Honesty comes before all else.
So none of the criminal trials ever got as far as actually looking at the disputed figures, because the defended would plead guilty or the PO would just lie about how the system worked to "prove" false accounting. As Justice Fraser wrote in the 2019 common issues judgement:
"This means that, even for disputed items going back as far as the year 2000, this litigation is the first time that there will be any independent consideration of disputed items showing in the branch accounts for the vast majority of the Claimants."
Indeed many of the disputes were not even caused by software errors but human error on the part of the SPM. Errors that they attempted to resolve but could not due to the way PO handled disputes.
So this is far more than just a code quality problem.
This should be something for the HN community to take note, as this scandal is for Fujitsu the chickens coming home to roost for an issue that started in 1999. We as software developers, the maker of the tools and the facilitators of change and advancement have responsibilities, and those responsibilities can have wide reaching consequences.
We saw in Myanmar that not understanding the locality and not having adequate moderation can have consequences, we see with this case that absolute faith in a system can be damaging and mixed with other factors (such as the UK's obscure Draconian laws around the mail service being able to prosecute people) can be catastrophic.
So, friends, when you are approaching a deadline and you feel as though you need to ship even though you know the product is bug riddled, just take a moment to write some extra tests or to document the shortcomings more thoroughly. Someone you don't know may be depending on it.