FUJ00165082
FUJ00165082
QUESTIONS ON RW AND SSC COMMENTS
ON DAYS 7 AND 8
DAY7
1. Pages 92 to 93 — Godeseth says that he has read a whole number of peaks showing the
standard practice in Fujitsu was to label transaction insertions into riposte to make it is clear
as possible that an insertion was being made —
(1) Could we have a list of peaks showing this being done, maybe by inserting an attribute
(as described in page 93) or calling it (what?) a different counter number (as described
in page 94).
(2) Could it be explained how calling counter I counter II worked, precisely?
2. Page 95, comment WM9 (from SSC) — it is said that SSC process required that we inform
the SPM when injecting transactions. Is that definitely right? Where is it evidenced?
(1) I Also, later on Godeseth was cross-examined about I transaction where the SPM
appears not to have been informed (see day 8, page 8), but it is unclear whether this
comment was made in relation to the branch transaction injection or the TIP Repair
injection — please confirm and provide comments.
3. Pages 105 and 115, comments R14 and R21 (from RW) — can FJ explain why the initial
project for a post office and DSS system was a disaster and had to be abandoned in favour of
a single project for post office does not in any way undermine his point that the original
horizon system represented a Greenfield?
(1) I What should I put to Coyne when he says the opposite?
4. Page 137, comment R43 (from RW) —
(1) Can Fujitsu please explain whether credence was a pure management information
system for post office or whether it was also part of the critical path information going
into Horizon as Godeseth described here?
FUJ00165082
FUJ00165082
(2) Can Fujitsu also explain whether there is any reason to think that there were data
integrity or other problems with credence making it an unsafe system to use for this
purpose?
(3) In the light of comment R 43, does RW need to correct his reports?
Pages 145 to 144, comments R47 and R48 — RW thinks there is an inconsistency here:
(1) Can Fujitsu explain whether the audit store was inside or outside riposte?
(2) Does RW need to correct his reports?
Pages 151-158. What is the position with respect to the document at F/1692 (post office
account — customer service problem management procedure)?
(1) I Was it issued for approval (and does is this what “.0” means)?
(2) Was it actually adopted/brought into effect?
(3) I Godeseth’s evidence in his witness statement was that this proposal was abandoned
when Salawu was replaced true. If so, why were further versions continually worked
on following Salawu’s replacement?
Pages 184 to 185, comment R87 (from RW) — what is the journal table dump referred to
here, exactly?
(1) Where is it stored on what does the table do?
(2) I Why can duplicate JSNs exist in this table — or rather, why isn’t it a problem that they
can exist in this table?
Page 187, comment R91 (from RW) — RW does not understand what Godeseth means when
he says that the BT tool was designed for one-sided transactions. Nor do I. Can Fujitsu
explain?
Page 187, comment on 92 (from RW) — RW points out that the relevant problem is a
harvester exception. Can it be precisely explained what a harvester exception is and why
such an exception might create a discrepancy in branch — why wouldn’t it simply create a
back-end problem?
10.
FUJ00165082
FUJ00165082
Page 190, comment WMD 26 (from SSC) — here, SSC seems to accept that this was a
transaction which got into the system even though it was not zero-sum basket. But I thought
it was impossible for non-zero-sum baskets to get into the system. What am I missing?
DAY 8
12.
Page 5, comment A2 — SSC say they are looking into F/432. Any update?
Page 8, comment A3 — in what sense is this comment uninformed?
(1) I Was it silly or stupid to ask whether the transaction may have caused a receipts and
payments error? If so, why.
Page 12, comment A6 — can it be explained why adding the discrepancy value into the
current value in the POLFS feed led to a transaction of just over $2000 being inserted in the
post office system?
Pages 13 to 14, comments A7 and A8 —I do not fully understand what is being said in A7
and thus I do not fully understand why it is wrong to say that the additional $1000 loss was
(or may have been?) a result of the difference in the 2 corrections made. Can this be
explained more fully please?
Page 20, comment AI! — can the change control documents showing that change controls
always required post office approval for any support action with an impact to stock value or
receipts and payments value be listed, please?
(1) I And can the documents specifying whether such consent approval was to be given
pursuant to an OCP or an OCR be listed question?
Page 25 — can it be explained why, if the BT was needed in relation to the transaction
discussed here, it wasn’t needed in relation to many other transactions?
(1) In other words, were there other procedures which could have been used to cure the
problem not involving remote access, and if so, why was the BT used here, and if not,
why hasn’t the BT been used in other similar cases?
(2) Was this case totally unique?
FUJ00165082
FUJ00165082
Pages 80 to 81, comments A26 and a 27 — by saying “no” and “incorrect” here, SSC appears
to be saying that the relevant bug did indeed only occur in September 2010. Is that really the
position?
(1) _ Ifso, are all the other mismatches in earlier months definitely due to other causes not
involving a bug?
(2) Ifnot, why say “no” and “incorrect” here? Am I missing something important?
Page 112, comment A31 — it is suggested here that my re-examination “got confused”. Here,
I was putting a point that Fujitsu had suggested that I put. In what sense was I confused and
what needs to be corrected or clarified to resolve the confusion?