  Could you give the court a rough idea of how long you
       have known it was possible?
   A.  In terms of inserting transactions, last year or so.
   Q.  Who told you about it?
   A.  Anything to do technically with Horizon I would get the
       information from Fujitsu.
   Q.  And was it Gareth Jenkins or was it someone else, can
       you remember?
   A.  I have not got this directly from Gareth Jenkins.
   Q.  Do you know if it originally came from him?
   A.  It probably did.
   Q.  Probably from Gareth.  When you say it would be visible
       in the audit trail at the end of 18.4 {E2/5/7}, what are
       you referring to as the audit trail?
   A.  So on the transaction log in branch.
   Q.  You say it would be visible on the transaction log --
   A.  In the branch.
   Q.  -- in the branch?
   A.  It would be visible to the branch that somebody other
       than that user had done a transaction.  That's my
       understanding of how that works.
   Q.  Do you -- or do you in fact mean -- if you don't
       actually know and you are having to sort of slightly do
       your best for the court --
   A.  Yes.
   Q.  -- that's not a fair thing for you to be asked to do,
       because you are on oath.
   A.  I've never see --
   Q.  Perfectly okay to say you're not sure.
   A.  I've never seen this so I can't speak on it from a
       working knowledge of actually seeing it.
   Q.  And you don't actually know whether it would be on the
       transaction log or actually in the audit store data in
       the audit, do you?
   A.  So my understanding was it would be in branch that it
       would be visible, but as I say I've never seen it so
       I couldn't actually ...
   Q.  Okay.
           My Lord, would that be a convenient moment for
       a break?
   MR JUSTICE FRASER:  Yes, it would.
           Mrs van den Bogerd, you will recall this from last
       time but you are now in the middle of your evidence so
       you are not to talk to anyone about the case.  We are
       going to have a short break for the shorthand writers.
       We will come back in at 5 to.
   (11.45 am)
                          (Short Break)
   (11.55 am)
   MR GREEN:  Mrs van den Bogerd, can I just ask you to have
       a quick look at page 45 of the transcript just before
       the break.
           We see at line 7 I said:
           "Question: And you know Andrew Winn, don't you?
           "Answer: Yes.
           "Question: And you have worked with him?
           "Answer: Not for very long; not directly."
   A.  Mm-hm.
   Q.  Just to clarify, could we look please at {F/1114}.  This
       is 7 August 2013.
   A.  Yes.
   Q.  This is the branch support programme terms of reference
       and this was the programme that you were in charge of,
       wasn't it?
   A.  That's correct.
   Q.  And at this stage Post Office had decided to try to
       address the concerns raised by some postmasters over
       recent years and at this stage Post Office commissioned
       the independent review and so forth and we see the scope
       there, first bullet point is:
           "Post Office's attitude to subpostmasters which is
       often defensive and unsympathetic, with a focus to
       recover assets rather than to identify the root cause of
       the problem."
   A.  Yes.
   Q.  This is what the interim report had identified,
       isn't it, as we see in the introduction to those bullet
       points?
   A.  Yes.
   Q.  And then the third one, for example:
           "Lack of timely, accurate and complete information
       provided to subpostmasters to support them in resolving
       issues."
           Yes?
   A.  Yes.
   Q.  And then the approach is then explained and if we could
       go over the page please {F/1114/2} you will see there
       the governance and key stakeholders are then set out and
       Alice Perkins -- who was Alice Perkins?
   A.  She was the chairman at the time.
   Q.  She was the chairman.  And Paula Vennells was the CEO
       until very recently?
   A.  Yes.
   Q.  And they have requested the establishment of the
       programme, which will be led by you?
   A.  That's right.
   Q.  With Gayle Peacock accountable for running it on an
       operational level?
   A.  That's correct.
   Q.  Then if we look down to the underneath table in terms of
       finance, on the programme board was Mr Ismay and then
       working group level, Mr Winn, Alison Bolsover and
       Paul Lebeter?
   A.  That's correct.
   Q.  That was the context, was it, in which you worked with
       Mr Winn on these issues?
   A.  Yes.  When we set up the mediation scheme as well,
       a number of the queries -- when we were investigating
       some of the issues then Andy Winn had been involved in
       some of them previously and had worked with Second Sight
       on some of the spot reviews as well, so I did -- that
       was the first time I had come across Andy Winn, and
       he -- FSD at the time was run by Rod Ismay and he was
       part of that team.
   Q.  And his responsibility and the reason why he was
       included in that group is in the right-hand column,
       isn't it?
   A.  Yes.
   Q.  The right-hand column is headed "Reason for inclusion
       within the group".
   A.  Absolutely, yes.
   Q.  And if we look at the right-hand corner it says:
           "Responsible for branch accounting and client
       settlement.  Also responsible for resolving specific
       branch accounting issues."
   A.  Yes.
   Q.  So in terms of that, that was sort of front and centre
       of what he was properly supposed to be dealing with,
       isn't it?
   A.  The second part of that, yes, the issues.  Not the
       branch account and client settlement.
   Q.  Okay, so Mr Winn was responsible for the specific branch
       accounting issues?
   A.  Yes, he did the dispute resolution, yes.
   Q.  So in context, he was the right person for that email we
       looked at before the break to have gone to, wasn't he?
   A.  Yes, he was.
   Q.  That doesn't improve the unsatisfactory response,
       does it?
   A.  Sorry, doesn't improve?
   Q.  Well, the fact that he is the very person who is
       supposed to be dealing with it doesn't make it any
       better, it makes it look worse?
   A.  No, it makes it worse actually, yes.
   Q.  Okay.  Can we now look at lottery transactions, which
       you deal with at paragraph 23 and following in your
       statement at {E2/5/8}.
           Now, you refer to the introduction of PING in 2012,
       paragraph 25.
   A.  Yes.
   Q.  So you then explain how the system worked before the
       introduction of PING and after the introduction of PING.
   A.  Yes.
   Q.  Before PING, 26.1 {E2/5/8}, SPMs had to activate
       scratchcards on the lottery terminal and rem them in on
       the Horizon terminal completely separately?
   A.  Yes.
   Q.  So if there was any difference between the two
       activities that would create a problem or a discrepancy
       of some sort?
   A.  Yes, it would.
   Q.  And at the end of 26.3 you say:
           "There are many explanations as to why the two
       figures do not match, the simplest of which is that the
       branch had (i) activated a pack without remitting it
       into Horizon or (ii) remitted a pack into Horizon
       without activating it."
   A.  Yes.
   Q.  Now, if an SPM activated on a lottery terminal but
       didn't rem it in on the Horizon terminal, if
       reconciliation worked correctly there should ultimately
       be a transaction correction sent to the Horizon
       terminal, is that right?
   A.  Yes.
   Q.  And that in itself depends on the reconciliation process
       working correctly, doesn't it?
   A.  Yes.
   Q.  And that in turn depends on the integrity of the client
       data being provided by Camelot?
   A.  Yes.
   Q.  And the process for determining whether a TC should be
       issued within Post Office?
   A.  Yes.
   Q.  And at 26.4 you say:
           "By around 2010, the level of TCs from lottery
       scratchcards had begun to grow.  A practice had
       developed in some branches whereby they would activate
       the scratchcards but wait for the transaction correction
       a few days/weeks later rather than actively remitting in
       each pack on Horizon.  They did this because it saved
       them a little bit of time, but it caused their accounts
       to become confused (because they would be selling
       scratchcards without first having recorded the inbound
       scratchcard stock)."
           Now, pausing there, you are suggesting there that
       that was a sort of conscious decision, yes?  But it
       could have been through a lack of understanding, or
       a simple mistake, couldn't it?
   A.  Yes, it could have.
   Q.  And there were, as we have seen -- we saw in the common
       issues trial there was a problem with very large lottery
       TCs arriving weeks later.
   A.  Yes.
   Q.  Some of which were in error.
   A.  Yes.
   Q.  Some credits, some debits.
   A.  Yes.
   Q.  Both unsatisfactory?
   A.  Um ...
   Q.  When they are in error?
   A.  Oh, I see: yes.
   Q.  So that was why the policy we looked at during the
       common issues trial was introduced, to try and cater for
       that, not do them immediately before the end of
       a trading period?
   A.  Yes.
   Q.  Amongst other things.  And SPMs effectively became used
       to quite a lot of lottery TCs coming in, didn't they?
   A.  From the numbers that we were processing, yes, I would
       say that would be fair.
   Q.  And what about if a subpostmaster remmed in on the
       Horizon terminal but didn't activate on the lottery
       terminal, what would happen then?  You would get a TC as
       well?
   A.  Yes.
   Q.  And what changed when PING was introduced, which is
       paragraph 26.7, is that lottery scratchcards were
       actually activated on the lottery terminal and there was
       no need to rem them in on the Horizon terminal any more?
   A.  It was an automatic, so the TA would come.
   Q.  So the information would go from the Horizon terminal
       back to Camelot and be sent to Post Office?
   A.  No, it would go from the Camelot terminal.
   Q.  Sorry?
   A.  The Camelot terminal.  It would go -- so the activated
       pack would go to Camelot and then come into Post Office
       and then it would come back to us.
   Q.  Exactly, yes.
   A.  Yes.
   Q.  So the SPM activates the pack on the lottery terminal?
   A.  Yes.
   Q.  That tells Camelot it has been activated?
   A.  Why he.
   Q.  Camelot tell Post Office?
   A.  Yes.
   Q.  And Post Office send an automatic TA?
   A.  That's right.
   Q.  And that's generated automatically?
   A.  Yes.
   Q.  To avoid manual reconciliation difficulties?
   A.  Yes.
   Q.  And provided that the data coming back from Camelot is
       correct and there are no problems in the data
       transmission from the lottery terminal, that should be
       correct?
   A.  Yes.
   Q.  And overall it was an improvement for SPMs?
   A.  Yes.
   Q.  And simplified the process?
   A.  Absolutely.
   Q.  And PING was not limited to scratchcards, it also
       applied to lottery online sales as well?
   A.  Yes.
   Q.  It did take quite a long time to implement, didn't it?
   A.  I can't remember exactly how long from when it was first
       suggested.
   Q.  Well, I mean lottery had been in branches since before
       the introduction of Horizon?
   A.  Yes.
   Q.  So before 1999/2000.
   A.  I can't remember exactly when, but ...
   Q.  For example, Mr Bates, you might remember, was actually
       required to have a lottery terminal in a prominent
       position in his branch and then built it up and then
       Post Office told him he couldn't have it any more?
   A.  What I'm saying is I can't remember when we actually
       started Camelot business.
   Q.  I understand.  Have a look if we may at {F/506}.  And if
       we look at page 3 {F/506/3} we can see on the version
       control, the document history, at 2.2, the initial
       version of this document is 2008, isn't it?
   A.  Yes.
   MR JUSTICE FRASER:  Where are you looking at?
   MR GREEN:  2.2, my Lord, "Document history", 0.1.
   MR JUSTICE FRASER:  Yes.
   MR GREEN:  10 October 2008.  And if we look at page 9
       {F/506/9}, we have the context of that document.  The
       background is:
           "In line with the commercial contracts with clients,
       a number of settlements made by product and branch
       accounting are based upon data provided by the client.
       Such an example is Camelot, where settlement is based
       upon data captured by the Camelot terminal in outlets
       rather than the data being captured at transactional
       source by [Horizon]."
           HNG-X, yes?
   A.  Yes.
   Q.  And it then explains:
           "The client data is uploaded into POL FS,
       [Post Office Limited Financial System] and compared with
       the equivalent [Horizon] data which has to be manually
       input ..."
           That's into the old system:
           "... by the agent/counter clerk.  Ideally the data,
       when compared, should be the same but a number of
       conformance issues have been identified where
       agents/counter clerks do not perform end of day routines
       correctly ..."
           And so forth:
           "This difference may require the issuing of
       a transaction correction."
           That's what we have just covered.  Then if you go
       down to the paragraph that begins "Although", do you see
       that?
   A.  Yes.
   Q.  "Although client based settlements are not a preferred
       settlement option it is recognised that the data being
       provided by clients such as Camelot is robust,
       controlled by reference data, and more accurate than the
       [Horizon] data stream due to the conformance issues
       mentioned previously."
           Yes?
   A.  Yes.
   Q.  So this was a big vote of confidence in the accuracy, or
       in fact robustness as it is termed there, of the Camelot
       client data stream?
   A.  Yes.
   Q.  "The solution is to automate a process that converts the
       client data file (eg Camelot) into a data feed into the
       branch ... the branch will then accept the data and thus
       avoid the issues described above."
           Yes?
   A.  Yes.
   Q.  So the whole point of this is that it is robust data
       from Camelot, with automatic TAs that are issued which
       then come into the subpostmaster's account effectively
       the next day, yes?
   A.  Yes.
   Q.  And they automatically have to accept those?
   A.  Yes.
   Q.  And you then explain this isn't in fact implemented
       until 2012, if we look at paragraph 25 of your witness
       statement please {E2/5/8}.  Yes?
   A.  Yes.
   Q.  If we look please at {F/1539}, on page 5 -- you probably
       noticed on the first page it is authored by
       Gareth Jenkins.
   A.  Yes, I saw that.
   Q.  And at 2.5 it says:
           "It would appear that PING has gone live for Camelot
       (at last)."
           Yes?
   A.  Yes.
   Q.  And this was an internal recognition that it had taken
       rather a long time, hadn't it?
   A.  Yes.
   Q.  This document is dated 9 August 2016.
   A.  (Nods).
   Q.  And it is clear -- if we look at {F/506/11}, just to
       confirm I think what you have accepted, this is the PING
       project interfacing client data into POL's systems
       document.  We can see that there is no provision there
       at all for anything other than acceptance, yes?
   A.  Of the --
   Q.  Of the TA?
   A.  -- transaction, correct.  Acceptance, right, okay, yes.
   Q.  The transaction acknowledgement as opposed to
       transaction correction?
   A.  Yes.
   Q.  So if you look at number 3 it says:
           "The branch will be presented with the value to
       'accept' only ... no opportunity to write off or ask for
       evidence due to the inherent robustness of the Camelot
       data provided."
           Yes?
   A.  Yes.
   Q.  So that was how it was intended to work and we will have
       a closer look at that with some of the specific examples
       with individuals at the minute, but that's the
       background.
           Let's look at Mr Latif's situation.  If we look at
       paragraph 98 of your witness statement please on page 24
       {E2/5/24}.  Now, in that paragraph the explanation you
       have given there is essentially that there was a data
       entry error by Post Office?
   A.  Yes.
   Q.  Now, can you see that clearly in paragraph 98?
   A.  Yes.
   Q.  And it is prefaced with the words:
           "To be clear, this was a data entry error by
       Post Office and not an issue with Horizon."
   A.  Yes.
   Q.  Now, the point about that sentence is that sentence
       tells the court that there's no problem with Horizon at
       all, it's just someone manually entering the wrong
       figure at Post Office, doesn't it?  That's what that is
       saying?
   A.  It says it is by Post Office, yes.
   Q.  So it's suggesting it is a manual error and not a system
       problem?
   A.  Mm-hm.
   Q.  But we have just established from looking at the
       documents, and you have agreed, that this process was
       completely automated, wasn't it?
   A.  Automated, yes.
   Q.  So what you have said there cannot possibly be right,
       can it?
   A.  No, from that, yes, it's --
   Q.  It's wrong?
   A.  The TA is automated.  The TC is manually uploaded into
       the system.  Yes.
   Q.  Yes.  So would it be fair to correct that part of your
       statement --
   A.  It would be actually, yes.
   Q.  -- to say "To be clear, this is an issue with Horizon
       and not a data entry error by Post Office"; is that
       a fair correction to make?
   A.  I have made a mistake in that the way I have worded
       that, absolutely, yes.  So, yes, it is fair.
   Q.  Would you agree with the formulation I have given, or
       would you prefer something slightly different?  What
       would you want the court to note as your evidence?
   A.  So the automated -- my understanding is the
       transaction -- the TA is the information that comes from
       Camelot to us and then it is passed through into
       Horizon, so in that respect Horizon just conveys it, is
       my understanding, and the information that's come from
       Camelot in that respect would be incorrect.
   Q.  So the point is that either way, it is not a manual data
       entry by Post Office?
   A.  No, it's not.
   Q.  Whatever it is, it's definitely not that, you agree with
       that?
   A.  Yes, absolutely not.  Agree, yes.
   Q.  What it could be is some problem with the information
       somewhere between the terminal in the branch --
   A.  The lottery terminal, yes.
   Q.  The lottery terminal in the branch and the matters
       showing up on the face of the Horizon terminal in the
       branch?
   A.  Yes, I agree that.
   Q.  Somewhere there?
   A.  Yes.
   Q.  So it's definitely not a user error, is it?
   A.  No, that's not user error.
   Q.  Right.  And on the face of it, it at least suggests some
       doubt as to the robustness and integrity of the Camelot
       data coming through in that automated system, doesn't
       it?
   A.  I would say yes.
   MR JUSTICE FRASER:  Is it also the case that the branch
       could have challenged the TA?  Because I understood the
       previous document to say they couldn't.
   A.  So they have to accept the T on the system, but if they
       see something wrong on there they can actually get in
       touch with FSC.
   MR JUSTICE FRASER:  On the helpline?
   A.  Or direct to FSC, because you can go to the helpline --
       you can always go through the helpline NBSC into the
       finance service centre anyway.
   MR JUSTICE FRASER:  So looking at your final sentence of
       paragraph 98 ...
   A.  Yes.
   MR JUSTICE FRASER:  I understood the previous document to
       say that the TA had to be accepted by the branch?
   A.  It does have to be accepted --
   MR JUSTICE FRASER:  It does have to be?
   A.  On the system, yes.
   MR GREEN:  And that's just a straight acceptance, isn't it?
   A.  That's my understanding of that, yes.  If I recall
       correctly, on the screen that comes through it does say
       to make sure it is right and if not then you would get
       in touch with FSC.  I don't know the exact wording but
       that's my recollection, on the TA.
   Q.  You are aware that some people had problems with TAs
       quite a few times?
   A.  Yes, from looking at the transaction corrections.
   Q.  Some people had duplicate TAs, some people had all sorts
       of problems with the TAs coming through, yes?
   A.  I would also say it is fair that most branches would
       just accept it anyway.
   Q.  They would just accept it anyway some of them?
   A.  Because they -- you know, on the understanding that it
       is an automatic pull-through from what they have
       activated, I would -- my working assumption is most
       people just hit the button and accept it.
   Q.  Well, you have to accept it, don't you?
   A.  Yes, but what I'm saying is they would do it --
   Q.  And you wouldn't do anything more --
   A.  -- yes, without even really looking at it.
   Q.  I mean in fact we can see that in -- slightly ahead,
       I was going to take you to some of the documents that
       show that that was actually internally anticipated by
       Post Office, that what was anticipated was there would
       be a basket and people would generally just hit "enter"
       to accept all?
   A.  Yes.
   Q.  That's what was anticipated, wasn't it, and I think you
       very fairly volunteered that just now?
   A.  Yes.
   Q.  Let's just have a little look at the helpline log please
       at {F/1223/1}.  It is 12 June 2014.  I think that is
       a native file so we just need to download that and it
       will come up in a second.  If we look on I think the
       "Individual incidents" tab and then if we go to row
       431 -- if we perhaps just type 431 in the top left-hand
       box.
           At row 431 you can see there National Lottery in
       column H, transaction corrections is column I.  If you
       scroll to the right slightly please, "Lottery TA issue",
       "PM says has still got a discrepancy due to the TA
       problem."
           Yes?
   A.  Yes, I see that.
   Q.  If we go down to row 2375 and you see "TC for
       National Lottery", you can see in column L:
           "TC not received for the TA that was duplicated from
       last Saturday for National Lottery.  Printed off TC
       report and on there as processed."
           So there's an issue there and if we just look at
       3176:
           "Lottery TAs: PM still not received TC for the TA
       problem."
           These are helpline call logs for the week of -- we
       just looked at entries for 9 June, 10 June and 11 June.
       These are the sort of problems that you fairly accepted
       people had from time to time?
   A.  Yes.
   Q.  And if we look at the document at -- well, I might be
       able to take it more quickly.  There were also problems
       that people were receiving TCs when they didn't know
       what it was for in relation to National Lottery.  That
       was internally acknowledged, we saw it in the common
       issues trial?
   A.  Yes, the lack of explanation on the TC.
   Q.  Yes.  And then people were raising queries about errors
       relating to the downloaded duplicate TA files in June.
   A.  Yes.
   Q.  2014.  Do you remember that?
           So there was a problem with duplicate TAs, wasn't
       there, from time to time?
   A.  From time to time, yes.
   Q.  Let's look please at {F/965/22}.  TAs were later
       introduced for Post and Go and Paystation transactions,
       weren't they?
   A.  Yes.
   Q.  And if you look on page 22 there you will see:
           "Crowns have incurred discrepancies as a result of
       missing TAs ..."
           Do you see the second box down, "Post and Go TA
       missing or duplicated"?
   A.  Yes.
   Q.  This is under "Current issues".  This is an integral
       fraud and conformance report.
   A.  Okay.
   Q.  So that's Crown offices, that's not SPMs, is it?
   A.  No, that's directly ...
   Q.  Yes, so they were experiencing them too?
   A.  Yes.
   Q.  And if we look back -- let's look at {F/1228} please and
       what we're going to here is a branch user forum record
       of input from users of Horizon and once we have loaded
       it we're going to look at line 74.  If we can go down to
       line 74 and you can see there -- go slightly to the
       right so we can get the whole ...
           "Lottery and Paystation: why can't I sell lottery on
       Horizon as other retailers do and similarly why is not
       the Paystation functionality contained on Horizon -
       saves messing around with separate tills or pots of cash
       with TAs the following day or even later."
           And so forth.  And he makes the point about the
       costs being borne by the subpostmasters at the bottom.
           Just pausing there, the lottery terminal and
       Paystation were not within the original design of the
       Horizon system, were they?
   A.  No, they weren't.
   Q.  Even though lottery, for example, was in use prior to
       the introduction of Horizon?
   A.  Yes, but it has always been Camelot's.
   Q.  So they were added on and that introduced the pre-PING
       issue of mistakes between the terminals?
   A.  Yes.
   Q.  And then questions about whether it is user error or
       something else that's gone wrong, there's an opacity
       there, and post PING it introduced the issue of
       duplicate TAs and difficulties with the integrity of the
       data stream all the way through?
   A.  Yes.
           Can I just say, when -- so when we brought in
       Horizon very early on, the format of a post office was
       very much that the post office was behind a counter,
       screened counter, and the lottery terminal was always
       designed to be on the retail counter and the Paystation
       was -- it was more of an out of hours, so it was never
       built as one, it was literally a very different part of
       the post office itself and the retail, so it was very
       different.
   Q.  It's a fair observation, but equally I think you would
       accept it is a fair observation to remind oneself that
       one of the conditions of appointment that Mr Bates had
       to agree to, from Post Office, was that he would operate
       a Horizon terminal from which Post Office would
       profit -- not in his shop for him to profit from it, but
       on behalf of Post Office --
   A.  Yes.
   Q.  -- he would operate a Horizon terminal in his branch.
       It was actually a contractual requirement for him, yes?
   A.  Run through Post Office, yes.
   Q.  So that --
   A.  I'm just explaining why it wasn't part of the -- you
       couldn't add it on to Horizon, it was very -- physically
       very different, that was what I was saying.  That was my
       point I was making.
   Q.  He was required to have it in a prominent position at
       his counter, wasn't he?
   A.  In some post offices it was.  All I was explaining was
       that in the vast majority it was physically apart from,
       distanced from the main counter, which was behind
       a screen.  That was the only opponent I was making.
   Q.  I understand.  Let's have a look at what you say about
       failed reversals please.  This is paragraph 154 of your
       witness statement {E2/5/34}.  Now, here you have moved
       on to respond to go what Mr Coyne says.  Do you see
       that?  It is on page 34.
   A.  Yes.
   Q.  E2/5/34, paragraph 154.  It is under the heading just
       above 153 "Mr Coyne's report", just to orientate you
       where you are in your witness statement.
   A.  Yes.
   Q.  And you say:
           "I address below a number of factual points raised
       by Mr Coyne in his report."
           Just pausing here.  Mrs van den Bogerd, you
       obviously took a lot of care over this witness
       statement, yes?
   A.  Yes.
   Q.  That's fair?
   A.  Yes.
   Q.  And probably had even greater focus perhaps and greater
       time to devote to this than someone dealing with an SPM
       ringing up might have?
   A.  Sorry, what do you mean?
   Q.  You had more chance to look into all these things you
       have dealt with, a greater opportunity to research and
       look into these things, to provide this witness
       statement?
   A.  In some areas, yes, I have, yes.  Not in all because of
       the way things were changing.
   Q.  But where you have been able to comment on a document
       you have been able to consider it carefully and so
       forth?
   A.  I have tried to, yes.
   Q.  And let's look at what you say about failed reversals.
       You say:
           "At paragraph 5.175 of his report, Mr Coyne has
       referred to a report prepared by Helen Rose dated
       12 June 2013 in the context of failed reversals.  The
       extracts taken from the report by Helen Rose referred to
       by Mr Coyne are taken out of context and mistakenly
       claim that the relevant reversal was issued in error by
       Horizon not the subpostmaster.  The Rose report makes it
       clear that:
           "The concerns were based on the fact that reversal
       were not being shown on the particular data sets
       reviewed/reports typically run by subpostmasters in
       branch on Horizon;
           "Transaction reversal data can be extracted from
       Horizon ... the issue was therefore surrounding how the
       transaction reversals were displayed/accessible in
       branch and that there was no issue with Horizon itself."
           Now, just pausing there, that is not something that
       you have changed, is it?
   A.  No.
   Q.  Let's have a look please if we may at the underlying
       document, the Helen Rose report itself, which is
       {F/1082}.  It is dated 12 June 2013.  Can you see that?
   A.  Mm-hm, I can.
   Q.  Is this the document you looked at?
   A.  Yes.
   Q.  Now, if we look at page 2 of that document please
       {F/1082/2}, it is obviously headed "Confidential and
       legally privileged" and it is "Horizon data - Lepton",
       the subpost office, yes?
   A.  Yes.
   Q.  Executive summary:
           "A transaction took place at Lepton
       [subpost office] ... for a British Telecom bill payment
       for £76.09; this was paid for by a Lloyds TSB cash
       withdrawal for £80 and change a given for £3.91.  At
       10.37 on the same day the British Telecom bill payment
       was reversed out to cash settlement."
           Looks pretty suspicious, doesn't it, so far?  Yes?
   A.  The running order of the times, do you mean?
   Q.  Does what we just read there look a bit suspicious?
   A.  Sorry, is this --
   Q.  Does it make it look as if it something strange has
       happened?
   A.  Oh, right, yes.
   Q.  The bill payment has been reversed out for cash?
   A.  Yes.
   Q.  "The branch was issued with a transaction correction for
       £76.09 which they duly settled ..."
           So all that means is that they press "settle"
       centrally, doesn't it?  They can't dispute it -- we have
       covered this lots of times.  They can't dispute it on
       Horizon, so all the postmaster can do is ring up and
       deny --
   A.  I'm not sure that he actually settled in this one,
       I think he made it good.  I can't remember actually.  In
       terms of the way she is referring to language.
   Q.  Let's have a look.
   A.  Because it doesn't say settle centrally.
   Q.  Okay, maybe settled it in cash.
   A.  Yes, that's my understanding.
   Q.  One way or the other, what we then see is:
           "The postmaster denied reversing this transaction
       and involved a forensic accountant as he believed his
       reputation was in doubt."
           So he is having to instruct a forensic accountant to
       help him over £76.09, is what we've got so far in the
       story, isn't it?
   A.  As telling in the first review, yes.
   Q.  Let's look at reviewing the Credence data.  Now, just
       pausing there, we know, don't we, from Tracy Mather's
       witness statement that the Credence data records the
       actual key strokes that have been used?
   A.  Yes.
   Q.  So that's a really important source if you are in doubt
       about what's actually been done by the postmaster, if
       they say "I had help from the helpline and I did
       a complicated set of key strokes", look at the Credence
       data, that will tell you, yes?
   A.  Yes.
   Q.  And so Helen Rose very sensibly looked at the Credence
       data and said:
           "... it clearly indicates that the reversal was
       completed by JAR001 (postmaster) ..."
           Yes?
   A.  Yes.
   Q.  At 10.37.
           "... and was reversal indicator 1 (existing
       refersal) and settled to cash.  An existing reversal is
       where the session number/automated payment number has to
       be entered to reverse the item."
           Yes?
   A.  Yes.
   Q.  "The Fujitsu logs were requested for this branch, but
       whilst waiting for these to arrive communications took
       place with Gareth Jenkins at Fujitsu for more details to
       gain an understanding of what had occurred at this
       branch."
           And then we get questions and extracts in various
       emails in response.  It doesn't come up very clearly,
       but black is the question and blue is the answer.
   A.  I can see that on here, yes.
   Q.  "Question - I am requesting Fujitsu logs for Lepton ...
       to look at a reversal that the postmaster denies
       transacting, do I need to request further details and
       also could you explain what happens when the system
       fails.  (Gareth looked at data at his end prior to me
       receiving the Fujitsu logs ...)
           "Answer - this shows that session 537803 was
       successfully saved to the [branch database] ..."
           Yes?
   A.  Yes.
   Q.  "... but when the user JAR001 ..."
           Which we know is the subpostmaster:
           "... logged on again recovery reversed the session
       in session 537805."
           Yes?
   A.  Yes.
   Q.  So what the Credence data showed was it had been an
       actual reversal by the subpostmaster, but in fact when
       Mr Jenkins got into it he found it was actually a system
       recovery that had reversed the session, didn't he?
   A.  Yes.
   Q.  And so he then goes on to say:
           "It isn't clear what failed, but if it was a coms
       error, then the system would have printed a disconnected
       session receipt and the clerk should have given the
       customer £80 and told him his bill was unpaid.  The fact
       that there is no indication of such a receipt in the
       events table suggests the counter may have been rebooted
       and so perhaps may have crashed in which case the clerk
       may not have been told exactly what to do."
           Yes?
   A.  Yes.
   Q.  "The reversal was due to recovery (Counter Mode Id =
       118) so this was not an explicit reversal by the clerk.
       This scenario is fairly rare so it is certainly quite
       easy for the clerk to have made a mistake and either he
       or the customer could be in pocket/out-of-pocket
       (depending on exactly what happened!)."
           Then this:
           "The system is behaving as it should."
           So there were issues, weren't there, where a problem
       could arise for a subpostmaster by design of the system;
       were you aware of that?  There was a whole category of
       PEAK codes for faults which are agreed between Fujitsu
       and Post Office to just stay like that as part of the
       design?
   A.  Okay.
   MR JUSTICE FRASER:  You actually put two questions there.
   MR GREEN:  I'm sorry.
   MR JUSTICE FRASER:  On the second one you went on to codes,
       so you really need to split them.
   MR GREEN:  That's entirely my fault.  I'm sorry,
       Mrs van den Bogerd.
           Let's take it in stages.  You can see there it says
       "The system is behaving as it should"?
   A.  Yes.
   Q.  That would be problematic for a subpostmaster,
       wouldn't it?
   A.  The way Gareth describes it here would be, yes, because
       what he says is that it would have printed the session
       receipt but it doesn't seem as if it did, when actually
       disconnection transaction receipts were actually printed
       in this example and a recovery receipt was printed.  But
       that's not referred to in here.
   Q.  Well, let's just have a look at that.  Let me just ask
       you my second question and then we will go on to probe
       that with more care.  The second point is are you aware
       of a closure code for Fujitsu for PEAKs which refers to
       faults which are known in the Horizon system but agreed
       between Post Office and Fujitsu to stay there?
   A.  I'm not aware of a closure code.
   Q.  You didn't know about that?
   A.  No, I don't know about that.
   Q.  Okay.
           Let's look at the question at the bottom of
       {F/1082/2} which we are on.  This is Helen Rose:
           "I can clearly see the recovery reversal on the
       Fujitsu logs received, but would this have been clear
       had we not previously discussed this issue?"
           That's her question.  And the answer comes
       {F/1082/3}:
           "Note that the standard ARG spreadsheet may not make
       it easy to confirm that the reversal was part of
       recovery, but the underlying logs used to extract them
       can show it."
   A.  Yes.
   Q.  So what Gareth Jenkins is pointing out to Helen Rose is
       that the standard ARQ data would certainly not make it
       clear but the underlying logs used to extract the ARQ
       data can show it, has the capability of showing it, yes?
   A.  Yes.
   Q.  And then he then talks about the Excel spreadsheets of
       what appear to be the underlying data and he says they
       are part of the standard ARQ returned and it says:
           "Rows 141 ..."
           In the second line of that paragraph:
           "Rows 141 to 143 of 4 to 25 October ... clearly show
       a reversal.  Also row 70 of [the events data] shows that
       session 537803 ... has been recovered and this event has
       the same time stamp as the reversal session.  Also
       row 71 ... shows that a receipt was generated from the
       session ... (not explicitly, but it was the only session
       at that time).  This receipt would have told the user
       that a roll back had taken place (but the logs don't
       make that explicit).  If that is sufficient for [your]
       purposes then you do have all you need in the standard
       ARQ."
           Then he says:
           "However what I was able to confirm from my look at
       live data a couple of weeks ago and is also held in the
       underlying raw logs is confirmation that the reversal
       was generated by the system (and not manually by the
       user).  What might also be available in the underlying
       logs is whether or not the system was rebooted -
       I suspect it was but have no evidence one way or the
       other (and it isn't in what was extracted this time
       either).  I can confirm that the user did log on
       again ..."
           And so forth.  Now, just pausing there, you would
       agree, wouldn't you, that if we go back to paragraph 154
       of your witness statement at page 34 {E2/5/34}, let's
       look at what you say there, top of page 35 if we may
       {E2/5/35}:
           "... mistakenly claim that the relevant reversal was
       issued in error by Horizon, not the subpostmaster."
           Do you see that?
   A.  Yes.
   Q.  Well, we can see from what Gareth Jenkins has actually
       said in the actual document you are looking at, that
       that's wrong, can't we, the reversal was not done by the
       subpostmaster, it was done by the system?
   A.  It was done by the system absolutely, yes.
   Q.  It was done by the system and not by the subpostmaster?
   A.  Yes, it was done by the system, yes.
   Q.  So it's fair to say that what we have at paragraph 154
       is wrong, isn't it?
   A.  "... the relevant reversal was issued in error by
       Horizon ..."
           It wasn't issued in error, it was actually issued by
       Horizon.  So I am obviously not be making myself clear,
       but yes, there's no question that that was done as part
       of that recovery and it was system generated.
   Q.  Well, look, Mrs van den Bogerd, be fair, what you have
       written there looks to a natural reading as suggesting
       that Mr Coyne has mistakenly claimed that the relevant
       reversal was issued in error by Horizon, not the
       subpostmaster.  The contrast is who has issued the
       relevant reversal.
   A.  I understand.  That wasn't what I was saying.  There's
       no question that it was issued by the system, it was
       definitely generated by the system and if that's -- and
       I have obviously not explained myself very clearly.
   MR JUSTICE FRASER:  So what was the mistake then that you
       refer to at the top of page 35?
   A.  That the reversal was an error, because the reversal
       itself wasn't an error and the receipts were printed.
       So in fact it was part of the recovery, the system did
       go down and the BT bill actually was reversed at part of
       that.  So I have just misread and misexplained that
       then, because I was -- there was no doubt in my mind
       what had happened there -- because this was part of the
       spot review by Second Sight, so I was familiar with the
       detail at the time and from Helen's report what she is
       saying is that it is not obvious that that reversal has
       been done by the system because it is against the
       Horizon user's ID and clearly it would be better if it
       was obvious and I fully agree with that.
   Q.  But to be fair, Mrs van den Bogerd, the point that this
       Helen Rose report makes is almost exactly the opposite
       point to the one you were making in your statement.  The
       point that this report makes is that it might appear to
       have been done, on the initial logs -- yes?  If you just
       read the beginning it shows -- look at page 2, let's go
       back to page 2, {F/1082/2}.  Do you see the first
       answer:
           "This shows that session 537803 was successfully
       saved ... but when the user JAR001 logged on again
       recovery reversed the session ..."
           Do you see that?
   A.  Yes.
   Q.  So it was the system that reversed the session not the
       subpostmaster, wasn't it?
   A.  Yes, it was.
   Q.  And that's the opposite to what you are saying at the
       top of page {E2/5/35}?
   A.  Okay, but that's not what I meant, because I clearly
       have always known that that was a Horizon system
       generated recovery and the reversal was part of that
       recovery.
   Q.  But you were suggesting there that Mr Coyne, the expert,
       had mistakenly claimed something.  You weren't just
       casually saying it in your witness statement, you were
       pointing out that Mr Coyne had got it wrong and
       positively suggesting the reversal had been done by the
       subpostmaster, weren't you?
   A.  That wasn't my intention and I'm sorry -- clearly I have
       not explained myself very well there at all, because
       I was in no doubt at all that that had been generated by
       the system.
   Q.  Well, let's go back, if we may, to F/1082 and look at
       page 3 please {F/1082/3}.  Halfway down there is
       a question:
           "I can see where this transaction is and now
       understand the reason behind it.  My main concern is
       that we use the basic ARQ logs for evidence in court and
       if we don't know what extra reports to ask for then in
       some circumstances we would not be giving a true
       picture."
           Do you see that?
   A.  Yes.
   Q.  And that's effectively what happened at the top of your
       witness statement -- on that page in your witness
       statement, isn't it?  You were not giving a true picture
       because you had got the wrong end of the stick?
   A.  Re-reading that, yes, that's correct.
   Q.  And then Helen Rose says:
           "I know you are aware of all the Horizon integrity
       issues ..."
           Can you tell his Lordship what those were?
   A.  At this point it would have been the fact that
       Second Sight were working with us investigating some of
       the cases.  That's all I can think that Helen means at
       that point.
   Q.  Did you have any particular ones in mind at that point?
   A.  It would have been this one particularly, because as
       I said this was one of the spot reviews, which is one of
       the early ones that Second Sight investigated before we
       actually went into the mediation scheme.
   Q.  Okay.  And it says:
           "... and I want to ensure that the ARQ logs are used
       and understood fully by our operational team who have to
       work with this data both in interviews and in court."
           And she says:
           "Just one question from my part - if the reversal is
       system created but shows as an existing reversal, could
       this not be reflected with a different code ie SR
       (system reversed) to clear up any initial challenges.
       My feelings at the moment are not questioning what
       Horizon does as I fully believe that it is working as it
       should, it is just that I don't think that some of the
       system based correction and adjustment transactions are
       clear to us on either Credence or ARQ logs."
           That's what she is saying, yes?
   A.  Yes.
   Q.  And that was a fair observation, wasn't it, by her?
   A.  That was a fair observation, yes.
   Q.  About the shortcomings of Credence and ARQ logs, yes?
   A.  Yes.
   Q.  And Gareth Jenkins' answer:
           "I understand your concerns.  It would be relatively
       simple to add an extra column into the existing ARQ
       report spreadsheet, that would make it clear whether the
       reversal basket was generated by recovery or not.
       I think this would address your concern.  I'm not sure
       what the formal process is for changing the report
       layout.  Penny, can you advise as to the process: is
       this done through a CR?"
           Do you know what a CR is?
   A.  Change request.
   Q.  Change request, okay.  Then at the bottom:
           "I do believe that the system has behaved as it
       should and I do not see this scenario occurring
       regularly and creating large losses.  However, my
       concerns are that we cannot clearly see what has
       happened on the data available to us and this in itself
       may be misinterpreted when giving evidence and using the
       same data for prosecutions.
           "My recommendation is that a change request is
       submitted so that all system created reversals are
       clearly identifiable on both Fujitsu and Credence."
           Do you know if that change request was acted on
       after that?
   A.  I don't believe it has been acted on.
   Q.  You don't believe it has been?
   A.  I don't believe so.
   Q.  Can we move please now --
   MR JUSTICE FRASER:  I have just got a question on this
       document which I'm going to ask now rather than at the
       end.  Could we go back one page please to page 2
       {F/1082/2}.  You will see at the bottom of that page
       there are three paragraphs in blue and then a paragraph
       in black.  In the bottom paragraph in blue it says:
           "This scenario is fairly rare so it is certainly
       quite easy for the clerk to have made a mistake and
       either he or the customer could be in
       pocket/out-of-pocket (depending on exactly what
       happened!).  The system is behaving as it should."
           Do you see that?
   A.  Yes.
   MR JUSTICE FRASER:  Am I right that the alternatives are --
       "he" meaning the subpostmaster.
   A.  Yes.
   MR JUSTICE FRASER:  So the subpostmaster being in pocket,
       that's one of the alternatives, is that right?
   A.  Yes.
   MR JUSTICE FRASER:  The subpostmaster being out-of-pocket.
   A.  Yes.
   MR JUSTICE FRASER:  Is that the other alternative?
   A.  Yes.
   MR JUSTICE FRASER:  And the other side of those two coins is
       either the customer being out-of-pocket or the customer
       being in pocket.
   A.  That's correct.
   MR JUSTICE FRASER:  Are those the options?
   A.  Yes, those are the options.
   MR JUSTICE FRASER:  Do I appear to have missed any out?
   A.  No.  That would have been -- on the disconnected session
       receipt it would have said on there whether to -- on the
       recovery whether the payment should have been made to
       the customer or not and that's what the person in branch
       would use to guide them with that printed receipt.  But
       you haven't missed any options, that's --
   MR JUSTICE FRASER:  I haven't missed -- I have covered the
       basics?
   A.  Yes, you have.
   MR JUSTICE FRASER:  And those are examples of the system
       behaving as it should?
   A.  Yes.  With the disconnected session and then the
       recovery receipt it would be, yes.
   MR JUSTICE FRASER:  Over to you, Mr Green.
           Just to clarify one point, you look at -- if we
       look -- let's do it from here while we are on there.  We
       are on this page.  Can you see the paragraph three up
       from the bottom:
           "It isn't clear what failed ..."
           Do you see that?
   A.  Yes.
   Q.  "... but if it was a coms error, then the system would
       have printed a disconnected session receipt ..."
   A.  Yes.
   Q.  And then the next sentence:
           "The fact that there is no indication of such
       a receipt in the events table ..."
           That suggests that the receipt you are talking about
       isn't recorded in the events table, doesn't it?
   A.  It does and that's the bit I don't understand on this,
       because the receipts were printed.
   Q.  But look at 155 of your witness statement {E2/5/35}.
       You say:
           "There is therefore no indication that the reversal
       was not notified to the subpostmaster."
   A.  Yes.
   Q.  But that's exactly the point that's being made,
       isn't it?
   A.  No, no.  The point is that he had the receipts.
   Q.  Well, I will give you one last -- I'm not going to press
       it too long, but let's go back to {F/1082/2}.  I just
       afford you the opportunity to reconsider that answer.
       You are on oath.  Have a look please --
   A.  Sorry, sorry.  So from this what he is saying is that
       the receipts would have been printed.
   Q.  Well, just look at the third paragraph up from the
       bottom.  If it is because you have misunderstood what it
       says let's clarify it now.  Third paragraph up from the
       bottom, I have just shown you the third line of that:
           "The fact that there is no indication of such
       a receipt in the events table suggests the counter may
       have been rebooted and so perhaps may have crashed in
       which case the clerk may not have been told exactly what
       to do."
           Do you see that?
   A.  Yes, I do see that.
   Q.  Right.  Now, just to be fair to you let's go back to
       your statement please, page 35 {E2/5/35}.  Let's take
       the context very carefully and let's look at 154.3
       please:
           "The issue was therefore surrounding how the
       transaction reversals were displayed/accessible in
       branch ..."
           Yes?
   A.  Mm-hm.
   Q.  Now, that is correct in part, isn't it, because there
       was an issue about what the subpostmaster could see in
       the branch, yes?
   A.  Okay, yes.
   Q.  Right, and you then say there is no issue with Horizon
       itself.  Let's just park that.  Look at 155.  You then
       say:
           "There is therefore no indication that the reversal
       was not notified to the subpostmaster.  When recovery
       was carried out a discontinued session receipt would
       have been printed and messages would have been clearly
       displayed to the user in branch during the recovery
       process."
           Now, go back please, having that in mind, to what
       I have just shown you, {F/1082/2}.  Third paragraph up
       from the bottom, three lines down:
           "The fact that there is no indication of such
       a receipt in the events table suggests the counter may
       have been rebooted and so perhaps may have crashed in
       which case the clerk may not have been told exactly what
       to do."
           So it's fair to say, isn't it, that there is an
       indication that the reversal may not have been notified
       to the subpostmaster?
   A.  So from reading that then absolutely.  All I'm saying is
       I know that those receipts were printed because I have
       seen them on an email between Second Sight and the
       postmaster, so I know they were actually printed.
       That's my only -- that's what I'm saying.  From reading
       what you said, absolutely agree, but all I'm saying here
       is -- and clearly Helen Rose and Gareth at this point
       didn't know that those receipts were actually printed.
   Q.  So you are actually giving evidence that contradicts the
       document that you have referred to?
   A.  So all I'm saying is I know -- I know that these
       receipts were printed in this scenario.
   MR JUSTICE FRASER:  Do you mean in this particular instance?
   A.  Yes.  So the Armstrong -- Mr Armstrong is the
       postmaster, Lepton is the branch.  This was -- it was
       a spot review undertaken by Second Sight early on before
       we went into the mediation scheme.  Now, when they did
       the initial investigation with the postmaster he didn't
       know he had the receipts, so it was written in that way
       and then when Second Sight went back to him before and
       said -- and he actually found the receipts.
   MR JUSTICE FRASER:  Understood.
   A.  Sorry if I have misled, I didn't intend to, but it's
       just that I know those receipts were printed and it's --
   MR GREEN:  I'm glad we took it carefully to work out
       precisely what you are saying.
   MR JUSTICE FRASER:  But as far as this document is
       concerned, if I understand what you have just said is
       your knowledge is wider than shown on this document
       alone --
   A.  Yes.
   MR JUSTICE FRASER:  -- because of the other documents you
       have mentioned.
   A.  That's what I'm saying, yes.
   MR JUSTICE FRASER:  Is that a useful place to stop and you
       perhaps --
   MR GREEN:  I've got --
   MR JUSTICE FRASER:  All right, put your question, but it
       sounded to me as if there was some documentary
       assistance stage right which you just might want to look
       at.
   MR GREEN:  I can understand.
           But, Mrs van den Bogerd, just two separate points in
       relation to this.  It may be that the subpostmaster was
       notified that a reversal had taken place, yes?
   A.  Yes.
   Q.  From a receipt which they don't refer to?
   A.  Disconnected session and the recovery receipts.
   Q.  But there's nothing in that documentation or what would
       have been displayed at the branch to have shown the
       subpostmaster how that was being recorded as who had
       done it, whether it was the subpostmaster or the system,
       is that fair?
   A.  I'm trying to recall the receipts.  I think that's fair.
   Q.  Thank you.
   A.  Yes, it -- the receipt would have shown it happened, but
       not --
   Q.  At whose behest?
   A.  I believe that's correct but I would need to check the
       receipt, but I do believe that's correct.
   Q.  And we have now seen it was a system recovery not
       a subpostmaster recovery?
   A.  It was always a system recovery.
   Q.  I'm grateful.
           My Lord, is that a convenient moment?
   MR JUSTICE FRASER:  Yes.  We will actually come back at 5
       past 2 just in case you need to pursue some documents.
           Same form as before, break until 5 past 2, please
       don't talk to anyone about the case and we will resume
       then.
   (1.00 pm)
                    (The luncheon adjournment)
   (2.05 pm)
   MR GREEN:  Mrs van den Bogerd, could you look please at
       {E2/5/35}, paragraph 156 of your witness statement.  You
       will see there you deal with changes to improve Horizon
       in branch and you make a reference to the fact Mr Coyne
       has picked up on the phrase "relatively small changes to
       Horizon could avoid errors/mistakes made in branch", do
       you see that?
   A.  Yes.
   Q.  Could we look please at {F/1258} and this is a document
       with a 1 October 2014 Opus date but "September 19"
       written on the front.  It is, as far as you remember,
       a 2014 document, isn't it?
   A.  Yes.
   Q.  And look at page 2 please {F/1258/2}:
           "There are five staff related wave 1 initiatives
       within finance with a cumulative run rate impact of 51
       FTE and £4.8 million ..."
           That's 51 full-time employees, isn't it?
   A.  That's correct.
   Q.  Let's just look at the initiatives.  "Near-term process
       improvements", and then "Demand reduction - make small
       change in Horizon to reduce errors in branch", 3 "Demand
       reduction - reduce housekeeping and other losses,
       4~"Demand reduction - negotiate Santander contract to
       remove manual transcription".  That fourth one is
       a reference to not manually transcribing data provided
       by Santander, isn't it?
   A.  Yes, moving away from paper based.
   MR JUSTICE FRASER:  Could you keep your voice up.
   A.  Sorry.
   MR GREEN:  And if we look at number 2, which is the one I'm
       inviting you to focus on, this is making a small change
       in Horizon to reduce errors in branch, yes?
   A.  Yes.
   Q.  If we go a few pages forward to page 18 please
       {F/1258/18}, do you see there under the "Executive
       summary":
           "A significant proportion of demand at FSC is driven
       by errors/mistakes made in branch with entering in data
       into Horizon.  Part of these errors can be avoided with
       relatively small changes to Horizon (eg related to two
       part transaction for bureau pre-order, mismatches with
       AEI third party kit ..."
           AEI is auto exchange information, isn't it?
   A.  Yes.
   Q.  That's automatic exchange of information with third
       parties?
   A.  Yes.
   Q.  "... top up and redemption ..."
           And just look at the end of that box:
           " ... the 'quantity' field for cheques and miskeys
       in general ..."
           Then:
           "Ease of implementation: medium.  There exists an
       interdependency on IT for changes in Horizon to reduce
       errors coming from upstream."
           Yes?
   A.  Yes.
   Q.  And so there is an interdependency with upstream data
       errors because if you have got problems in the data for
       example coming in from Camelot, that's not something
       that the miskeying problem is going to deal with on its
       own, but it is right, isn't it, that pretty much since
       Horizon was introduced it had been noted that miskeying
       did represent a source of errors in the branch --
       I think you even said that in your witness statement for
       the first trial?
   A.  That's correct.
   Q.  And we can see that the changes that were being
       contemplated in October 2014 are said to be relatively
       small changes, yes?
   A.  Yes.  From this, yes.
   Q.  And could we look please at {F/476/1}.  Now, this is
       a "Summary of IS review".  What does IS stand for?
   A.  I would assume information security.
   Q.  And who was Peter Laycock?
   A.  I don't know.  I've never come across him before.
   Q.  The document itself says it is dated 2008.  The first
       item he deals with there, number one,
       "Deliberate/accidental miskeying (clerk)", that's
       a reference to a subpostmistress or subpostmaster,
       isn't it?
   A.  Yes, somebody in branch, yes.
   Q.  Or their assistant?
   A.  Or their assistant, yes, but it could also include
       a Crown branch as well.
   Q.  It could be a Crown branch.  And the recommendation is:
           "Double entry and cross-validation of freeform
       transaction values at the counter for all financial
       products ..."
           So that would be one a way of dealing with it,
       wouldn't it?
   A.  Yes.
   Q.  An alternative might be simply to say any transaction
       which is over £500 you just have to click "confirm" or
       "enter" twice?
   A.  On the --
   Q.  That's another way of doing it?
   A.  On the Horizon system you mean, do you?
   Q.  Yes, on the Horizon system?
   A.  Yes.
   Q.  That would avoid some quite large losses potentially,
       wouldn't it?
   A.  Yes.
   Q.  And the operational impact:
           "May add some time to process at counter (minimal)."
           Yes?
   A.  Yes.
   Q.  "System impact":
           "Change to Horizon required to prompt for double
       entry and cross-validation should not be overly
       complicated or expensive ..."
           And then "Opportunity" --
   MR JUSTICE FRASER:  What does "PL" mean in brackets?  If you
       don't know, don't worry.
   A.  I don't know exactly what that means in this context.
   MR GREEN:  I think, my Lord, we have read it as possibly
       referring to comment by Peter Laycock himself, but we're
       not sure.
   MR JUSTICE FRASER:  Fine.
   A.  Yes.
   MR GREEN:  Then "Opportunity":
           "c.80% reduction in disputes and claims - saving
       800k per annum in compensation, agent debt and business
       write-off for bill payments ..."
           Etc and then:
           "Staff reduction potentially at P&BA."
           And if we go over the page, "Other business
       benefits":
           "Major improvement of point of transaction data
       integrity."
           That would be a good thing, wouldn't it, to improve
       point of transaction data integrity?
   A.  Yes.
   Q.  "Double entry, customer sight and validation ... will
       minimise balancing issues ..."
           That's a good thing?
   A.  Yes.
   Q.  Deter fraudulent behaviour, that's a good thing?
   A.  Yes.
   Q.  And help to target investigations where there's lower
       volume of those sort of problems?
   A.  Yes.
   Q.  And then "Operational efficiency", you see fourth bullet
       point:
           "Less balancing errors, productivity savings, less
       calls to NBSC and partner banks."
           Yes?
   A.  Yes.
   Q.  Good thing?
   A.  Yes.
   Q.  "Increased confidence to the customer, partners,
       clients, agents ..."
           Et cetera.  Yes?  All pretty sensible and good
       ideas?
   A.  Yes.
   Q.  And then we say -- we just note at the bottom there:
           "Please Note – financial savings may have increased
       since 2008."
           So that tends to suggest the document had been
       updated.  Is this a document you had ever seen before or
       not?
   A.  No.
   Q.  But you were aware of miskeying as an ongoing issue?
   A.  Yes.
   Q.  And if we look at {F/994/4} please.  In fairness to you,
       Mrs van den Bogerd, sorry, could we look at page 1
       {F/994/1} so Mrs van den Bogerd can see what it is.
       This is the miskeyed project feasibility study in 2012,
       do you see that?
   A.  Yes.
   Q.  And if we go please to page 4 {F/994/4}, you will see
       under the background section at 3.1:
           "As part of the P&BA centre of excellence drive, one
       of the areas of concern is the number of instances of
       mis-keyed transactions that occur and much to the
       detriment of P&BA.  A mis-keyed transaction occurs when
       an incorrect value is input by the counter clerk, which
       causes a poor customer experience.  The mistake can have
       a significant impact on the branch and resource is
       required in P&BA to manage the client and address the
       error.  A very large value mis-keyed transaction will
       put the viability of a branch in doubt."
           Yes?
   A.  Yes.
   Q.  And we've got some statistics about how much it is
       costing a year in terms of total of investigations and
       so forth.
           So on any view, that had potentially a pretty
       serious impact for a branch, didn't it, and that was
       internally recognised in that document?
   A.  Miskeying, yes.
   Q.  And on page 10 {F/994/10}, there's a risk and issues
       section at 5.5:
           "If nothing is done to prevent miskeying within
       Horizon the current bill of £10 million will spiral out
       of control."
           "Recommend resolve.
           "Make it happen to improve the situation where
       miskeying occurs."
           Then:
           "Many of the workforce is not aware of the miskey
       situation.  It is part of this project to get the
       message across to the network ..."
           Et cetera.  And if we then go forward -- I'm not
       going to take you to them, but just looking at sample
       helpline logs, if I can deal with it generally -- I'm
       happy to show you them if you want but they are those
       Excel ones which take a while to download.
   A.  Yes.
   Q.  But in the NBSC call logs, in the helpline call logs in
       the week of 13 April 2014, which we've got in the bundle
       at {F/1201}, we've got, for example, 10,789 rather than
       1,089.37; in the 16 August 2015 {F/1368} we've got
       19,400, should have been 1,940, that's row 1171.  We've
       got row 1578, 14,000 instead of 1,400.
           At {F/1429}, 7 February 2016, we've got 3,400 keyed
       in as 34,000 and 4,428.58 entered as 44,298.58.
           Pausing there, I'm not going to show you but those
       are the sorts of numbers you recognise --
   A.  Yes.
   Q.  -- as the sorts of miskeying problems that people were
       encountering, is that fair?
   A.  Yes, that's fair.
   Q.  In your first witness statement you identified miskeys
       as a cause of shortfalls, which I think you fairly
       accepted earlier?
   A.  Yes.
   Q.  And can we look please now at {F/1449}.
           Now, this document is from 31 March 2016 and
       the heading is "Overkeyed online banking cash deposit
       customer present - end of procedure" and then it says:
           "Is the amount of the miskey greater than £150?" and
       there is "No" on the left and this is what to advise
       branch, it looks like the helpline:
           "Advice the branch that if the discrepancy is still
       apparent when balancing they must make the amount
       good ..."
           And then:
           "If deemed necessary a transaction correction will
       be sent at a later date to counteract against the
       initial discrepancy."
           Then "Yes":
           "Advise branch that if the discrepancy is still
       apparent when balancing they must make the amount good
       (loss or gain)."
           Yes?
   A.  Yes.
   Q.  So in some of the examples we have seen, if it is a very
       substantial sum, ie greater than £150, the right-hand
       branch of that would see a branch being advised that if
       the discrepancy is still apparent when they are
       balancing they must make it good.
   A.  Yes.
   Q.  And that could be a very large number and that's why we
       saw the previous -- the document we saw from 2014 was
       saying that a large discrepancy could put the viability
       of a branch in issue?
   A.  Yes.
   Q.  And the suggestion -- if you come down the middle of the
       flowchart, the first box in the middle says:
           "Open word document ie 'miskey A&L' or miskey
       non-A&L."
           That's miskey Alliance & Leicester or non-Alliance &
       Leicester, isn't it?
   A.  Yes, that's right.
   Q.  Because I think Alliance & Leicester inherited the
       Girobank?
   A.  Yes.
   Q.  "Complete the appropriate sections of the form by using
       the information included within the incident log."
           And:
           "To find the miskey form follow the file path
       P:\Core K Base\Miskey Forms\Miskey A&L or Miskey Non
       A&L."
           And then "Save a copy of the completed form to 'my
       documents' and paste a copy into the incident log."
           And so forth.
           So that fairly reflects what's actually happened
       with miskeys since the documents we have been looking
       at, does it?
   A.  That looks like what you have read out off the NBSC.
   Q.  Let's now move to your examples of changes that have
       been introduced, which are found at paragraph 156 and
       following {E2/5/35}.  These are other other changes.  If
       we look, for example, at the three you have chosen to
       identify in paragraph 158 -- so just for some context,
       you have obviously been in charge of branch support and
       latterly improvement?
   A.  Yes.
   Q.  So you are the person to go to for this evidence,
       aren't you?
   A.  For some of it, yes.
   Q.  And you have chosen these as -- are these good examples
       of the sorts of improvement that have been made?
   A.  They are some examples, yes.  These are the ones that
       were affecting the customers and postmasters in
       particular.
   Q.  Okay.  Well, let's take them in turn, if we may.  Let's
       start please with Drop & Go, that's at 158.1 and 158.2
       in your witness statement {E2/5/35}, isn't it?
   A.  Yes.
   Q.  So you point out in 158.1 that:
           "In November 2014 as a result of some branches
       settling Drop & Go sessions to cash instead of the
       customers pre-paid account, a screen prompt was added to
       the Horizon journey advising branches to not press
       settle at the end of the session before they had pressed
       end mails."
           Then two and a half years later:
           "... this was taken further as a new Drop & Go
       transaction was introduced which prevented users
       pressing the settle button before end mails ... the
       screen prompt was removed at the same time as it was now
       redundant."
   A.  Why he.
   Q.  That sounds all quite a happy experience, doesn't it, on
       the face of your evidence there?  It sounds as if it all
       went quite well?
   A.  It took a while and there were quite a few problems with
       Drop & Go actually.  It did take a while.  So there was
       a first fix but a recognition at the time that it needed
       to have a complete review of the transaction.
   Q.  Yes.  That extra bit doesn't leap off the page, does it?
   A.  No, it doesn't.
   Q.  The problems and the difficulties?
   A.  But it was -- yes, it has been which is why it was two
       stages in here, the first and then the second, but
       I take your point.
   Q.  Let's have a look, if we may please, at {F/1346}.  So
       here we're looking at 2 June 2015 and this is the ATOS
       presentation and if we go over the page to page 2 please
       {F/1346/2} it is referring to the online mails project,
       yes?
   A.  Yes.
   Q.  "... to introduce a new online channel for customers
       looking to send parcels via the Post Office by use of
       a service called Click & Drop.  The project was also
       scoped with the task of aligning this new service to the
       existing Drop & Go service and offering an online
       channel for those SME customers.  As the project
       progressed, Click & Drop was descoped
       during September 14 and the Drop & Go changes were
       deployed on 5 November 14."
           Yes?
   A.  Yes.
   Q.  That's the deployment of the changes you are referring
       to in your witness statement, isn't it?
   A.  Yes.
   Q.  And then it says:
           "The enhanced Drop & Go service introduced a new D&G
       database hosted on CDP ..."
           Do you know what CDP is?
   A.  Core digital platform.
   Q.  The core digital platform:
           "... and aligned the Horizon Drop & Go basket
       behaviour with the standard Horizon basket
       functionality."
           If we just go down to the bottom of that paragraph:
           "This reinstated the existing known issue of locking
       customer accounts if the correct counter processes are
       not followed."
           You see that?
   A.  Yes.
   Q.  "Since go live a number of issues have been encountered
       that have resulted in a large number of blocked customer
       accounts, a resultant high volume of calls into NBSC and
       dissatisfaction within branches and for impacted
       customers."
           If we look over the page please on {F/1346/3}, you
       can see the fifth bullet point there is:
           "Lessons learned that have been a captured to ensure
       that other similar projects don't suffer from the same
       pitfalls."
           Yes?
   A.  Yes.
   Q.  And if we go over the page {F/1346/4-5} we can see what
       was effectively happening.  There is a timeline there.
       And if we can just go please to the "Lessons learned" at
       page 12 {F/1346/12}, "Commercial decisions over customer
       experience" is the heading, "Summary":
           "The decision was taken to allow a known design flaw
       to continue to enable customers to use their D&G
       accounts to purchase other mails products."
           And then under "Limited and non-effective
       communications":
           "Limited training and communications were issued to
       branches as this was assumed to be an incremental change
       to an existing product."
           And the recommendation there is:
           "Branch staff should be engaged during user
       acceptance testing to provide a user's view on the scale
       of the change to inform the type of training and
       communications approaches that should be applied."
           These are the sort of things that you were referring
       to in your answer that added to the text of your witness
       statement, the sort of difficulties that were
       encountered?
   A.  Yes.  I wasn't aware of the exact details of what, but
       I knew there were problems.
   Q.  Okay.
   A.  I can't remember the date but at some point Drop & Go
       came to the branch user forum for feedback.
   Q.  Yes.  If you want to see that I can show you where that
       is.  That's at {F/1194}.  There's a branch user forum on
       20 March.  Is this the one you are thinking of,
       halfway --
   A.  No, I wasn't at this one.  So the product manager for
       Drop & Go, I invited her to come and take the input from
       the branch user forum whilst they were looking at the
       design of the improvements going forward, so she came
       a couple of times, but I wasn't at this particular one
       and I usually chaired this.
   Q.  Will you forgive me if we just sort of press on.
   A.  Yes, certainly, yes.
   Q.  What I was going to say is you will understand why
       I suggested to you that the difficulties with Drop & Go
       hadn't leapt off the page?
   A.  Yes.
   Q.  And it is fair to say that it gives -- what you had
       actually written included, gave a slightly rosy picture,
       is that fair?
   A.  Without going into all that detail, yes, that's fair.
   Q.  And if we look please now at {F/1549}, so we're now
       looking at the June 2017 changes and what was
       anticipated for those changes on 7 October 2016, let's
       just have a look at that please.  You will see that's
       the front of the document and if we look please at
       page 3 {F/1549/3}.  You will see there "Summary",
       "Issues and options", and the box at the top says "Money
       out of thin air upon timeout (introduced by prototype".
       Were you aware of this difficulty that was encountered
       at the time?
   A.  I've never heard of that terminology before.
   Q.  You haven't?
   A.  Not creates money out of thin air, I've never heard of
       that.
   MR JUSTICE FRASER:  I'm really sorry, I'm having great
       difficulty hearing what you are saying.
   A.  Sorry.  What I said is I've never heard of that
       terminology: creates money out of thin air.
   MR GREEN:  Had you heard of the problem by another name?
   A.  No, not that I can recall.  As I say, I knew there were
       problems, but not to that detail.
   Q.  And what type of document is this, if we just go back to
       the front?  {F/1549}.  Is this an operational document,
       does it look like, or ...?
   A.  No, there would normally be another facing sheet to
       a presentation, but if it was to existing prototype it
       typically would be part of the project, the project --
   Q.  Right, an IT document?
   A.  Yes.
   Q.  I'm just asking because if we look at page 3 again
       {F/1549/3}, there is a redacted column on the right.
   A.  Yes.  It's not something I have seen before so ...
   Q.  Okay.  If we go to page 5 please {F/1549/5}, just to
       sort of take this in stages, can you see on page 5 we've
       got the same heading at the top: A money out of thin air
       upon time out", do you see that?
   A.  Yes.
   Q.  And then number one in the table is "Do nothing", 2 is
       "Post settle script to detect", 3 is "Monitor available
       balance outside of basket", 4 is "Operate Horizon in
       Drop & Go mode disabling Fast Cash and allowing settle
       only against D&G account."
   A.  Yes.
   Q.  And the various risks and so forth are identified there.
       Then look at number 1:
           "Excess cash in till, compared to Horizon,
       incentivises fraud.  Remedy by FSC involves a debit
       'transaction correction'."
           Have you ever heard of a discussion of a debit
       transaction correction being used in quotes in that way?
       Is there anything ...?
   A.  No.  That doesn't make any sense to me.
   Q.  Okay.  Let's see -- if we go over the page to 6
       {F/1549/6}, number 3 says:
           "A new mode like 'back office' mode.  This mode
       would remove 'Fast Cash' and constrain 'settle' to just
       the D&G account; also to be used by time out."
           Are these things you understand or is this not
       something you can really speak to, this document?
   A.  No, this is something I wasn't involved in, so it's not
       something I can actually talk to.
   Q.  Just have a look if you would very kindly -- I will go
       more speedily on this then.  Page 10 please {F/1549/10}.
       There is a diagram of what would happen and I should say
       this appears to be the prototype that's being proposed:
           "Start mails looks up - but does not debit - the
       customer's account balance."
           And then on the right:
           "And credits that amount to the basket, effectively
       creating cash out of thin air."
           Do you see that?
   A.  Yes, I see that.
   Q.  So what we can see from this is that even reasonably
       well thought through proposed changes to Horizon could
       at least potentially have unintended effects.  We can
       see that, can't we?
   A.  Sorry, in the prototype you mean?
   Q.  Yes.
   A.  Okay.
   Q.  I mean it is obvious that this prototype has been
       developed as a proposed prototype which is being
       discussed fairly carefully and we can see from that that
       tweaking one aspect of one function might trigger
       something in another aspect of Horizon?
   A.  Yes.
   Q.  That's what we can see?
   A.  Yes.
   Q.  And if we go please to page 13 {F/1549/13}, this is
       "Proposed detection and correction" with Chesterfield
       effectively investigating and then at (e):
           "A matching credit to a 'write-off' account so that
       the cash that appeared out of nowhere returns to
       nowhere."
           That seems to be sort of getting the imaginary cash
       to vanish out again out of the account?
   A.  Yes.
   Q.  On the face of it, doesn't it?
   A.  It does.
   Q.  Then just look over the page, last point on this,
       page 14 {F/1549} if you would:
           "1.  The reason for the transaction correction
       is ..."
           This is the critique:
           "The reason for the transaction correction is opaque
       to the branch ..."
           For a reason that perhaps we can't see.  That would
       obviously be undesirable, wouldn't it, to have --
   A.  Yes.
   Q.  And then 3:
           "The initial credit to the basket offends against
       the conventions of double entry bookkeeping."
           Yes?
   A.  Yes.
   Q.  Because you are getting a figure pop up without
       a corresponding figure.  And then:
           "Routine use of write-offs is inappropriate and
       unacceptable to Ashley Hall's team."
           So that's the critique.  Having seen that, there
       were clearly not only some problems with Drop & Go that
       you were aware of and its development, but also some
       others, is that fair?
   A.  That would be fair.
   Q.  If we just look very quickly at {F/1640}.  That's
       an email chain and if we can look at page 3 where it
       begins {F/1640/3}, if you look at the bottom message,
       you will see:
           "Hi,
           "This is to get feedback from model office and for
       POL to confirm the release to the 54 pilot branches with
       the pilot to commence on Friday 28.04."
           Do you see that?
   A.  Yes.
   Q.  And then "Minor fix is required" in the next message up
       on the right-hand side, "Currently in progress for the
       pilot"?
   A.  Yes.
   Q.  And then 27 April at 11.26:
           "From discussions with Jeff Smyth due to the
       negative impacts of 15.92 on other projects the POL
       recommendation is that the Reference data for Pilot is
       NOT to be released on Friday."
           And then at the top:
           "After having had a conversation between yourselves
       on the below impact ... it now transpires that the code
       cannot be released ... as per the agreement yesterday.
           "Can you please have a discussion and let me know
       what has been decided and as a result the retrospective
       action be taken in the next 30 minutes.  Note the code
       has gone into MO and we are planning to move ahead to
       pilot tomorrow unless someone lets me know in the next
       30 minutes."
           And if we go to page 2 {F/1640/2} at the bottom we
       see:
           "My understanding is that the live code that is
       deployed in the MO has broken the existing live code for
       telecoms that is already live in the rest of the network
       estate.  If we were to deploy it (beyond MO) we will
       create a P1/P2 issue in the live network estate which is
       a crazy thing to do."
           Do you see that?
   A.  Yes.
   Q.  And then if you just look at the penultimate paragraph:
           "I've already checked with Rob H and Mick M on this
       one - we cannot consciously put code live if we know it
       will break the bigger network, especially on the run-up
       to a bank holiday."
   A.  Yes.
   Q.  Did you ever hear about the fact that there was a risk
       of a break to the bigger network in the run-up to a bank
       holiday, sort of 30 minutes away?
   A.  No.
   Q.  Did that -- no?
   A.  No.  I mean the whole point of testing the model is
       obviously to see what the impact of it would be and
       obviously that's what they said will be the impact if it
       were to go live.
   Q.  So it's fair to say, isn't it, that there's quite a lot
       that goes on in the background that maybe doesn't get
       quite up to your level, would that be fair?
   A.  This would be routine.  To go into the model office for
       testing, it would be routine, yes.
   Q.  But it was going beyond the model office to the pilots,
       that's the point.  That's what they're about to do,
       isn't it?
   A.  But once it got the go ahead from the model office --
   Q.  Yes.
   A.  And I haven't read all of it, but just reading what you
       have taken me to, what I read from this is that it found
       that it would break something else.
   Q.  So they stopped it just in time?
   A.  That's what I would anticipate from this, yes.
   Q.  Okay.  Let's look at cheque rem out, the cheque rem out
       problem, that's paragraph 158.3 of your witness
       statement at page 36 {E2/5/36}.  Now, errors in relation
       to the remming out of cheques go back pretty much to the
       beginning of Horizon.  They have been an issue over
       time, haven't they?
   A.  They had been an issue, yes.
   Q.  And you have pointed here to an improvement that's made.
   A.  Yes.
   Q.  And you say there:
           "This was a change to bring all the steps in the
       cheque rem out process onto a single screen and automate
       the calculation of the cheques total for the counter.
       This was prompted because branch staff occasionally
       forgot to complete the process fully due to
       interruptions, allowing extra cheque transactions to
       occur mid-process from other counter positions."
           Now if we look please at F/1129 and have a look at
       page 4 of that document please {F/1129/4}.  This is the
       Horizon improvements workshop and 1.5 is "Rem out for
       end of day cheque processing", "Summary":
           "Customer cheques accepted during the trading are
       processed at end of day.  The clerk performs the
       following process ..."
           You see remming them out and then over the page
       {F/1129/5} we have got:
           "This is a lengthy process and unnecessarily runs
       the report twice.  It also increases the potential for
       the clerk to enter an amount which does not match the
       report total amount."
           And then just below the diagram:
           "Having reconciled the results against the physical
       cheques they then navigate to back office, rems and
       transfers, out to cheques, cheques and enter the total
       amount to rem out and settle ... they then have to
       navigate again to back office, reports, cheques and
       preview or print the report.  Finally they have to cut
       off the report.
           "Suggested improvements.
           "An improvement could be to provide an additional
       button on the counter daily cheques listing screen to
       invoke the cheque rem out process and return when
       complete."
           So it is quite a good idea, yes?
   A.  Yes.
   Q.  So that fairly indicates what the problem was and what
       the proposed solution was going to be.  If we look at --
       so that's 5 September 2013.  If we go to {F/1225}
       please, 25 June 2014 and you will see if we go to
       page --
   MR JUSTICE FRASER:  This has got August 2013 in the middle
       and September 2013 in August writing at the bottom.  It
       doesn't seem to have June 2014.  I'm not saying it
       isn't June 2014, I just can't see June 2014.
   MR GREEN:  My Lord, in fact if I can just deal with the
       version control.  If we go -- we've got the wrong
       document up there.
   MR JUSTICE FRASER:  Do you want 1125 or 1225?
   MR GREEN:  We want 1225.
   MR JUSTICE FRASER:  That's probably the explanation.
   MR GREEN:  I probably called it out wrong.
   MR JUSTICE FRASER:  No, no -- it doesn't matter.  Is that
       the one?
   MR GREEN:  That's the one, 25 June 2014.  And this is branch
       support programme.  So this is June 2014; you were
       involved in the branch support programme at this time?
   A.  Yes.
   Q.  And we can see at the bottom KPIs, key performance
       indicators.
   A.  Yes.
   Q.  "The programme will track performance against the
       following key performance indicators ..."
           And then your name is alongside Gayle Peacock's at
       the bottom there.
   A.  Yes.
   Q.  If we go over the page please we see there is "Programme
       progress" and if we go forward a page to item 4.4 on
       page 3 {F/1225/3}:
           "Horizon system transaction improvements ... the end
       of day cheque remittance process has been reviewed and
       a new solution to drive out errors has been identified.
       This will be scheduled for the next software release
       in October 2014 will reduce the calls into NBSC and the
       transaction corrections issued by [the FSC]."
           Yes?
   A.  Yes.
   Q.  So that was what the intention was, was October 2014.
       Then if we go forward please to {F/1323}, this document
       is -- at the very bottom of it says 9 March 2015, do you
       see that?
   A.  Yes.
   Q.  And if we go to page 4 please {F/1323/4} we can see that
       the first draft of the document was actually
       3 February 2015, yes?
   A.  Yes -- sorry, yes.
   Q.  And if we go forward one further page {F/1323/5},
       "Associated documents", we have the change proposal for
       cheque rem out just there at item 7 in the top box, do
       you see that?
   A.  Yes.
   Q.  So it hadn't actually come in in 2014 October, had it?
   A.  No.
   Q.  And in fact your witness statement confirms it was
       brought in in August 2015?
   A.  Yes.
   Q.  And when it did come in it made things a lot easier for
       subpostmasters?
   A.  Yes.
   Q.  Was there any reason why that couldn't simply have been
       done within a couple of years when the problem had been
       noticed?
   A.  So when I picked it up in the branch support
       programme --
   Q.  I'm not criticising you, I'm just saying is there any
       reason why that change could not have been made in say
       2002?
   A.  Well, it depends what else was going on at the time, but
       I mean on the face of it --
   Q.  There isn't any real reason, is there?
   A.  On the face of it that was quite a straightforward ...
   Q.  Yes.
   A.  And it did take -- it took me longer to get it
       introduced than I would have liked anyway.  It didn't
       cost an awful lot of money to get it sorted either, so
       ideally I would have liked it --
   Q.  Ideally it would have been better to do it a long time
       ago?
   A.  Yes.
   Q.  If we can move forward now please to bureau de change at
       paragraph 158.4 in your witness statement which is back
       on page {E2/5/36}.  Do you see there:
           "Bureau de change automated 2nd receipt.  This was
       a change to Horizon in September 2014 to introduce
       a function defining the number of copies of a receipt
       for a transaction which will automatically be printed.
       This was prompted because some branches were being
       defrauded by the use of fraudulent debit cards for
       bureau transactions.  If the counter clerk did not print
       a second receipt for the branch to keep (showing
       4 digits from the card number ... (a fraudulent
       transaction would not be refunded by the card issuer."
   A.  Yes.
   Q.  So from a practical point of view what was going to be
       proposed was that there would be assistance to try and
       help avoid the problem of a second receipt not being
       printed?
   A.  Yes.
   Q.  Shall we just look at how that actually happened.  If we
       look at {F/1129} and you can see this is back in the
       Horizon improvements workshop in 2013.
   A.  Yes.
   Q.  And if we go to page 5 of that document please
       {F/1129/5}.  That is effectively -- if we just look at
       "Travel money debit card branch receipt" at the bottom,
       do you see that?
   A.  Yes.
   Q.  2.1:
           "When a clerk sells travel money to a customer, and
       the customer pays by a debit card, the security check
       responses performed by the transaction and card details
       are printed on the session receipt at settlement.
           "The clerk can request a duplicate session receipt,
       but if they do not and proceed with another customer,
       then the reprint is no longer available to the clerk."
           So a moment's inattention and they are at risk of
       underwriting any fraud which otherwise the bank would
       bear?
   A.  Yes.
   Q.  And that has always been like that?
   A.  Yes.
   Q.  For years?
   A.  Yes.
   Q.  And it says there -- it explains:
           "If the bank subsequently challenges the
       transaction, the postmaster needs to provide proof that
       the security checks were performed.  They cannot do this
       without a receipt and may therefore have to make good
       the transaction and be out-of-pocket."
           Then the suggested improvement is to print
       a duplicate branch receipt when the travel money is paid
       for by debit card:
           "In order to be consistent with other receipts this
       should probably indicate it is a branch copy."
           That's pretty straightforward, isn't it?
   A.  Yes.
   Q.  And it avoids that risk of a moment's inattention taking
       literally one key press beyond where they can get back
       to get what they need, yes?
   A.  Yes.
   Q.  And if we look please at {F/436} you can see there this
       is an old document, this is a 2008 document, and we can
       see on page 57 of that, if we may, {F/436/57} that the
       system automatically prints the customer receipt for the
       transaction, yes?
   A.  Yes.
   Q.  At the bottom.  Then if we go over the page {F/436/58},
       this is the point about subpostmasters being told:
           "You must print a duplicate  ...(Reading to the
       words)... information is mandatory ..."
   MR JUSTICE FRASER:  Where are you reading from?
   MR GREEN:  Just at the top, my Lord "Please note".
   MR JUSTICE FRASER:  Yes.
   MR GREEN:  So press the "feed" key on the printer, so you
       have to remember to print the "feed" key on the printer
       and then request a duplicate receipt as a branch copy,
       that's what they have to remember to do.
   A.  That's right.
   Q.  So that was the old system.  The short point is that was
       actually quite a simple thing to change again,
       wasn't it?
   A.  Yes it was.
   Q.  It was a good idea which was fairly simple to execute
       which you brought in?
   A.  Yes.
   Q.  And there wasn't really any big reason why that couldn't
       have been done ages ago?
   A.  My view was -- when I learned of the problems that
       people were having with it, is that -- if a receipt is
       required as part of the transaction my view was it
       should have always been automatically produced.
   Q.  Yes.
           Now can we move forward to one of your responses to
       what Mr Coyne has said in relation to MoneyGram please.
       If we look at paragraph 161 of your witness statement
       {E2/5/36}, you say:
           "Turning to the MoneyGram example referenced by
       Jason Coyne, in October 2015 MoneyGram and Post Office
       agreed jointly to introduce debit/credit card acceptance
       for payment ... this change was aimed at making
       MoneyGram at Post Office more attractive for
       customers ..."
           And then at 162:
           "Accepting debit/credit cards as a method of payment
       is familiar to branch staff, however, processing
       a refund for MoneyGram card payments is different so the
       instruction given to branches is to liaise with NBSC to
       ensure the branch is walked through the correct
       process."
   A.  Yes.
   Q.  "This was an added step check to prevent the branch
       making an error."
           So one way of reading that is to suggest that when
       this was brought in branches were told to liaise with
       NBSC.  That's what it naturally reads, is that right?
   A.  Yes.  It did have instructions of what to do, but the
       fallback was ring NBSC, because it didn't happen very
       often anyway.
   Q.  Okay, well, let's just look at that in a tiny bit more
       detail, if we may.  Let's look please at {F/1382}.  This
       is a 1 October 2015 document, "New and improved
       MoneyGram service", and it says in the highlighted
       passage:
           "The Horizon transaction screens have been designed
       to help you transact these services simply and
       compliantly.  Please refer to Horizon Help for
       operational instructions and further information."
           Yes?
   A.  Yes.
   Q.  And then over the page {F/1382/2} on the back it says:
           "Important note - the refund/cancellation process
       for MoneyGram has changed.
           "If a customer wants to cancel their transaction and
       obtain a refund on the same day, you must ... go through
       the MoneyGram refund button to cancel the transaction.
       You then need to use the back office reversals process
       to reverse out the transaction (this is the case for
       both debit card and cash transactions)."
   A.  Yes.
   Q.  So that's if it is on the same day.  And then:
           "All refunds/cancellations the following day or
       later ... are refunded to cash through the MoneyGram
       refund button (regardless of method of payment), the
       back office reversals process doesn't need to be used."
           So what this is saying is if it is on the same day
       you have to use the back office reversals process as
       well as the refund button?
   A.  Yes.
   Q.  And otherwise you don't, yes?
   A.  Yes.  It says cancel and then reverse, yes.
   Q.  Does it mean the same calendar day?
   A.  Yes.  When you say -- well, it means on the day of its
       being -- yes.
   Q.  The day, so if it's Tuesday, it means on Tuesday?
   A.  On the same day, yes.
   Q.  What about if it is after polling time?  Do you know
       what polling time is?
   A.  From -- so the instruction -- my understanding was it
       was if it was on the same day and if it was after -- and
       if it was the next day you would need to treat it
       differently.  That was my understanding.
   Q.  Yes.  But taking it in stages, it just says "the day"
       there, doesn't it?
   A.  Yes, it does, it just says the day.
   Q.  And it doesn't say you must speak to the NBSC, does it,
       when you do a reversal?
   A.  Not on this one, no, it doesn't.
   Q.  No.  So that's the October 2015 one and then if we look
       please at paragraph 163 of your witness statement
       {E2/5/37} where you refer to further communications to
       remind branches, yes?
   A.  Yes.
   Q.  And you are referring there to the first one is
       3 March 2016 Branch Focus article.  Can we look at
       {F/1440} please.  Now, we haven't got this in the
       context of the Branch Focus magazine, we can't see where
       it sits amongst some of the other content, but it could
       be quite varied, couldn't it, in the Branch Focus?
       There could be lots of --
   A.  I see, yes, there could be quite a number of pages in
       there.
   Q.  We had "summer sizzlers" and other things last time.
   A.  Yes.
   Q.  But we've got the text that is said to be what was said
       and we can see here, "MoneyGram refunds".  If you come
       down to just below "To manage customer expectations",
       can you see:
           "To process a MoneyGram refund to a debit card
       please contact NBSC ... before completing the
       transaction."
           Yes?
   A.  Yes.
   Q.  So that's on 3 March 2016 and you can see there that
       this one now does say contact the NBSC, where
       the October 2015 one didn't say that?
   A.  Yes, that's right.
   Q.  And Mr Patny had his problem with MoneyGram on
       23 February, about ten days before this new guidance
       came out, yes?
   A.  Yes.
   Q.  So for the purposes of Mr Patny could we fairly rewrite
       what you say at paragraph 162 on page 36 of your witness
       statement, {E2/5/36}, would it be fair to say that the
       instruction given to branches from 3 March 2016 was to
       liaise with NBSC?
   A.  If that's the date of that one, yes.
   Q.  Yes.  Then the next thing you refer to at 163.2
       {E2/5/37} is the flowchart at {F/1767}.  Can we have
       a look at that please.  So we are now in
       19 February 2018.  Again, we don't know exactly where or
       how this was in the Branch Focus article?
   A.  My understanding is it was attached to that version of
       it.  That's my understanding, it went out with that
       version of ...
   Q.  Well, this one seems to be 19 February 2018 this one.
       Is that right?  You said it is 19 February 2018 in your
       witness statement?
   A.  Yes, because the actual diagram itself isn't dated, the
       flowchart itself isn't dated on here.
   Q.  Yes.
   A.  So when I located it it was with that Branch Focus is
       how it was presented to me as that's when it went out,
       on that date.
   Q.  On which date, 19 February?
   A.  Whatever I have said, date -- yes.
   Q.  I'm just pointing out we haven't got the actual Branch
       Focus --
   A.  No.
   Q.  -- we have just got the flowchart that you say went out
       with it and it says "MoneyGram decision tree - refund
       customer diagram V0.2".  Did this definitely go out?
   A.  Yes.
   Q.  Okay.  And can you come down from "Customer wants
       a refund", if you come down to "Cancellation approved?",
       if you go straight down the "Yes" arrow, "Same day
       refund?" and the arrow then says "Before 7 pm cut off?"
   A.  Yes, I see that now, yes.
   Q.  Now, that is the polling time at which the branch's
       business day ends, isn't it, and the branch's business
       day transactions are then sent off?
   A.  Yes, I mean transactions continue after, but yes.
   Q.  They do, but that's the polling time at the end of the
       business day for the branch?
   A.  Yes.
   Q.  You will probably remember sitting in court when
       Mr Patny was saying he was on the phone to the helpline
       to try and do his reversal at just after 7 o'clock?
   A.  Yes, I remember him saying that.
   Q.  He said a couple of times that was past polling time.
   A.  Yes, I remember him saying polling time, yes.
   Q.  Let's have a look, if we may please, at {F/1664}.  Now,
       this is an operations board agenda for July 2017.  A lot
       of it is blank, but if we could go please -- well, now
       blank.  If we go please to page 86 {F/1664/86}, it was
       at least at this stage finally recognised I think
       that -- if we go down:
           "The main root causes for the increase in items
       were:
           "Branches retrying transactions that had failed due
       to timing out."
   MR JUSTICE FRASER:  Where are you reading?
   MR GREEN:  Under "The main root causes", the second bold
       heading, my Lord.
           Do you see that, Mrs van den Bogerd?
   A.  Yes.
   Q.  "The main root causes for the increase in items were:
           "Branches retrying transactions that had failed due
       to timing out."
           Because that had become a bit of a problem,
       hadn't it?
   A.  Yes.
   Q.  Timeouts in relation to MoneyGram transactions.
           "Branches reversing a transaction but not cancelling
       the AP part of the transaction."
   A.  Yes.
   Q.  Is that the authorisation process?
   A.  Yes.
   Q.  "General lack of understanding in branch of how the
       process works, especially for refunds."
           And then:
           "MoneyGram and POL systems were out of sync re
       cut-off times.  Poll cut off at 19.00 and MoneyGram at
       24.00."
           Then in bold:
           "Towards the end of 2016 changes were made to the
       transaction journey and the cut-off time.
           "MoneyGram changed their cut off time ..."
           Et cetera and also:
           "The end to end transaction journey time was
       extended from approximately 8 seconds to 30 seconds."
           That was in order to try and stop the timing out
       happening?
   A.  Timing out, yes.
   Q.  And then there are some next steps listed:
           "More work being done to establish the root cause of
       error (branch conformance/product complexity)."
           So it is right, isn't it, that in fact the issue of
       MoneyGram was a little bit more complicated, because
       initially it didn't say "Contact the NBSC"?
   A.  Yes it changed, yes.
   Q.  And then that then changed.  And then there was
       a flowchart which for the first time mentioned the 7 pm
       rather than the calendar day, yes?
   A.  Yes.
   Q.  And distinguished between pre-7 pm and post and then
       eventually at the end of 2016, as reflected in the 2017
       document, the polling days for Post Office and MoneyGram
       were then aligned?
   A.  Yes.  The flowchart was the best, I think, communication
       out to the network around the stages to go through.
   Q.  Okay.  Let's just look, if we may, at {F/1556} please
       and if you could look very kindly -- you see that's
       a 21 October 2016 document.  Have you seen that before?
   A.  I don't recall.  I don't recall seeing this actual
       document, but I was party to some of the operations
       duplication work.
   Q.  Let's look please at page 4 {F/1556/4}.
   A.  Sorry, what date did you say this document was?
   Q.  This document seems to be 21 October 2016.  Does that
       make sense?
   A.  Okay, yes.
   Q.  If we have a look at page 16 of it {F/1556/16} we've got
       the document history and you will see the version
       history at the bottom and it says initial draft,
       30 September 2016, Andy Greening.  Who is he?
   A.  He is one of the change analysts.
   Q.  One of the change analysts.  Do you know him?
   A.  Yes.
   Q.  And you have worked with him?
   A.  Yes.
   Q.  Review by Dawn Brooks.  Who is she?
   A.  Dawn used to be a change analyst, one of the managers,
       in FSC.
   Q.  And is Andrea Horner doing the same job as well?
   A.  Andrea Horner was a project manager with us, she was
       a contractor with us for a short time.
   Q.  Okay.  So it seems to be anyway a document, the date of
       which is in October 2016, and if we just look please at
       page 4 {F/1556/4} there is a management summary:
           "Post Office branch 'back office' balancing and
       accounting processes are still rooted in legacy
       practices and methods.  These operations processes and
       associated workflows are inefficient and susceptible to
       losses and fraudulent activity.  Many business rules are
       ill-defined and the language is confusing to those who
       operate the processes."
           Now, some of those who are operating processes are
       in FSC dealing with supposed shortfalls from
       Post Office's point of view, yes?
   A.  They would be, yes.
   Q.  People deciding on transaction corrections?
   A.  Yes.
   Q.  And also of course one shouldn't forget the
       subpostmistresses and subpostmasters who have got the
       other end of these processes in a sense?
   A.  Yes.
   Q.  And it is the reconciliation of the back office data and
       the front office entered data that causes a discrepancy,
       isn't it?
   A.  Yes.
   Q.  And if we look at "Drivers for change", 1.2, it says:
           "There is a requirement for balancing and
       reconciliation processes to be intuitive and easy to
       operate and for those processes to use simple language
       and applied business rules that negate the requirement
       for excessive training in branches and back office
       support services."
           So what really you are saying there is this should
       be absolutely intuitive on the Horizon system and in the
       processes that relate to it?
   A.  Yes.
   Q.  So that you don't have to go and explain super
       complicated things to people in branch and then have
       mistakes which then have to be sorted out.  Is that
       fair?
   A.  Absolutely.
   Q.  And that's obviously a good idea?
   A.  Of course.
   Q.  And there's a requirement for integration of data into
       Post Office Limited's front office systems, which is
       basically the Horizon facing system for the SPMs?
   A.  Yes.
   Q.  And that would obviously be a good idea as well, in that
       if the front office data and the back office data was
       visible in the front office, they would be able to see
       both?
   A.  Yes.
   Q.  These are all good ideas and it's not surprising that
       you think they are good ideas because you are the person
       who signs off on this document at the end on page 17
       {F/1556/17}.
   A.  Yes.
   Q.  Did it begin to seem familiar when we were going through
       it?
   A.  So -- yes, absolutely.  So ATM, so ... I don't
       absolutely recall this particular document if I'm
       honest, but --
   Q.  You remember the broad points, trying to streamline the
       back-end office processes?
   A.  Absolutely.  So at this time -- and I can't be sure of
       the dates, but I was the director of support services,
       so FSC was under my remit at the time and what
       I particularly wanted to do was try to get things right
       at the front end to stop driving -- obviously right for
       the front end in the first place but then to stop
       driving, as you have described, the workarounds and the
       activity to put right things that should have been right
       in the first place.
   Q.  I understand.
   MR JUSTICE FRASER:  Mr Green, you have to keep an eye on the
       clock for the shorthand writers.
   MR GREEN:  My Lord, would that be a convenient moment?
   MR JUSTICE FRASER:  I imagine you have more than five
       minutes.
   MR GREEN:  My Lord, I have.  I'm not going to finish with
       Mrs van den Bogerd today but I will be finished well
       before lunch tomorrow and we are timetabling everyone
       else to fit in --
   MR JUSTICE FRASER:  This is day one of your four days.
   MR GREEN:  Indeed.
   MR JUSTICE FRASER:  Mrs van den Bogerd, you heard that
       exchange so apparently you are still going to be in the
       witness box tomorrow but we will have a break now for
       the shorthand writers so we will come back at 20 past 3.
       Same as before, don't talk to anyone about the case.
   (3.14 pm)
                          (Short Break)
   (3.22 pm)
   MR GREEN:  Mrs van den Bogerd, just following through
       a couple of further documents in relation to MoneyGram.
       Can you please look very kindly at {F/1502} please.  You
       will see that's a MoneyGram quarterly business review,
       18 July 2016.
   A.  Yes.
   Q.  And if you could look please at page 24 {F/1502/24}.
       "Problem statement":
           "Multiple instances of system latency have been
       reported since 24 January 2016."
           Yes?
   A.  Yes.
   Q.  "The latency was exposed by changes to the IT
       environment in the January release."
           Yes?
   A.  Yes.
   Q.  So it has come to the fore after the January release of
       the software, yes?
   A.  Yes.
   Q.  And it identifies the purpose and benefits and then
       fourth bullet point at the bottom of that page:
           "This latency is causing some transactions on the
       POL system to time out, resulting in high traffic to the
       operational contact centres and obvious customer
       impact."
           Yes?
   A.  Yes.
   Q.  If we go over the page please {F/1502/25} we can
       actually see the "Instances of latency since February"
       diagram shows weeks commencing 22February onwards, can
       you see that?
   A.  Yes.
   Q.  And 22 February 2016 is the highest one by quite
       a margin, isn't it?
   A.  Yes.
   Q.  And 23 February is when Mr Patny has his difficulties,
       isn't it, with MoneyGram?
   A.  Yes.
   Q.  If we go forward please to {F/1502/29}, do you remember
       Mr Patny thought that he had got a duplicate
       transaction?
   A.  Yes.
   Q.  3,100 and then he seemed to have 6,200 --
   A.  Yes.
   Q.  -- was what he thought had happened.  Look at the bottom
       of that page, "Reconciliation":
           "Duplicate transactions are created in [MoneyGram]
       systems as a result of Post Office time outs.
           "A joint MG/POL team is working to ascertain impact
       on settlement and implementation of an appropriate plan
       of action."
   A.  Yes.
   Q.  It wasn't wildly well advertised to subpostmasters that
       this problem was being dealt with at the time?
   A.  No.  People were aware there was an issue, clearly, in
       branch and then --
   Q.  People in the branch had an issue, but they were
       probably being blamed for user error, weren't they?
   A.  Well, they knew there was an issue and they were ringing
       into NBSC at the time.
   Q.  Sorry?
   A.  They knew there was an issue and they were ringing into
       NBSC at the time.
   Q.  Well they were claiming there was an issue but they
       might be met with "This is user error"?  They didn't go
       "Oh, by the way, have a look at the graph which shows
       there has been a spike in the very week I'm complaining
       about" because they didn't have that information,
       did they?
   A.  Not unless there was a message on the helpline system.
       I'm not aware --
   MR JUSTICE FRASER:  When did you first see this document?
   A.  This document, now.
   MR JUSTICE FRASER:  You haven't seen this before today?
   A.  I've not seen this before, no.
   MR GREEN:  Okay.  Let's look, if we may, at {F/1555}.  This
       is a "Small project proposal", "Hot house - MoneyGram
       issue", "Pricing for additional work".  And then
       "Description of requirement" box halfway down:
           "Background:
           "MoneyGram summary.
           "For the last several months Post Office has
       experienced a live operational issue with MoneyGram
       transactions across the branch network.  In the event of
       a transaction timing out at the counter, a system error
       message is displayed to the user ... and the transaction
       is aborted.  This leaves no record of the transaction at
       the counter and the transactions and funds may or may
       not have been committed in the MoneyGram domain.  This
       causes significant issues for Post Office and MoneyGram
       and for customers."
           And then:
           "End to end testing has been commissioned to
       identify the cause of the problem.  Within the first
       hour of testing the ATOS test team were able to identify
       the issue as CDP returning an incorrectly formatted
       (SOAP error message) response to the counter which
       results in a system error rather than a time-out
       response ..."
           And if we just look at the bottom three lines of
       that box:
           "A quick fix has been identified by the ATOS team to
       amend the existing AP-ADC reference data to treat the
       system error as a time-out response and trigger the
       correct counter behaviour."
           Do you see that?
   A.  Yes.
   Q.  So that does suggest, doesn't it, that there was a real
       problem that was not very easy for the person at the
       counter to deal with correctly?
   A.  Yes.  They had an error message, but yes.
   Q.  This document is dated 19 October 2016, as we can see,
       top right?
   A.  Yes, I can see that.
   Q.  Yes.  And it is addressing something that was a result
       of the January release of software?
   A.  Yes.
   Q.  So that had actually persisted for the whole of
       February, March, April, June, July, August, September,
       October, nearly the whole of October as well, yes?
   A.  From the graph you showed, yes.
   Q.  If Post Office had commissioned ATOS earlier -- if they
       had been asked to look at it earlier, it looks as if
       they would have got an answer more swiftly, because they
       found it in the first hour of testing, didn't they?
   A.  Yes.
   Q.  Just looking at that from Post Office's position, does
       that look satisfactory to you?
   A.  I don't understand why it took so long.
   Q.  No.
   A.  On the face of it, no.
   Q.  Let's look at other matters that you deal with in your
       witness statement, if we may.  Paragraphs 178 and 179
       {E2/5/41}, you are talking there about volume of
       branches and volume of customer sessions.  You can see
       that the point you make at 179 over the page {E2/5/42}
       is you give the example of the Barkham branch as
       an example, being the branch that Mrs Stubbs was
       subpostmistress for, raises a number of matters that you
       have responded to.  You were able to see that the
       workload for this particular branch in 2001 involved an
       average of 1,047 weekly customer sessions and in 2007
       this increased to 1,836, yes?  And that's information
       that you got from the spreadsheet that we see on the
       previous page referred to there, is it?
   A.  Yes.
   Q.  Can we look please at {H/172}, which is a letter from
       Post Office's solicitors dated 18 January.  You will see
       there the second line of 1.1 -- do you see in the
       middle?
   A.  Yes.
   Q.  They are talking about your witness statement:
           "It has come to our attention that this spreadsheet
       contains an error due to the internal referencing that
       means, while the actual volumes of customer sessions is
       correct, they do not correspond correctly with the
       relevant branches.  We therefore write to enclose a copy
       of the correct version of the volume of customer
       sessions spreadsheet for your reference."
           Yes?
   A.  Yes.
   Q.  Were you made aware of that change?  Do you know that
       that change happened?
   A.  No.
   Q.  Okay.  This may not be fair to you, but let's just look
       at it.  You have referred in your witness statement, we
       have seen, to the Barkham branch as an example.
   A.  Yes.
   Q.  I have just shown you that the point you make about
       customer sessions going up from 1,047 to 1,836.  Let's
       look please at the corrected spreadsheet which is at
       {F/1837}.
           Is this something that you actually know about, or
       is this something someone asked you to put in your
       statement?
   A.  What, the volume of branches in the --
   Q.  Yes?  I mean is this really your evidence at all?
   A.  I know about branches and how they have changed over the
       years, yes.
   Q.  Okay, so you felt comfortable giving the evidence?
   A.  Well, I spoke to Kjetil, a colleague of mine.
   Q.  Spoke to ..?
   A.  Kjetil.
   Q.  Full name?
   A.  Kjetil Fuglestad.
   Q.  That's the person you mention in your witness statement
       {E2/5/41}.
   A.  So Kjetil manages all the data on the network, the size
       of -- the number of branches, volume of transactions,
       everything that goes through that; that's what he does.
   Q.  So he would be a reliable person to get this information
       from?
   A.  Yes.
   Q.  If we look at line 254 please if we look there we can
       see the Barkham branch and we can see there that the
       figures between 2001 and 2007 were in fact 398 and 368;
       do you see that?
   A.  Yes.
   Q.  And they weren't the figures that you gave in your
       witness statement of 1,047 and an increase up to 1,836.
       So in fact they have gone down those figures, haven't
       they?
   A.  Mm-hm.  Yes.
   Q.  And they are quite a lot smaller?
   A.  Yes.
   Q.  Do you know how that error has come about?
   A.  I don't.  Because I had a conversation with Kjetil
       around the data and then told him what I wanted and he
       provided it to me, so I don't know --
   Q.  Could you speak up a tiny bit.
   A.  Sorry.  I had a conversation with Kjetil around the
       information that I was looking for and then he provided
       it to me, so I have no idea how this data would have
       changed, or been incorrect in the first place.
   Q.  Okay.  Well, let's move on to your description of the
       implementation of Horizon and Horizon Online.  Let's
       look at paragraph 180 of your statement please
       {E2/5/42}.  You say there:
           "In relation to the practical implementation of
       Horizon and Horizon Online in branches, Post Office
       representatives were present as explained below."
           Then you say at 181:
           "It is my understanding that branches were required
       to do their final paper based cash account before the
       change over to Horizon.  The information was then
       transferred onto the Horizon system ..."
           Using the physical cash and stock held, et cetera,
       as could be agreed with the subpostmaster:
           "The transfer onto Horizon was supported by the
       Horizon field support officer ... and also the RNM.  If
       there were any issues as to discrepancies/general
       functionality ... [they] would assist the branch."
           Then Horizon Online, 183:
           "Before the change to Horizon Online a cash check
       was completed in branch by the HFSO."
           That's the Horizon field support officer?
   A.  Yes.
   Q.  You have defined in 182:
           "Branches were notified in advance that this cash
       check would be carried out.  I recall that this
       mandatory cash check across the entire network caused
       a temporary spike in declared losses.  I suspect that
       this was due to branches tidying up their accounts
       before the cash check and therefore losses coming to the
       surface that had previously been ignored or covered up."
           Now, that doesn't on the face of it suggest any
       difficulties particularly encountered by SPMs on the
       introduction of either Horizon or Horizon Online,
       does it, your account?
   A.  No, it just states what the process was, yes.
   Q.  Yes, it is process based, but it doesn't hint at any
       difficulties they encountered that might explain a spike
       in discrepancies, for example when Horizon Online was
       introduced, anything like that?
   A.  So that was done -- so what I'm referring to there is
       the spike before it went onto online.
   Q.  Okay.
   A.  So it was at the point at which it was done in advance
       of and then carried over.
   Q.  Well, let's have a look.  Let's just take a step back.
       We might be able to take it more shortly.  I mean do you
       actually personally know what the introduction of
       Horizon was actually like for the SPMs affected by it?
   A.  Going from paper based to Horizon, is that what you ..?
   Q.  Yes.
   A.  Yes I do.
   Q.  And there were quite a lot of problems, weren't there?
   A.  There were some.  I was supporting -- when I refer to
       the RNM, I was one of -- I was an RNM at the time, yes.
   Q.  And there were quite a lot of problems with the software
       and hardware?
   A.  At that time I can only talk about what my experience
       was.
   Q.  And what was your experience?
   A.  So we had some issues in -- where I worked in the local
       area, but we were out in branch supporting and making
       sure that everything was okay.  So there would be the
       odd -- from recollection, because it was a long time
       ago, there were the odd one or two issues that we had.
       I wasn't aware of the picture across the business at the
       time.
   Q.  Okay.  Let's just have a quick look, let's look at
       {F/16/2} please.  This is PEAK 0027887.  And it is
       created on 21 July 1999 and this is the one that we have
       seen at the end of last week, on Thursday last week.
       There's a receipts and payments misbalance which is
       escalating and it gets up to £1.05 million.  Do you see
       that?
   A.  Yes.
   Q.  And that's quite a large number for a small branch to
       have gone astray, isn't it?
   A.  Well, for any branch.  Any branch that would be --
   Q.  That's not going to be a real amount of cash --
   A.  No.
   Q.  -- that the SPM has put in her pocket, is it?
   A.  No.
   Q.  Or quite a lot of stamps or anything like that?
   A.  No.
   Q.  And look at page 3 please {F/16/3}.  You can see in the
       bottom box:
           "The initial balance brought forward for this CAP
       was [1.196 million].  This was multiplied twice to give
       a total ... of 2,279,189.04."
           There was a £1 million discrepancy:
           "This was due a known software error which has no
       been resolved."
           Is that "not been resolved" or "now been
       resolved" --
   A.  I don't know.
   Q.  -- can you remember this one?
   A.  No.
   Q.  Okay.  So that's quite a serious one.  Let's have a look
       at a couple more examples.  {F/22} please.  This is
       10 November 1999.  PM Dungannon.  This is PEAK 0033128
       and you can see there that there's a discrepancy of
       £43,000.  Halfway down the first box:
           "Outlet has a discrepancy of £43,000 after balancing
       [stock units] and doing office snapshot."
           Do you see that?
   A.  Yes.
   Q.  "... investigating why this misbalance occurred."
           If we look at the entry for 3 December 1999, which
       is down towards the bottom:
           "I have talked with development ref this problem.
       It is seen as a one-off.  No fault can be found and
       developments do not expect to be able to find a fault
       with the evidence available.  There is no additional
       information available as evidence.  I suggest this call
       be placed on monitor for 1 month."
           And then 18 February 2000, if we go over the page on
       page 2 {F/22/2}, if you look there can you see "Further
       examination of the event logs", do you see that?
   A.  Yes.
   MR JUSTICE FRASER:  Whereabouts?
   MR GREEN:  It is the big box, 18 February, 17.07, my Lord,
       and we come down --
   MR JUSTICE FRASER:  Yes, got it.
   MR GREEN:  "Further examination of the event logs for these
       two counters indicate that counter 5 looks suspect
       (C drive nearly full with big gap of no messages).
       Calls from PO into HSH for period ... indicate a reboot
       (counter not specified but would tie in with counter 5
       event log) ..."
           Do you see that?
   A.  Yes.
   Q.  And then straight after that, can you see underneath
       Saturday 31 October 1999 it says:
           "The evidence in the message store was that messages
       continued to be written to the message store but that
       all the 'payment' transactions which should have been
       recorded in the rollover trailer messages failed to
       appear (although others did, such as rem out and
       transfer out totals).  This indicates that the problem
       was not one of running out of disc space but of failing
       either to retrieve, or write out, transaction totals for
       one particular node in the node hierarchy.
           "Given that there were known problems with corrupted
       persistent object indexes at about this time, it is
       possible that an update on an EPOSS nodes object failed
       to be registered correctly at the outlet, causing the
       node accumulation to fail."
           Now, on the face of it there was a £43,000
       discrepancy in November 1999 and the underlying
       information wasn't actually rechecked until February the
       following year, on the face of this, yes?
   A.  That's what it looks like, yes.
   Q.  Let's go forward please to page 3 {F/22/3} and we're
       going to look at the entry for 7 April which you can
       see -- sorry, if you look at the top can you see there
       there's a further occurrence for 9,000, do you see that?
   A.  In Appleby?
   Q.  Yes.  And then that's then escalated and then at the
       bottom of that box a fix, penultimate paragraph:
           "A fix has been developed and is currently in
       testing."
           Do you see that?
   A.  Yes.
   Q.  And if we go down to 4 July 2000 at the bottom, it says:
           "Root cause of stock unit integration problem.
           "Data trees have been failing to build fully, and
       the system has not been detecting this.  Consequently,
       discrepancies in the balancing have been occurring.  In
       the case of Dungannon a whole payments node was missing.
       There have been a number of calls relating to this kind
       of issue.  A fix has been put in at CI4 which will
       prevent this happening.  The root cause identified ...
       is as follows: 'data server trees have failed to
       build ..."
           So what we can see there is the sort of problem that
       was being encountered is not only an error occurring but
       the system failing to spot that the error was occurring,
       yes?
   A.  Yes.
   Q.  And if we go forward please to {F/89} we will see the
       ICL Pathway customer service management support unit
       monthly incident review for November 2000 and if we look
       at page 6 quickly {F/89/6} -- sorry, I shouldn't -- fair
       to you.  If we look at page 6, do you see the second
       bullet point:
           "The most frequently occurring incidents in November
       were both types of receipts and payments incidents
       (migration and post migration) with 31 incidents per
       category.  The migration incidents have remained at the
       same level ... post migration occurrences have
       increased.  This was followed by 17 transactions polled
       by TIP but not by HAPS, these were due to delayed
       transactions as reported ..."
           And so forth.  Then they were added back into normal
       processing.
           We can see there that at the top bullet point, key
       point:
           "During November the number of incidents received by
       MSU increased to 109, in comparison to October where 91
       received and resolved ... a total of 98 incidents were
       cleared and 13 will be carried forward ..."
           So there was quite a lot of activity in terms of
       trying to clear incidents that were coming in of this
       type, weren't there?
   A.  Yes.
   Q.  And we can see in December, which is at {F/84}, that's
       the December document of the same type, let's have
       a look again at page 6 quickly {F/84/6} --
   MR DE GARR ROBINSON:  My Lord, I rise simply to say these
       are Fujitsu documents, these are not Post Office
       documents, but I'm sure your Lordship is well aware of
       the limitations in this witness' ability to speak to
       these documents.
   MR JUSTICE FRASER:  Well, she has given broad evidence of
       these types of things.  You might want to just explore
       when she has first seen them.
   MR GREEN:  Yes.  Have you seen these documents before?
   A.  I haven't seen these.
   Q.  You haven't?
   A.  No.
   Q.  But you were in aware in your support role at the time
       of activity of this sort going on?
   A.  Not in 2000 I wasn't, no.
   Q.  You weren't?
   A.  No.  So in 2000 as the retail network manager I operated
       in an area as part of a team of the same, so I was not
       aware of this in the background.
   Q.  Were they giving you any feedback about the sort of
       problems that Fujitsu were encountering?
   A.  No, no.  So the process we had in place was we had
       a roll-out plan of the branches that would migrate.  We
       organised ourselves in what we called a cluster group at
       the time and we made sure that we had one of us as
       an RNM was on-site to support the migration and then
       subsequently some of the balances going forward as well,
       because there were some issues with the balances, but
       I wasn't aware of any of this in the background.
   Q.  So you are the person listening to the SPMs saying
       "Look, I have a problem with it".
   A.  Yes.
   Q.  But you weren't being provided with any background
       information --
   A.  No.
   Q.  -- about what the categories of problems being
       encountered were?
   A.  What I would do in my role then is if I had a -- if one
       of my postmasters had a problem then I would go and
       check their accounts with them, go through all the
       information and there was one issue that I do remember
       where there was an issue with a postmaster that always
       balanced and on this particular day he didn't.  It was
       very unusual.  I checked everything and then I made
       a call to Chesterfield, which wasn't called FSC then, it
       was called P&BA and they resolved the problem from
       there.  I don't know exactly.  All I know is he was
       happy it was resolved.
   Q.  Let's look at that specific example.  They didn't give
       you feedback about what had gone wrong, if anything?
   A.  No.
   Q.  So you weren't able to form a view about that?
   A.  No.  So my understanding was what to do in the
       situation, if we had a situation like that, then I would
       escalate it and then the process would kick in behind
       because I was on to the next Post Office to support.
   Q.  Okay.  So is it fair to say --
   MR JUSTICE FRASER:  Just pausing there, this might speed
       things up because there are two different strands you
       are running at the same time.
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  The first thing is just a point of
       clarity.  I don't think these are actually Fujitsu
       because I think at the time it was ICL; that's right,
       isn't it?
   A.  That would have been at the time, yes.
   MR JUSTICE FRASER:  You were a retail network manager.
   A.  Yes.
   MR JUSTICE FRASER:  So the problems that you experienced,
       that you had direct experience of, they were the
       problems from the branches that were under your
       umbrella, is that right?
   A.  Yes, that's correct.
   MR JUSTICE FRASER:  Did you get any communications back from
       other areas about problems they were having, or was your
       involvement solely in respect of the branches in your
       area?
   A.  So my responsibility was the branch -- it wasn't just my
       area, so I think at the time there were about eight RNMs
       worked out as a team so we covered a geographical area
       that would cover the whole of say Cardiff, Swansea,
       West Wales, which is quite a geographical error, and
       then we would talk to our counterparts in Newport and so
       forth, but I'm not aware of any formal communication
       back to us at the time of anything to do with this at
       all, other than we were there to help postmasters
       resolve the issues at the time.
   MR JUSTICE FRASER:  Understood.  And then the second point
       is when you were preparing your witness statement and in
       particular the paragraphs at 180 to 183 --
   A.  Yes.
   MR JUSTICE FRASER:  -- did you do any investigation in
       respect of what might have been happening that you
       didn't know at the time in 2000 --
   A.  Not back to 2000, no.
   MR JUSTICE FRASER:  -- or in 2010 when the change was from
       Legacy Horizon to Horizon Online?
   A.  So in 2010 I was in a different role and had broader
       responsibility and I knew what -- what we did, again we
       replicated a similar approach to make sure we supported
       branches at the time, but as for any detail of
       information, I didn't research into that, no.
   MR JUSTICE FRASER:  Right, Mr Green.
   MR GREEN:  I'm most grateful.
           So just quickly to follow through on your role at
       the time, if you just look at page {F/99.1}.  Now, this
       is a document about Mr Bates' branch.  It's not one we
       went to in the common issues trial.
   A.  Yes.
   Q.  But this is June 2001 and I think you were still an RNM,
       is that right?
   A.  Yes.
   Q.  And this is an audit of his office and can we go down
       please to page 4 {F/99.1/4}.  Do you see cash management
       accurately declared at the close of business; you see
       that?
   A.  Yes.
   Q.  And this is the sort of document you might have seen if
       an audit had been done at one of your branches?
   A.  Yes.
   Q.  And if you look under "Control gaps - high risk" and
       then the next one says "Comments", it says -- and this
       is from the auditor:
           "A correct assessment of cash holdings could not be
       made because the Horizon system intermittently adds the
       previous day's cash holdings to the daily declaration."
   A.  Right.
   Q.  That's the sort of problem that you might have had to
       deal with if that had happened?
   A.  If it had happened, then yes --
   Q.  In one of your branches --
   MR JUSTICE FRASER:  Mr Green, can you let the witness
       finish.
   A.  That wasn't anything I was aware of at the time.
       I never encountered that problem in one of my branches,
       or in my colleagues' branches; that isn't something
       I was aware of.
   MR GREEN:  Very well.  What you would have done was taken
       that forward and tried to help the SPM resolve it?
   A.  Yes.  I mean if I had an audit report from any of my --
       from my branches then I would go through it in quite
       some detail, yes.
   Q.  I'm grateful, thank you.
           Can we look now at migration to Horizon Online.
       Now, this is something you would have more knowledge of,
       is that fair?  What happened during the migration to
       Horizon Online?
   A.  Well, I wasn't out in the field supporting it, so it
       depends when you say "more involvement".
   Q.  Shall we just look, if we may quickly -- I will take
       this reasonably quickly, just because I think you have
       had a role in talking about the introduction of
       Horizon Online.  Let's look at {F/555} please.  This is
       the Horizon Online induction training document.
   A.  Yes.
   Q.  And if we look at page 8 {F/555/8} can you see that
       Horizon Online:
           "... is about reducing the Post Office's operating
       costs ..."
   A.  Yes.
   Q.  "... and is a fundamental element of forward:
       three2eleven"?
   A.  Yes.
   Q.  What's "forward: three2eleven"?
   A.  It was the strategic plan at the time called
       three2eleven, so it was -- if I remember correctly, it
       was 2003 to 2011, the years.
   Q.  Okay.  This is a 7 December 2009 document.
   A.  Yes.
   Q.  And did you have some knowledge of what that strategy
       was at the time?
   A.  I had a knowledge because that was -- the whole business
       was engaged in what that was.
   Q.  Yes.
   A.  In terms of recalling the detail of it, I can't recall
       the detail of it at this point.  But it was a strap line
       that everybody was aware of, absolutely.
   Q.  Okay.  We can see from the slide that Horizon Online was
       based on the principle of business equivalence: "Will
       look and feel very different to existing Horizon but
       will not radically change transaction processes".  Yes?
   A.  Yes.
   Q.  And it has "been designed to run on existing branch
       equipment".
   A.  Yes.
   Q.  Then if you look in the notes, the sort of speaking
       notes under the slide, we can see again it is reiterated
       it is about reducing operating costs, yes, and that
       Horizon is a major cost and then do you see it is not
       about new functionality, on the left-hand margin just
       above "Business equivalence"?
   A.  Yes.
   Q.  "It is based on the principle of Business equivalence.
       Business equivalence is the primary design principle for
       the programme - we are not about new functionality."
           Yes?
   A.  Yes.
   Q.  So this wasn't really an upgrade in terms of
       functionality, was it?
   A.  No.
   Q.  It was business equivalence and a way that would be --
       there would be cost savings from doing it this way --
   A.  Yes.
   Q.  -- is that fair?
   A.  Yes.
   Q.  Let's look at {F/610}.  This is April 2010.  This is
       "Operations functional report" and if we look at the
       bottom, "Horizon Online high volume pilot", you can see
       a few bullet points there and the bottom one is:
           "Problems reported and decision made to halt further
       migrations allowing further investigation and resolution
       of issues."
           Now, first of all did you see this document at the
       time?
   A.  I don't recall seeing this, no.
   Q.  Were you aware of these issues at the time in the pilot?
       Did you know it hadn't gone --
   A.  No not -- no.
   Q.  At all?
   A.  No, not that I can recall, but I wasn't close to the
       absolute beat rate of how things were --
   Q.  Okay.  And did you have a look at any of this sort of
       documentary history before you wrote your witness
       statement?
   A.  No, I haven't --
   Q.  Research what happened?
   A.  So a broad understanding of what happened, but I have
       not seen these documents in that respect.
   Q.  Okay.
   MR JUSTICE FRASER:  Mr Green, you did interrupt again.
   MR GREEN:  I'm so sorry.
   MR JUSTICE FRASER:  Please don't.
   MR GREEN:  I'm sorry, my Lord.
           If we look please at {F/588} that's PEAK 0195380 and
       if we look at page 4 please {F/588/4}, 5 March 2010.  In
       the second box down -- do you see that?
   A.  In the green text, the big box?
   Q.  Yes, the big box.
   A.  Yes.
   Q.  "We have received notification from POL regarding the
       problems at this office ... on the 1st of March at the
       close of business we found that on node 5 the cash was
       short of £1,000.  All of the figures for that day match
       the figures presented at the time of each transactions.
       An instant saver withdrawal for £1,000 was transacted
       that day, but I was unable to find this transaction
       using the online report facility.  I feel very anxious
       as I believe a system error has occurred at the time of
       this transaction.
           "On the 2nd of March a transaction for a cash
       withdrawal was completed where the system commanded
       a member of staff to issue the money to the customer on
       screen but the receipt printed for that transaction
       printed out a decline slip.  The customer was honest
       enough to bring back the decline receipt a day later
       with the money.
           "On the 2nd of March on node 5 a £220 cash deposit
       was authorised on screen but 20 minutes later the
       customer brought back a receipt that stating the
       transaction had declined.  We contacted the NBSC as and
       when the customer produced the receipt.  NBSC stated
       that the transaction approved on the system and had no
       idea why the money was not deposited and why the decline
       slip was printed.
           "A rem was scanned in our system and all the figures
       had doubled up.  The helpline team was notified at the
       time, to which they seemed more confused as to why it
       happened than me!
           "Another error occurred on the system when 10 items
       of postage seemed to disappear for no reason halfway
       through a customer's transaction.  The system commanded
       no money to be taken from the customer on screen or by
       receipt.
           "A transaction on node 2 where a car tax was entered
       disappeared ..."
           And so forth.  Can you see?
   A.  Yes.
   Q.  This is a litany of issues that this person is raising
       and it says at the bottom, about six lines up from the
       bottom of that box:
           "I have a deep regret in initially volunteering to
       take part in this new pilot scheme, as I did not expect
       to have these complications with such poor services from
       the helpline.
           "Unfortunately our migration officer is away ...
       leaving us with no one with the means to correct our
       issue today.  Why should it be my liability to recoup
       all the losses in this already declining business, when
       these systems should have adequate contingencies for any
       such problems that could or would arrive.  It seems that
       the helpline have left me for seven hours now without
       any intention of calling us back as no one again wants
       to take ownership over this problem."
           Now, pausing there, were you aware at all of that
       level of difficulty being encountered by people who had
       volunteered for the pilot for Horizon Online?
   A.  No.
   Q.  Let's look at {F/614} please.  This is a Post Office
       document.  It is Horizon Online programme update,
       8 April 2010, do you see that?
   A.  Yes.
   Q.  And can we go to page 4 please {F/614/4}.  You can see
       there 614 branches live, almost 90% have a router
       installed, high volume pilot suspended, NFSP have raised
       concerns but remain supportive, business case and
       benefits secure, Fujitsu initiated red alert and
       independent reviews.  Were you made aware of that within
       Post Office?
   A.  No.
   Q.  Let's have a look at {F/639} please, the following
       month, May 2010.  If we go to page 6 please {F/639/6}
       because quite a lot of it is redacted.  We can see that
       the document is authored by Paula Vennells, the network
       director, and if we come back to page 3 {F/639/3}:
           "Executive correspondence (flag cases): the
       executive correspondence team are still receiving some
       letters from MPs despite the current purdah.  Each case
       is dealt with individually ensuring that we tackle any
       relevant customer issues but stay within current purdah
       guidelines.  Internally we are dealing with requests
       from subpostmasters who would like compensation in
       relation to closures following the Horizon issues over
       the Easter period."
           Were you aware there had been Horizon issues over
       the Easter period in 2010?
   A.  Not that I can recall, no.
   Q.  Okay.  Did you see this document, can you remember?
   A.  It would be hard to recognise it now, wouldn't it?
   Q.  It's very difficult, I agree.
   A.  So I don't think I did.
   Q.  Sort of striped with redactions, it's difficult.
           At {F/930} we've got a pack for James Arbuthnot and
       Oliver Letwin meeting scheduled for 17 May and if we
       look at page 2 of that {F/930/2} we can see who the
       attendees were: James Arbuthnot, who was then an MP for
       North East Hampshire, Oliver Letwin, Alice Perkins, the
       chairman of the Post Office, Paula Vennells, the
       chief executive, Susan Crichton and Lesley Sewell and
       then case review, Mr Ismay and yourself.
   A.  That's correct.
   Q.  And this was essentially the planning pack for that
       meeting, wasn't it?
   A.  Yes.
   Q.  And if we look at page 3 please {F/930/3}, in item 1,
       fourth bullet point, it says:
           "We are open to feedback and we will provide you the
       information we have available, our aim is to be open and
       transparent."
           Yes?
   A.  Yes.
   Q.  And the plan was that that was to be said at the
       meeting.  And if we look at page 12 {F/930/12} you can
       see there under "Background" do you see:
           "In 2010 Horizon underwent an upgrade.  The upgraded
       system was tested and has the full support of the NFSP."
           Bottom hole punch?
   A.  Yes.
   Q.  It wasn't really an upgrade from the perspective of
       functionality for SPMs, was it, as we have already
       identified?
   A.  No, agreed.
   Q.  And that doesn't give a completely clear explanation of
       that, does it?
   A.  No, it is quite limited in what it says there.
   Q.  And it could be read as meaning it had been upgraded
       from their perspective when the truth was it hadn't?
   A.  Yes.  I mean I don't know what was said around that
       because I wasn't in the whole of the meeting.
   Q.  And then it says:
           "The upgraded system was tested and has the full
       support of the NFSP."
           It is true, as far as we can tell from the
       documents, that NFSP was still supportive, but they had
       actually raised concerns about it, hadn't they?
   A.  Yes.
   Q.  We saw that on page 4 of {F/614}.  So that doesn't give
       the complete picture to the MPs either if that's what
       was said?
   A.  If that's what was said, because I wasn't party to that
       conversation.
   Q.  Let's look please now at {F/658} please.  It is the
       functional report operations, June 2010, and on page 1
       you can see Horizon Online:
           "Successful completion to pilot of new
       Horizon Online migration package.
           "Horizon Online high volume pilot successfully
       recommenced."
           So what had happened there was that the halt that
       had been placed on it had been lifted so it could
       recommence, is that fair?
   A.  Yes.
   Q.  And this is something that you must have known about at
       this time, is that fair?  You know the --
   A.  Sorry, I don't recall ...
   Q.  No?  You didn't know that the pilot had been stopped and
       then recommenced?
   A.  I don't recall that it had actually been stopped and
       recommenced, no.
   Q.  Let's look on page 2 please {F/658/2}:
           "Horizon Online losses: over the last month a small
       number of branches have raised concerns in regard to
       system integrity and the creation of losses.  So far no
       evidence of system integrity issues has been found."
           Were you aware that branches were raising system
       integrity issues at this time?
   A.  No, I don't recall any information flowing out of that.
   Q.  Let's look please at {F/657} which is the network
       functional report, June 2010 and if we look at page 3
       please {F/657/3}, Horizon Online, we've got:
           "High volume pilot resumed, 100 Crowns migrated with
       further 170 Crowns planned this week.
           "Ongoing delay to the programme impacting on
       resource available to deliver migration support.
       Support team continually reducing as people leave on VR
       terms ..."
           That's voluntary redundancy, isn't it?
   A.  Yes.
   Q.  So the support team was reducing because of voluntary
       redundancies, yes?
   A.  Yes.
   Q.  "Initiating contingencies (eg more subpostmasters) but
       to deliver the roll-out as planned we will now refuse
       any further requests for people to leave the team to
       take up roles in POL."
           So that was one facet of provision of support to
       SPMs who were affected by some of these issues, yes?
   A.  Yes.
   Q.  And if we go forward please to {F/708} this is a period
       12 09/10 sales report.  What is period 12?  Just to
       orientate ourselves by date?
   A.  That would be March.
   Q.  That would be March?
   A.  Yes.
   Q.  Of 2010?
   A.  We run April to March in our periods, yes.
   Q.  April to March, okay.  So March 2010 it looks as if
       we're in.  And if we can go please to page 8 of that
       document {F/708/8}:
           "Horizon migration - a number of issues have
       occurred with the data centre which have impacted [or]
       have affected live service to all Horizon Online
       branches.  Following on from this a decision has now
       been made to extend the duration of the pilot period to
       allow for further testing to be undertaken on the
       system.  As a consequence the planned date for the full
       roll-out schedule to commence has now been postponed
       until satisfactory progress has been made in achieving
       stability of branches within the pilot."
           Did you have any visibility of this at the time?  If
       you didn't I can take it even more swiftly.
   A.  I must have had some but I can't recall to be honest.
   Q.  You can't recall.  Did you look at any of this history
       to --
   A.  No, no.
   MR JUSTICE FRASER:  To what?
   MR GREEN:  In order to give the description you give in your
       witness statement?
   A.  Sorry, that's what I understood.
   Q.  That's what you understood?
   A.  Yes and the answer was no.
   Q.  I will take it a little bit more quickly then.  If you
       just go forward all the way to {F/1402}.  This is an
       extract from "Lessons Learned log" and under the area
       "Audits" at 4.1 we can see certain complaints about
       audits and so forth:
           "Existing branch discrepancies not discovered during
       audit of branch".
           And so forth.  And "Rationale for change:
           "Postmasters perceive a lack of knowledge of person
       conducting branch audit.
           "POL needs to be confident that colleagues
       performing audits are competent and there is confidence
       across the network.
           "POL need to be able to have clear sight of
       losses/gains throughout the network."
           Is this a document you are familiar with?
   A.  Yes.
   Q.  It is from 11 November 2015.  It has your name at the
       bottom left-hand corner, hasn't it?
   A.  Yes.
   Q.  Were you the author of the document?
   A.  Yes.
   Q.  You were?
   A.  Yes.
   Q.  And what was the purpose of this document?
   A.  So the purpose was to look at any areas that I thought
       we could improve.  So at this time -- so we had been --
       so we had -- branch support programme had been running,
       we had done the initial complaint to the mediation
       scheme and we had some business as usual areas that we
       were investigating, some issues as well, so what I took
       the opportunity to do then was look at what I believed
       were enhancements/improvements that we could make to how
       we operated to better support postmasters.
   Q.  And let's just identify the column headings.  The
       left-hand column is area, the second column is "Issue
       identified", the third column is "Rationale for change"
       and the fourth column is "Consideration for
       Post Office"?
   A.  Yes.
   Q.  If we go forward please to page 5 {F/1402/5}, 9.2:
           "Failure to be open and honest when issues arise eg
       roll out of Horizon, HNGx migration issues/issues
       affecting few branches no the seemingly publicised."
   A.  Yes.
   Q.  That was something you were specifically aware of,
       wasn't it?
   A.  So that was out of -- there were some cases that we had
       investigated as part of the mediation scheme.
   Q.  So you knew that there had been an issue about failing
       to be open and honest during both the roll-out of
       Horizon itself and the migration to HNGx which is
       Horizon Online, yes?
   A.  Yes.  But it was from a very limited source, it was just
       from what I had investigated as part of this.
   Q.  So in your investigations you didn't look at any of the
       sorts of documents that I have been showing you?
   A.  No.
   Q.  And do you accept that they might have helped to inform
       you as to the extent to which those complaints were
       justified?
   A.  Yes.
   Q.  If you had looked into it?
   A.  Yes.
   Q.  And the next box across says:
           "Need to understand and action learnings from
       Horizon, HNGx roll-out ..."
           We can't see what that says.  Without saying
       anything that's privileged, what is the point you are
       making there, in as far as you are able to tell
       the court?
   A.  Sorry, in which part, the middle bit?
   Q.  No, the top bit, where you say "Need to understand and
       action learnings from Horizon, HNGx roll-out ..."
   A.  So what I was trying to say was a broad -- a broad
       concept really of as we do things across the business
       then we should always look to see what lessons we could
       take from whatever we do, in some cases they could be
       very small things and in other cases they could be
       larger things and this was a document then that I was
       handing over into the business as usual environment as
       a lessons learned, so "These are what we have observed
       and these are the things that we would suggest", which
       is why it said "for Post Office consideration".
   Q.  Because you go on in the bottom of that box to say:
           "Adopting a mindset of that no hidden agenda/no
       'skeletons in cupboard' will demonstrate that lessons
       have been learned from Horizon, HNGx ... and business
       shows its willingness to change."
   A.  Yes.
   Q.  You felt that would be an important change, didn't you,
       and a helpful one?
   A.  Yes.
   Q.  What's proposed to do in the right-hand box is:
           "Use standard communications in branch to reinforce
       consistent message to customers and stakeholders."
           And:
           "Engage champions of POL and stakeholders throughout
       the process of change."
           That didn't actually specify any particular measures
       for openness or transparency, did it?
   A.  No.  No, that was quite broad.  I mean so these were say
       areas of consideration, they weren't really specific.
       In some cases I would have been more specific, in others
       I was being quite broad.  What I meant by that were
       things around -- so the branch user forum, for instance,
       is actually bringing in postmasters and people from
       within the business to take their B1 input and to
       share -- to get their insight into things.  That was in
       a broad sense.
   Q.  Do you know what happened to that proposal?  What action
       was taken, if any, to implement what you had suggested?
   A.  I don't -- I don't know what happened to that one.
   Q.  Is there any evidence that you are aware of something
       actually happening?
   A.  With branch user forum, yes, things did happen, as you
       have seen and that was a very, very open forum and that
       was used to engage particularly product managers or any
       changes that were happening to the network would come to
       the branch user forum and share in confidence a very
       early view and take input and that kind of fed through
       in terms of the new technology.  IBM we were talking to
       at the time in the front office and also the hardware
       swap out that we did later on as well.  Those are the
       kind of things.
           So some of it did happen, but I can't say I have
       tracked that through.
   Q.  Is it fair to say in the branch user forum, when we
       actually look at the branch user forum logs a lot of
       that is actually feedback from SPMs to Post Office,
       rather than Post Office saying "Look, we're aware that
       this has gone wrong and so" --
   A.  It changed actually.  So when it was first set up it
       very much was that, so it was taking a lot of input from
       postmasters and there was a Crown VM there as well, an
       ABM, so it was across the business.  It was a lot of
       taking input and then it changed.  As it got more --
       I suppose as everybody got more comfortable and we
       addressed some small areas and some quick wins in that,
       it changed then to being seen as the forum to come to
       get input to things like, you know, the new hardware,
       the keyboard, the layout of the keys, for instance, on
       the keyboard, even down to that basic level around what
       that layout should look like in terms of the operation,
       but other things as well.
   Q.  So sort of just looking at a practical point on that is
       if you have buttons very close together to each other
       which can easily be confused, that tends to increase the
       risk of miskeys, doesn't it?
   A.  No, what it was on this -- there were two or three
       keyboards that were being considered, size, shape, and
       what we were asking -- you know, postmasters in the
       forum, the users, "From how you operate the system,
       which keyboard suits you better around the functionality
       of easy of operation?"
   Q.  You mean a normal keyboard --
   A.  Yes.
   Q.  -- not how it is presented on the Horizon scene?
   A.  This was about the hardware, which was --
   Q.  Not the touch screen?
   A.  So that would be the software.  This was about the
       hardware.
   Q.  Okay.  Let's briefly turn, if we may --
   MR JUSTICE FRASER:  Now, Mr Green, I am going to suggest you
       stop now.
   MR GREEN:  I will.
   MR JUSTICE FRASER:  We've got a number of matters to deal
       with.
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  This is going to take about ten minutes.
       You are very welcome to stay in the witness box if you
       would like to, but you don't have to.  You are still
       going to be giving evidence tomorrow so please don't
       discuss the case.
   A.  Of course.
                           Housekeeping
   MR JUSTICE FRASER:  I have some items and then we will come
       back to some of the other items.
           Mr De Garr Robinson, I am going to ask you to do
       a task please, which is a redaction review.  I have been
       noting the different documents that have been redacted
       and it seems to me that some of them are crying out for
       a review by leading counsel of the basis on which they
       have been redacted.  I will just go through the list.
       There is F/1549, F/1225, F/1664, F/610, F/639, F/658,
       F/657, F/708 and F/1402.
           I'm obviously not going to impose a time on you and
       perhaps tomorrow you could just give me an update as to
       how long you think it might take you.  It might be
       necessary to recall witnesses, hopefully it won't be,
       but in the first instance I'd like you to review those
       reactions please.
   MR DE GARR ROBINSON:  My Lord, yes, so it is F/1549, F/1664,
       F/1225, F/610, F/639, F/658, F/657 and F/1402.
   MR JUSTICE FRASER:  And there was F/708 as well.
           The first couple appeared I think probably at the
       beginning of the afternoon, the last half a dozen have
       all been within the last 25 minutes.
           That's the first point.  The second point is --
       I mean simply as an example, Mr De Garr Robinson, to
       explain, F/708 is the monthly sales report
       for March 2012.  It has got quite a few redactions in
       it.  All of those I'm asking for a specific reason
       I just think they are -- it is sensible to have it
       reviewed.
           Secondly, there's the outstanding point relating to
       7/8/9 May which I mentioned at the end of last week.  Do
       any of you have any issues with it being the Tuesday and
       Wednesday rather than --
   MR DE GARR ROBINSON:  I have no issues with the 7th or the
       8th.
   MR JUSTICE FRASER:  I didn't think you would have.  Thank
       you very much.  Mr Green?
   MR GREEN:  My Lord, one member of our team can't do the 7th
       at all and it is not easy for the rest of us but we can
       do it.
   MR JUSTICE FRASER:  All right, 7th and 8th then, I'm afraid.
       It is going to have to be.
   MR GREEN:  I understand.
   MR DE GARR ROBINSON:  My Lord, that's true of both -- I have
       similar issues but ...
   MR JUSTICE FRASER:  As far as I'm concerned, unless it was
       physically impossible it's going to be the Tuesday and
       Wednesday, so it's going to be the 7th and 8th.
   MR GREEN:  We understand.
   MR JUSTICE FRASER:  I will keep -- because I'm in an
       indulgent mood I will keep the time for your lodging
       your closings as it was before.  So you won't lose
       a day.
   MR DE GARR ROBINSON:  I'm afraid I have forgotten when our
       closings are due.
   MR JUSTICE FRASER:  I believe it was -- well, Mr Green is
       about to tell me when it was.  Do you remember when it
       was?
   MR GREEN:  I've got it I think in a different notebook.
   MR JUSTICE FRASER:  I think your junior behind you is trying
       to ...
   MR GREEN:  My Lord, I think your Lordship had a day in mind.
       I'm not sure we actually crystallised what it was.
   MR DE GARR ROBINSON:  I don't remember a day being actually
       specified.
   MR JUSTICE FRASER:  Well, then, that's my fault and
       I apologise.
   MR GREEN:  It may be ours.
   MR JUSTICE FRASER:  I will give you until noon on the Friday
       before.
   MR DE GARR ROBINSON:  That's very kind of your Lordship.
   MR JUSTICE FRASER:  Because I have taken a working day off
       you, so that's up to me.  And the Monday is in any case
       a bank holiday but that's not a problem from my point of
       view.  So noon on the Friday before, which I think is
       3 May.
           Right, well, that -- just give me one second.  Are
       we having an opposed application at 10 o'clock tomorrow
       for a third party disclosure order against Royal Mail?
   MR GREEN:  My Lord, I was expecting that Mr Warwick would be
       coming over to do that at 10 o'clock.
   MR JUSTICE FRASER:  That's why I'm asking.
   MR GREEN:  Precisely.
   MR JUSTICE FRASER:  I don't have the witness statement that
       you issued in respect -- I don't need it this minute but
       if you have got one --
   MR GREEN:  We've got a copy just in case --
   MR JUSTICE FRASER:  Right, hand that up please.
   MR GREEN:  I'm afraid it has not been hole-punched.
   MR JUSTICE FRASER:  That doesn't matter.
   MR GREEN:  My Lord, I have literally just been passed
       a yellow sticker saying "Some E&Y audits have arrived".
   MR JUSTICE FRASER:  All right, well, you can update me
       tomorrow, or through my clerk later on.
   MR GREEN:  I'm most grateful.
   MR JUSTICE FRASER:  But these are the papers for tomorrow at
       10 o'clock?
   MR GREEN:  My Lord, they are.
   MR JUSTICE FRASER:  Do you know -- we will just have to take
       that as and when.
   MR GREEN:  If we find out anything this afternoon we will
       let your learned clerk now and then --
   MR JUSTICE FRASER:  All right, that's excellent.
           That's all my housekeeping.
   MR DE GARR ROBINSON:  My Lord, I have one piece of
       housekeeping which is your Lordship asked for an order
       to be prepared reflecting what your Lordship indicated
       this morning about the witness statement.
   MR JUSTICE FRASER:  Yes please.
   MR DE GARR ROBINSON:  Could I hand this up and then I will
       hand it to my learned friend.
   MR JUSTICE FRASER:  Thank you very much indeed.
           (Pause).
           I have just signed that -- well, I have read it,
       checked it, signed it.  I will give it to my clerk and
       it will be sealed by the QB.
           So no more housekeeping?
   MR DE GARR ROBINSON:  My Lord, there is one other matter
       which is one of my witnesses, Mr Membery.  Your Lordship
       may recall from the PTR that he has a serious illness
       and your Lordship may recall that we indicated that he
       would be having tests in hospital this week and we would
       do our best to ensure that it was possible.  My Lord, he
       is currently due to be giving evidence on Wednesday.
       The difficulty is he is in hospital today for tests --
       this is to do with prostate cancer and --
   MR JUSTICE FRASER:  There is no need to go into -- you don't
       have to go into details.
   MR DE GARR ROBINSON:  But he has been told that he needs to
       attend for further tests tomorrow and Wednesday.
   MR JUSTICE FRASER:  All right.
   MR DE GARR ROBINSON:  And at the moment he is not sure
       whether he will be needed on Thursday as well.
   MR JUSTICE FRASER:  I will tell you what we should do with
       him, there is flexibility in the timetable next week,
       without -- I don't want that gentleman to feel under any
       pressure at all.
   MR DE GARR ROBINSON:  I'm grateful.
   MR JUSTICE FRASER:  Take instructions, if with some sensible
       interleaving or creating of what I imagine based on the
       length of his witness statement is not going to be more
       than half a day at the most if not probably an hour with
       Mr Green, that could be fitted in either the week we're
       not sitting, or if necessary some other time.  But
       I don't want him to feel under pressure to come here if
       he is actually in hospital at the moment.
   MR DE GARR ROBINSON:  I'm very grateful to your Lordship.
   MR JUSTICE FRASER:  We will deal with that creatively.
           Mr Green, anything from you?
   MR GREEN:  My Lord, the only thing that occurred to me, just
       thinking about the sticker that has just arrived --
   MR JUSTICE FRASER:  The sticker?
   MR GREEN:  Sorry, the E&Y audits that apparently some have
       now been received.  Would your Lordship be amenable to
       us requesting, if appropriate, that what's due to be
       heard tomorrow morning at 10 might be postponed until
       the following morning at 10 if it looks as if we are
       likely to have greater clarity on it?
   MR JUSTICE FRASER:  Yes, but you will have to take relevant
       steps with Royal Mail and whoever they have
       instructed --
   MR GREEN:  My Lord, of course.
   MR JUSTICE FRASER:  And you will have to notify the court.
   MR GREEN:  It may be easier just to come at 10 tomorrow and
       get it clear.
   MR JUSTICE FRASER:  Let's put it this way: I'm here anyway
       from quite a lot earlier than 10 so it is not a problem
       for me to come down at 10 and go back upstairs and I'm
       flexible.  But you have my clerk's email address so it
       is not an issue.  If you would rather do it on
       Wednesday -- well, just see how the land lies and just
       decide.
   MR GREEN:  I'm grateful.
   MR JUSTICE FRASER:  But if it is to be moved from the date
       in the order I made on Friday, you will have to notify
       Royal Mail of that.
   MR GREEN:  My Lord, we would of course do that.
   MR JUSTICE FRASER:  What I don't want is Royal Mail coming
       at very short notice and the claimants not being here.
   MR GREEN:  Of course.
   MR JUSTICE FRASER:  All right.  Is that everything?
   MR GREEN:  My Lord, yes.
   MR JUSTICE FRASER:  Thank you all very much.
           Actually there is one point I should add to
       everyone.  If there is an opposed application tomorrow
       at 10 o'clock, Mrs van den Bogerd, there is an outside
       chance that we might not be ready to resume your
       evidence bang on 10.30.
   A.  Okay.
   MR JUSTICE FRASER:  So please don't be concerned about that.
       If you turn up and you look through the window and you
       see -- I mean it will be open to the public, but you
       probably won't want to come in given you will be giving
       your evidence.  If we're not able to start bang on 10.30
       with you then I will probably start at say quarter to
       11, but if the Royal Mail do come tomorrow, I imagine it
       won't take a full 30 minutes, so it probably won't
       arise, but just to tip you off basically, so you don't
       worry that we have started without you.
           Right, so, 10 o'clock for you, 10.30 potentially
       10.45 for the Post Office, unless you want to sit in on
       the Royal Mail.  Thank you all very much.
   (4.32 pm)
         (The court adjourned until 10.00 am on Tuesday,
                          19 March 2019)