> Fallibility is fine, the problem is that Fujitsu and the Post Office were so scared that admitting fallibility was somehow a massive moral failing. It's not.

> Fallibility is fine, the problem is that Fujitsu and the Post Office were so scared that admitting fallibility was somehow a massive moral failing. It's not. Fujitsu was a contractor; the contract was a PPI (Public-Private Investment) contract. The deal was that Fujitsu developed the system at their own cost, and ran it in exchange for a cut of each transaction. I know something about this work; back in the 1980s, I worked on a bid for what was then called Post Office Counter-mechanization. This was probably the third attempt to automate Post Office counters; previous attempts had failed. FWIW, we lost the bid; I think it was won by ICL, the forerunner of Fujitsu. That third attempt failed, of course. To work on this bid, I was loacked away in an office on my own for 6 months. The requirements were delivered as two metres of lever-arch files, full of paper forms; each form described an input screen, or a report, or a processing operation, or a storage requirement. 90% of these forms were blank, apart from the name of the operation. It was impossible to do a serious job when so little of the required data was supplied. And we weren't allowed to talk to the Post Office about the requirements; that would have risked the entire tender being overturned in court. If the Fujitsu PPI contract was handled in the same way by the Post Office, and I assume it was, then it would have been impossible for Fujitsu to deliver a working system. But now we see Alex Chalk, the (current) minister in charge of the Post Office, suggesting that the compensation money should be extracted from Fujitsu. It may turn out that Fujitsu have some hard questions to answer; but for now, it looks as if this mess falls entirely on the shoulders of incompetent Post Office managers.