Archive
TR

The user code in the Lepton record

A bill payment was reversed at Lepton in October 2012.

CSINT-NEWS-2026-010Open in archive

Reading. Findings published by other researchers, courts or institutions, read from the source and retold. The findings are theirs.The findings belong to their sources.

Documents

The yellow highlights were added later. The original image is unchanged.

Who made the reversal?

Credence, Post Office's data for reviewing branch work, names the subpostmaster.

  1. He denies it
  2. The user code
  3. Number keyed in

Post Office, Helen Rose report POL00030214, DRAFTPDF page 212 June 2013

Evidence published by the Post Office Horizon IT Inquiry. Personal details are covered by the inquiry's GRO redactions, which are kept as they are.

What does 118 say?

Fujitsu blames recovery, the system tidying up a session itself after a failure.

  1. On logging on
  2. Not the clerk

What it does not show. 118 is a Counter Mode Id value, not an error code.

Post Office, Helen Rose report POL00030214, DRAFTPDF page 2, extract from an e-mail of 30 January 201312 June 2013

Evidence published by the Post Office Horizon IT Inquiry. Personal details are covered by the inquiry's GRO redactions, which are kept as they are.

Only in the raw logs

The ARQ, the standard extract requested from Fujitsu, shows this only indirectly.

  1. Not explicit
  2. System-made
  3. Reboot only suspected

What it does not show. I have not seen the raw logs; the reboot is unproven.

Post Office and Fujitsu, e-mail thread POL00134138PDF pages 2 and 311 February 2013

Evidence published by the Post Office Horizon IT Inquiry. Personal details are covered by the inquiry's GRO redactions, which are kept as they are.

A missing column

Jenkins calls a recovery column easy to add, and Penny confirms a change request is the route.

  1. The extra column
  2. The process question
  3. Rose's separate code

What it does not show. A proposal, not proof the column was added.

Post Office and Fujitsu, e-mail thread POL00134138PDF pages 1 and 213 February 2013

Evidence published by the Post Office Horizon IT Inquiry. Personal details are covered by the inquiry's GRO redactions, which are kept as they are.

What the court saw

The court finds the record named the subpostmaster, though the system made it.

  1. Record and event differ
  2. Used in investigations

What it does not show. It covers this reversal, not every Horizon transaction.

High Court, Bates v Post Office Ltd (No 6) [2019] EWHC 3408 (QB)Paragraph 227, PDF pages 72 and 7316 December 2019

The judgment as published in PDF on judiciary.uk.

Full account and source transcriptMethod, timeline and full text. The findings are in the documents above.

The claim reviewed

The Credence data put a reversal of 4 October 2012 under the code JAR001, and the subpostmaster said he had not made it. The question is whether that code shows who started the transaction.

Limits of the evidence

  • The Rose report is marked DRAFT, version 1.
  • I have not seen the raw logs and have not reproduced Horizon's behaviour. Everything read here is in published documents.
  • The court's finding, the technical explanation of the time and the short notes on this page are separate things.

Context of the documents

According to the Rose report's summary, a British Telecom bill payment of £76.09 was reversed at the Lepton branch on 4 October 2012. The branch settled the Transaction Correction it was issued. The subpostmaster said he had not made the reversal and engaged a forensic accountant.

The report's first page gives its last edit date as 12/06/2013. Its recommendation is a change request so that every reversal created by the system can be told apart in both Fujitsu and Credence data.

Paragraphs 224 to 226 of the judgment record that the Post Office witness Mrs Van Den Bogerd accepted in cross-examination that the reversal was generated by the system. In the cross-examination quoted later in paragraph 227, the same witness says she does not believe the change request was acted on. That is a witness's belief, not a finding of the court.

Sources and method

The three PDFs were downloaded from their original addresses on 17 October 2026, the Rose report and the e-mail thread from the Post Office Horizon IT Inquiry site and the judgment from judiciary.uk. The files are kept unchanged.

Their SHA-256 values, in order, are as follows. POL00030214.pdf 2c726a6cf579983e76a61dfceb0d9599df4647a7cf826ac291445863b15ef6c4, POL00134138.pdf 2406460454f5d8d9adae6d05b44cbe21bca60351ed07e89b5dff422d9a4eb085, bates-v-post-office-judgment.pdf 65c65586199bcecec16cb6d920f4bbfe0f6d5d4f8eadd03d73d9de6ef652cea1.

The images are renders of the PDF pages made with PyMuPDF at 200 DPI. Where a passage runs past a page break, crops of the two pages sit one under the other in page order with a plain grey band between them. No pixel was changed, and the inquiry's redactions stay as they are.

The yellow highlights are not part of the images. Their boxes were measured from the PDF text layer and sit on a separate layer. With the highlights switched off, the document shows unmarked.

Sources