Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
If the speed camera has not been calibrated, according to the manufacturers instructions for periodic servicing, why should we trust it?
To start with, the Horizon system had no internal self-audit.
The system I am working on currently, writes logs and creates database entries for *everything* it does. This is to create a deliberate audit trail, so that you can follow every transaction as it moves through the system. A timeline of actions.
...or so the IT guys tell us.
At least it tries. The point is also that if you don't find something in the audit sub-system(s), but it happened, then alarm bells should start ringing,
The way that Horizon swallows evidence is quite suspicious to me.
Yes I agree, having worked in IT all my career I wholeheartedly agree that transaction logging is a must for systems that are susceptible to fraud or the suspicion of it.
I hope, vainly, that the question of where the money went will eventually be addressed, at some length.
There wasn’t any money that went anywhere, except in the mind of the computer.
If I give you £2, you have £2. If you computer says I gave you £20, you still have the correct £2, even if your boss’s lawyer looks at the computer and says you’re £18 short.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
I actually felt slightly better-disposed towards Kissinger, after reading that, which is usually the case with polemics.
Rolling Stone’s take on foreign affairs is no more attractive than Kissinger’s.
I often find that when I read about someone I have negative views on, in more depth, I am more sympathetic towards them, which can be really annoying as I am expecting my prejudices to be reinforced.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
When you present text messages in court, you don't just hand the judge your mobile. There is a process by which (for text messages) the mobile company provides "legal grade" evidence, based on data from their servers.
Yes but I assume that is reported and audited (like should be the case in all good IT systems)
@NickPalmer was suggesting a full independent review of all the code if there’s an accusation it’s false. Is that really feasible or is it a (retired) politician just suggesting something that sounds good (even if he has a background in IT)
That is not feasible on a court case by court case basis.
But for, say, ATM transactions, the system should be previously audited and signed off by regulators as good/bad/shit.
It should be -
1) The system is audited and verified, on a regular basis 2) The system is audited and certified as part of the court case
In banking, there is already quite a lot of audit of systems, and more is coming.
The problem we are looking at has two problems: Justice requires that the prosecution prove their case, and the burden is on them. Placing burdens on the defence invites malicious or mistaken prosecution - the latter seems to have occurred wholesale in the PO scandal.
It is wholly unfeasible to prove your case if you cannot rely upon the infallible nature of automated systems in a manner which admits vast amounts of hearsay evidence.
In finding a solution, the only one I can think of is to give the judge a guided discretion as to when to require strict proof of evidence X with equal access for both sides to expert evidence.
The PO case would be an obvious example, where (SFAICS) there was generally no corroboration, and the reliability of IT was effectively 100% of the case.
Of course as well as protecting the innocent, such a move would provide a field day for the guilty seeking to complexify basically simple evidence, like phone calls and messaging.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Something like a speeding fine isn't minor if someone drives for a living and their contract of employment or local authority licence depends on a certain number of points (taxis, HGVs etc). And the courts often do decide the camera was potentially faulty, as Dura Ace will describe in great detail shortly.
The IT stuff is a red herring. There is complexity in almost everything, whether it's the engineering of railways, financial services, environmental protection. We can't expect the courts to understand it all, nor us.
The error was not with an ignorance of the IT, but with the broader legal system that could see hundreds of these prosecutions go forward without someone checking the victims' bank accounts/beds/spending behaviour and thinking "huh - where is all this stolen money going?".
Hm. The PO employs its own investigators and prosecutors. The police don't have anything to do with it. Maybe that has something to do with that mentality?
In something like a banking audit, you'd look at the broad financial trends to see if they fit with your expectations of the business and the wider economy, and have a look for any weird transactions or patterns. If you find something, you'd check if it was one person or branch or team, or distributed across the firm. If the latter, you'd set your IT specialists on it and see if there is some sort of financial reporting error.
Simple stuff really. This wouldn't have got past a vaguely interested graduate internal auditor if they had access to all the data. I assume this is where the PO's prosecutors took over.
Some interesting points but it is more law than politicalbetting really and quite a long article again
No - it is about Parliament: how it operates and what it should do now and about its responsibilities to voters, about the proper balance between state and citizens - see the section on fairness which sums it up, which indeed sums up why all these scandals - not just this one - matter.
The relationship between state and citizen is the very essence of politics. If you can't see that what on earth are you doing on a politics site?
(And it is shorter than yesterday's.)
Still a bit long for HYUFD to comprehend, clearly.
Like some others, I confess the PO scandal would barely be on my radar if not for your headers. Thank you.
Why are the news outlets not making more of it? Could it be that the arts graduates who run them are struggling to understand because it's, you know, techy stuff?
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
I hope, vainly, that the question of where the money went will eventually be addressed, at some length.
Which money?
In many cases subpostmasters put their own money in to make up the discrepancies. The PO took this money. It was effectively taking money it was not owed - theft in short.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
When you present text messages in court, you don't just hand the judge your mobile. There is a process by which (for text messages) the mobile company provides "legal grade" evidence, based on data from their servers.
Yes but I assume that is reported and audited (like should be the case in all good IT systems)
@NickPalmer was suggesting a full independent review of all the code if there’s an accusation it’s false. Is that really feasible or is it a (retired) politician just suggesting something that sounds good (even if he has a background in IT)
That is not feasible on a court case by court case basis.
But for, say, ATM transactions, the system should be previously audited and signed off by regulators as good/bad/shit.
It should be -
1) The system is audited and verified, on a regular basis 2) The system is audited and certified as part of the court case
In banking, there is already quite a lot of audit of systems, and more is coming.
I agree with all that. But it’s necessary not sufficient for a criminal conviction in my view
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Something like a speeding fine isn't minor if someone drives for a living and their contract of employment or local authority licence depends on a certain number of points (taxis, HGVs etc). And the courts often do decide the camera was potentially faulty, as Dura Ace will describe in great detail shortly.
The IT stuff is a red herring. There is complexity in almost everything, whether it's the engineering of railways, financial services, environmental protection. We can't expect the courts to understand it all, nor us.
The error was not with an ignorance of the IT, but with the broader legal system that could see hundreds of these prosecutions go forward without someone checking the victims' bank accounts/beds/spending behaviour and thinking "huh - where is all this stolen money going?".
Hm. The PO employs its own investigators and prosecutors. The police don't have anything to do with it. Maybe that has something to do with that mentality?
In something like a banking audit, you'd look at the broad financial trends to see if they fit with your expectations of the business and the wider economy, and have a look for any weird transactions or patterns. If you find something, you'd check if it was one person or branch or team, or distributed across the firm. If the latter, you'd set your IT specialists on it and see if there is some sort of financial reporting error.
Simple stuff really. This wouldn't have got past a vaguely interested graduate internal auditor if they had access to all the data. I assume this is where the PO's prosecutors took over.
Nowadays you audit almost every single mouse click.
And each system audits individually so you can check that the audit details matches across systems
On AI, I approach a lot of analysis and opinion with scepticism; I have no doubt that it is MASSIVE. Huge. And underappreciated as agent of fundamental societal change.
But that said, like crypto, blockchain and various other tech fads it is also a magnet for Tate-osphere nonsense. And unlike e.g. NFTs, which I think to all clear-thinking folk was patently a scam, it is actually quite difficult for the inexpert to sort through the chaff.
With that in mind, I did find something genuinely useful from my own field (marketing & advertising), using LLM to build perceptual maps of perceptions of car brands. For an academic article, it's a fairly lucid read and quite an interesting case study of the sort of things LLM will soon be able to do for us, and will no doubt have equivalent examples in lots of other fields.
I should have said early (distracted by @leon's attempts to deflect the conversation away from @Cyclefree's header and the conversation on Kissinger):
Excellent thread header @Cyclefree. Very thought provoking, although I do find the whole thing depressing. This stuff just repeats itself over and over again. Contrary to the statement always made, lessons are never learned.
I hope, vainly, that the question of where the money went will eventually be addressed, at some length.
There wasn’t any money that went anywhere, except in the mind of the computer.
If I give you £2, you have £2. If you computer says I gave you £20, you still have the correct £2, even if your boss’s lawyer looks at the computer and says you’re £18 short.
It's a form of quantitive easing really.. creating money where it hadn't existed previously
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
Trying to save money was probably one of the reasons behind the removal of S.69. But you make an important general point - one which I have tried to make repeatedly - persistent underfunding of the criminal justice system leads to injustice.
He's right. I listened to extracts from yesterday's HO Select Committee hearing. I normally defend Civil Servants, but the two who appeared yesterday (Rycroft + another) were embarrassingly hopeless - ill-prepared and incapable of answering basic questions. This led to the bizarre scenario of both the Chair (Diane Johnson, Lab) and Lee Anderson joining forces to tell them how useless they were.
It would appear that in their desperation to meet targets for removing the backlog, at least 17,000 asylum seekers have simply had their claims discontinued. When asked where they were, the response was simply "no idea".
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
If you imagine anyone within the legal system who actually has any authority over how it works is going to go down that route, I will show you an NHS union activist who knows how much expenditure on the NHS would be too much or a Home Secretary who knows what is the acceptable level of murder in our country.
I actually felt slightly better-disposed towards Kissinger, after reading that, which is usually the case with polemics.
Rolling Stone’s take on foreign affairs is no more attractive than Kissinger’s.
I often find that when I read about someone I have negative views on, in more depth, I am more sympathetic towards them, which can be really annoying as I am expecting my prejudices to be reinforced.
I daren't pick up a book on Trump.
I wouldn't worry about that. I've read a few on him and however bad you think he is beforehand, it turns out to be worse.
He's right. I listened to extracts from yesterday's HO Select Committee hearing. I normally defend Civil Servants, but the two who appeared yesterday (Rycroft + another) were embarrassingly hopeless - ill-prepared and incapable of answering basic questions. This led to the bizarre scenario of both the Chair (Diane Johnson, Lab) and Lee Anderson joining forces to tell them how useless they were.
It would appear that in their desperation to meet targets for removing the backlog, at least 17,000 asylum seekers have simply had their claims discontinued. When asked where they were, the response was simply "no idea".
It’s almost as if there’s a problem with certain elements of the Civil Service going out of their way to undermine the government in a number of policy areas.
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
Hmm, good question. I think it depends on the severity of the punishment. I'd accept an 0.1% risk that I was improperly convicted of, say, shoplifting, with a fine and embarassment, but if the risk is life imprisonment then I want the risk to be more like 0.0001%. With a death penalty then the figure becomes precisely 0, which is partly why I'm against the death penalty.
I remember the case of a prisoner released 20 years after a wrongful conviction due to a botched lab test. Asked by a journalist eager for a scathing quote what he felt about the lab technician responsible, the ex-prisoner said mildly, "Well, we all make mistakes." The very best tradition of British understatement.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
What the auditors were doing needs to be known by others as well as God. Any proper audit would discover the discrepancy between IT alleged money/transactions and money/transactions existing in the real world. If the discrepancy can't be/isn't disclosed then it isn't a proper audit because the one thing a proper audit CAN'T allow is that any information given to it is actually true of the actual world. It is exactly that question the audit is asking, and cannot be answered in advance.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
If you imagine anyone within the legal system who actually has any authority over how it works is going to go down that route, I will show you an NHS union activist who knows how much expenditure on the NHS would be too much or a Home Secretary who knows what is the acceptable level of murder in our country.
Well I wouldn't expect it to be debated openly by (eg) the Home Secretary, nevertheless it's key.
"It is better that X guilty go free than one innocent is wrongly convicted."
This is central to the CJS and the Order of Magnitude (not the precise number obviously) of that X should imo be thought about and generally agreed. 10 or 100 or 1000 are the realistic alternatives and you get a very different system depending on which one it is.
But tbh I'm more just interested in what people on here think. Eg you. 10 or 100 or 1000 iyo? Which feels most right to you?
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
The default defense for a speed camera was it was not calibrated which is why many police forces now employ a senior admin person to keep all that data up to date in 1 place.
When the police were first allowed to use laser guns and cameras as primary evidence of a criminal offence, many forces had a somewhat laissez-faire attitude to keeping the equipment correctly stored and regularly calibrated, and the officers formally trained on their use, as was required by law.
It was only when defence lawyers (led by Nick Freeman) started questioning in court the reliability of such evidence, that the police forces actually paid attention to the issue. There will have been thousands of unsafe convictions for speeding, but most will have accepted the penalty without complaint or challenge.
Some interesting points but it is more law than politicalbetting really and quite a long article again
The article is about the serious failure of Parliamentary legislation. What do you think the purpose of governments should be, if not to legislate wisely ?
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
What the auditors were doing needs to be known by others as well as God. Any proper audit would discover the discrepancy between IT alleged money/transactions and money/transactions existing in the real world. If the discrepancy can't be/isn't disclosed then it isn't a proper audit because the one thing a proper audit CAN'T allow is that any information given to it is actually true of the actual world. It is exactly that question the audit is asking, and cannot be answered in advance.
I also don't understand (understatement) the attitude of the SPM's union. Far more interested, it would seem, in brown-nosing the PO Board than looking after the interests of it's members!
Another excellent piece on this scandal of scandals, as we’ve come to expect from our resident investigator. @Cyclefree you really should write the book.
"A Handbook of British Cock-Ups" with on its cover the quote by @Leon: "a decent writer with interesting opinions". 😁
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
The problem is that this should have been sorted out at the investigation stage. Rather than assume that there was theft the PO should have looked at why discrepancies were arising and got those resolved. It didn't. It simply did not have the necessary IT people who saw it as their job to resolve issues in software and get Horizon working properly and reliably. As any IT person will tell you, no software is free of bugs.
They knew it didn't work. When discrepancies arose they jumped straight into prosecute mode with no intervening stage of inquiry into what was happening with Horizon and why. It is only once you have done that then you can even begin to contemplate prosecution. And when you do, precisely because you have investigated and audited your own system, the presumption that the discrepancies can only have been caused by theft is stronger because you can show how the system worked.
What was missing here was any sort of proper internal IT audit and then proper investigation.
Put it this way, if if some issue like this came to me in my professional life, the very first thing I would do would be to spend bloody ages understanding the systems before jumping to any sort of presumption that there must have been theft. You work by a process of elimination - you eliminate all other possibilities. None of that happened and that is why the prosecutions were fundamentally flawed. The investigations were fundamentally flawed. Indeed I wouldn't even call them investigations in any meaningful sense of the word.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
Is not that number one an unanswerable case for such contracts simply being null and void?
A contract is not valid if signed with no sight, aiui.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
Trying to save money was probably one of the reasons behind the removal of S.69. But you make an important general point - one which I have tried to make repeatedly - persistent underfunding of the criminal justice system leads to injustice.
I think it's wrongly seen as overhead rather than (soft but crucial) infrastructure.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
If you imagine anyone within the legal system who actually has any authority over how it works is going to go down that route, I will show you an NHS union activist who knows how much expenditure on the NHS would be too much or a Home Secretary who knows what is the acceptable level of murder in our country.
Well I wouldn't expect it to be debated openly by (eg) the Home Secretary, nevertheless it's key.
"It is better that X guilty go free than one innocent is wrongly convicted."
This is central to the CJS and the Order of Magnitude (not the precise number obviously) of that X should imo be thought about and generally agreed. 10 or 100 or 1000 are the realistic alternatives and you get a very different system depending on which one it is.
But tbh I'm more just interested in what people on here think. Eg you. 10 or 100 or 1000 iyo? Which feels most right to you?
Very interesting. While I would not conjecture about how to quantify this in theory, there is one very significant practical problem: How do you know.
Our criminal system has these checks in it: police, CPS, prosecutor, defence, judge, jury, the concept of 'beyond reasonable doubt' and the Court of Appeal. It is to that structure that one must look for reliability, while infallibility is beyond all that and all of us.
At each stage prosecutions fail. From police saying 'NFA' to the CA quashing a conviction. What cannot be quantified is this: How many people convicted were actually innocent; nor can you quantify How many not convicted were actually guilty.
There are some cases where there is abundant and clear proof of innocence is someone who has been convicted. The PO scandal unusually has brought to light a particular stream of them, but this may be exceptional.
So while your question is interesting, the data is unobtainable.
hancock claims he told PM on 13th Mar that UK must lockdown.
KC says - odd, nothing in your book about that date.
Hmm....
This is the issue with the inquiry. Rather than the bun fight/blame game exercise we see now, it should have been a truth and reconciliation exercise. Too many of the participants are trying to protect their reputations at the expense of others, and there is a lot of benefit of hindsight going on. I don't blame anyone who had been subject of criticism already from trying to counter it, but the whole thing has been unedifying.
The BBC considerably more fawning, weirdly. Jim Naughtie positively fanboi-ing this am, though perhaps having just spent 2 days with HK recording an interview might have something to do with it. I bet along with many others he punched the air when he heard the news, though for different reasons.
Some interesting points but it is more law than politicalbetting really and quite a long article again
You're at liberty to skim through if you want to! I've no complaints about the length of any article that someone volunteers.
And we focus on so few subjects - Tories still in the 20s, 2024 presidential pre-pre-election second guessing - that introducing another hobbyhorse into our agenda is not unwelcome.
I wouldn't have heard about the Post Office scandal at all if it weren't for Cyclefree. I certainly wouldn't have understood it - I'd have just classified it as a highly regrettable error and moved on.
Personally I don't find the subject exciting - not unimportant, but unexciting. I understand why this is a big issue, but instinctively my brain wants to classify it as a 'computer says no' error. But I'm very pleased that someone DOES care about it. Because it clearly deserves more caring about than its getting.
I've liked, but just wanted to say (which of course I know you know) that the issue isn't about the 'computer says no error'. The computer error although triggering this is not the issue at all. It is all about what happened (or didn't happen) after that. Of course the computer shouldn't have got it wrong and those flaws need to be reviewed, but it is programmed by humans and these things will happen. What shouldn't have happened were the litany of consequences of that computer error.
I put my card in a Santander ATM and asked for £100. Nothing came out. So like an idiot I did it again and this time got the cash. When my Barclays statement arrived I had, indeed, been docked £200. After such a delay, and not a shred of evidence, I decided it wasn't worth the effort to make a fuss, and therefore became complicit. If disgruntled customers don't get on their case they'll just carry on carrying on, won't they?
My only recourse has been to give Santander a wide berth, which is not difficult as the very name to my ears sounds like a dodgy suburb of Torremolinos. But there again, bank corporate IDs are a mystery to the uninitiated. Chase Manhattan? Why on earth would you need to chase them? The question answers itself. And whatever prompts the Hong Kong and Shanghai Banking Corporation to hide behind a cypher? Above my pay grade.
An excellent and pungent piece, @Cyclefree - demonstrating the huge problems that can arise from individuals making supericial assumptions, that go unchecked.
It's the first day of real winter around here, so happy #plonkatonka day to everyone, when anyone who has spent £40k on an SUV, or £25k on a fake SUV, suddenly imagines they have received the snow-driving skills of Stig Blomqvist by osmosis.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
The default defense for a speed camera was it was not calibrated which is why many police forces now employ a senior admin person to keep all that data up to date in 1 place.
When the police were first allowed to use laser guns and cameras as primary evidence of a criminal offence, many forces had a somewhat laissez-faire attitude to keeping the equipment correctly stored and regularly calibrated, and the officers formally trained on their use, as was required by law.
I've got off with approximately 50% of all of the speeding infringements I've ever had and was 100% guilty in all cases.
Only a small minority (two, I think) of those were based on a successful challenge of equipment or training. All of the other were because the case fell apart because the prosecution's paperwork was wrong or late or otherwise fucked up or the relevant copper was a no show or on the sick or fuck knows what.
The whole system is based on getting an incriminating statement (How fast do you think you were going?) and is very poorly optimised for getting a prosecution on an uncooperative and belligerent defendant.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
Is not that number one an unanswerable case for such contracts simply being null and void?
A contract is not valid if signed with no sight, aiui.
Yes. Hence the Bates litigation which first cracked open this case. It is a very long judgment. The judge was deeply unimpressed - to put it mildly - by the PO witnesses. He found 25 different ways of describing them as liars - without actually using the word. It is a most impressive display of legal snark.
And what was the PO's reaction? To try and get the judge thrown off the case for bias because if they lost the cases it would be disastrous for the Post Office's commercial viability. (That is what they actually wrote internally.) Shades of Denning's infamous judgment in the Birmingham Six case.
That attempt (in which, shamefully, Lord Neuberger, a former Supreme Court judge was involved) was thrown out and described by Court of Appeal judges as wholly without merit.
But it shows the mindset of the PO and of the government - which was using our money to fund this nonsense.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
Yes, practicality does have to come into this. The CJS is creaking and under-resourced. It needs to be funded properly. If you make changes that lead to longer trials and more technical expertise needed by the Crown you have to be prepared to invest accordingly.
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
If you imagine anyone within the legal system who actually has any authority over how it works is going to go down that route, I will show you an NHS union activist who knows how much expenditure on the NHS would be too much or a Home Secretary who knows what is the acceptable level of murder in our country.
Well I wouldn't expect it to be debated openly by (eg) the Home Secretary, nevertheless it's key.
"It is better that X guilty go free than one innocent is wrongly convicted."
This is central to the CJS and the Order of Magnitude (not the precise number obviously) of that X should imo be thought about and generally agreed. 10 or 100 or 1000 are the realistic alternatives and you get a very different system depending on which one it is.
But tbh I'm more just interested in what people on here think. Eg you. 10 or 100 or 1000 iyo? Which feels most right to you?
Very interesting. While I would not conjecture about how to quantify this in theory, there is one very significant practical problem: How do you know.
Our criminal system has these checks in it: police, CPS, prosecutor, defence, judge, jury, the concept of 'beyond reasonable doubt' and the Court of Appeal. It is to that structure that one must look for reliability, while infallibility is beyond all that and all of us.
At each stage prosecutions fail. From police saying 'NFA' to the CA quashing a conviction. What cannot be quantified is this: How many people convicted were actually innocent; nor can you quantify How many not convicted were actually guilty.
There are some cases where there is abundant and clear proof of innocence is someone who has been convicted. The PO scandal unusually has brought to light a particular stream of them, but this may be exceptional.
So while your question is interesting, the data is unobtainable.
Yes very true. By definition it is. There'll be a way to roughly estimate what X is in our current system but not to measure it. But that doesn't matter for what I'm driving at.
(A) It is better that 10 guilty men go free than 1 innocent man be convicted. (B) It is better that 100 guilty men go free than 1 innocent man be convicted. (C) It is better that 1000 guilty men go free than 1 innocent man be convicted.
I'm wondering which feels most right to people. Just instinctively.
Just had an interesting experience (Well not particularly interesting) - Facebook sprang up an ad in my feed which was remarkably well targeted and something I could use. Hit the link, price was £13.99. Headed to Temu - same item being sold for £1.43 (With postage !!).
I've ordered said item anyway and the last item I bought from Aliexpress (Which is a similar explicitly direct from China site) was fine so I think this one ought to be.
One has to wonder though how much of the current Amazon/Facebook (Or other social media ads)/Ebay goods/shops are simply marked up resale from Temu/AliBaba/Aliexpress. & How many people check the friendly Chinese suppliers directly ?
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
I’m not sure that is practical.
Perhaps it’s just to downgrade the standing of IT evidence to corroboration. Ie - not sufficient to convict *in the absence of other evidence*
AIUI one of the mysteries in the PO scandal was “what happened to the cash”? The sub-postmasters weren’t living extravagant lives and their bank accounts weren’t inflated. The PO said “they are just skimming cash and hiding it somewhere - the computer says it’s missing and the computer is right”
If they had actually had to prove the first part of the proposition would that have provided sufficient protection?
The allegedly missing sums were parked in suspense accounts and then, AIUI, taken by the Post Office. Quite what their auditors were doing, God only knows. But remember - at the corporate level these sums were insignificant but very very material at a branch level. Horizon may well have worked fine at a corporate level but not at a branch level for what were separate businesses.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and 2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
Is not that number one an unanswerable case for such contracts simply being null and void?
A contract is not valid if signed with no sight, aiui.
Yes. Hence the Bates litigation which first cracked open this case. It is a very long judgment. The judge was deeply unimpressed - to put it mildly - by the PO witnesses. He found 25 different ways of describing them as liars - without actually using the word. It is a most impressive display of legal snark.
And what was the PO's reaction? To try and get the judge thrown off the case for bias because if they lost the cases it would be disastrous for the Post Office's commercial viability. (That is what they actually wrote internally.) Shades of Denning's infamous judgment in the Birmingham Six case.
That attempt (in which, shamefully, Lord Neuberger, a former Supreme Court judge was involved) was thrown out and described by Court of Appeal judges as wholly without merit.
But it shows the mindset of the PO and of the government - which was using our money to fund this nonsense.
I repeat; what on earth were the SPM's Union doing? Can't imagine the RMT (for example) putting up with this sort of nonsense.
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
hancock claims he told PM on 13th Mar that UK must lockdown.
KC says - odd, nothing in your book about that date.
Hmm....
This is the issue with the inquiry. Rather than the bun fight/blame game exercise we see now, it should have been a truth and reconciliation exercise. Too many of the participants are trying to protect their reputations at the expense of others, and there is a lot of benefit of hindsight going on. I don't blame anyone who had been subject of criticism already from trying to counter it, but the whole thing has been unedifying.
Good morning
Frankly the whole enquiry is a massive 'gotcha' exercise when it should be a forensic investigation into lessons to be learnt
Furthermore, hindsight is a wonderful thing but those involved were dealing with a once in a generation pandemic which nobody had had any experience off
I give everyone involved, including Johnson, sympathy for the decisions they made of life and death with huge unknowns
hancock claims he told PM on 13th Mar that UK must lockdown.
KC says - odd, nothing in your book about that date.
Hmm....
This is the issue with the inquiry. Rather than the bun fight/blame game exercise we see now, it should have been a truth and reconciliation exercise. Too many of the participants are trying to protect their reputations at the expense of others, and there is a lot of benefit of hindsight going on. I don't blame anyone who had been subject of criticism already from trying to counter it, but the whole thing has been unedifying.
Good morning
Frankly the whole enquiry is a massive 'gotcha' exercise when it should be a forensic investigation into lessons to be learnt
Furthermore, hindsight is a wonderful thing but those involved were dealing with a once in a generation pandemic which nobody had had any experience off
I give everyone involved, including Johnson, sympathy for the decisions they made of life and death with huge unknowns
"A Handbook of British Cock-Ups" with on its cover the quote by @Leon: "a decent writer with interesting opinions". 😁
It'll catch fire, trust me.
The PO affair is an utter scandal, the absolute pits, but I wanted to ask you a more general question:
Say S69 comes back and there's a case where a vital piece of evidence for the prosecution is some output from a big company's IT system. So the Crown now have to 'prove' this piece of output is ok.
Can they rely on the Auditors having signed off on the key systems and controls as part of the statutory audit? I mean, would that normally be enough or does the legal process tend to need something more specific and rigorous?
Talking of crime. Yesterday I was cycling back from lunch along Piccadilly - traffic was solid - when a guy on one of those electric food delivery-type bikes swooped in front of me between the cars and over to the pavement, and in one motion plucked a mobile phone from the hands of a tourist outside the Royal Academy and then off he went at speed.
I had to admire his dexterity.
I then cycled on and came upsides a police ERV and tapped on the window as I thought they might be vaguely interested. They completely ignored me. Some bloke tapping on the window of a police vehicle and they just looked at me and drove off.
hancock claims he told PM on 13th Mar that UK must lockdown.
KC says - odd, nothing in your book about that date.
Hmm....
This is the issue with the inquiry. Rather than the bun fight/blame game exercise we see now, it should have been a truth and reconciliation exercise. Too many of the participants are trying to protect their reputations at the expense of others, and there is a lot of benefit of hindsight going on. I don't blame anyone who had been subject of criticism already from trying to counter it, but the whole thing has been unedifying.
Good morning
Frankly the whole enquiry is a massive 'gotcha' exercise when it should be a forensic investigation into lessons to be learnt
Furthermore, hindsight is a wonderful thing but those involved were dealing with a once in a generation pandemic which nobody had had any experience off
I give everyone involved, including Johnson, sympathy for the decisions they made of life and death with huge unknowns
Hope all is well with you Big_G.
Thank you
I have some way to go with three different consultants but remain positive and keep taking my pills
"A Handbook of British Cock-Ups" with on its cover the quote by @Leon: "a decent writer with interesting opinions". 😁
It'll catch fire, trust me.
The PO affair is an utter scandal, the absolute pits, but I wanted to ask you a more general question:
Say S69 comes back and there's a case where a vital piece of evidence for the prosecution is some output from a big company's IT system. So the Crown now have to 'prove' this piece of output is ok.
Can they rely on the Auditors having signed off on the key systems and controls as part of the statutory audit? I mean, would that normally be enough or does the legal process tend to need something more specific and rigorous?
The answer is no. You need to understand what the auditors are signing off on. What are they testing and what is the margin of error they are happy to accept? If they have tested the bit of the system which produces the data being relied on that might be fine.
But remember a system may be working fine but the data may still be unreliable eg if what has been inputted is wrong or if the assumptions underlying it are wrong etc.,. That is why you need to understand the whole process. And you generally need to dig deeper than a general sign off on key systems and controls.
Some interesting points but it is more law than politicalbetting really and quite a long article again
The article is about the serious failure of Parliamentary legislation. What do you think the purpose of governments should be, if not to legislate wisely ?
Yes but this website is called 'Politicalbetting.Com' not 'In depth scrutiny of Parliamentary legislation.com'.
Now we have some posts on other topics eg sports betting, Eurovision winner and occasionally broader political topics but this is quite a loose expansion of the heading of the site. Which is fine but that is what it is
He's right. I listened to extracts from yesterday's HO Select Committee hearing. I normally defend Civil Servants, but the two who appeared yesterday (Rycroft + another) were embarrassingly hopeless - ill-prepared and incapable of answering basic questions. This led to the bizarre scenario of both the Chair (Diane Johnson, Lab) and Lee Anderson joining forces to tell them how useless they were.
It would appear that in their desperation to meet targets for removing the backlog, at least 17,000 asylum seekers have simply had their claims discontinued. When asked where they were, the response was simply "no idea".
I'm not sure on that one.
My MP Lee Anderson came across to me as a grandstanding barrack-room-lawyer - asking statistical questions based on multiple sub-categories, then having a go at Civil Servants for not having the answers immediately available in their heads.
Perhaps if he had given appropriate notice (eg 48-72 hours) of his complex questions Civil Servants would have been in a position to provide him with an answer.
I am not convinced that he actually wants answers, rather an opportunity to indulge in performative outrage on GB News / social media.
I'm open to correction by anyone with a more complete knowledge, but that is my impression from coverage I have seen.
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
Are we going through this nonsense of everybody is off to Mastadon again...hows Threads doing these days, seems to have about as much popularity as Piers Morgan tv show.
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
The North won by a full scale invasion of the South, not by a popular uprising. Using a massive army, with thousands of Soviet supplied tanks and armoured vehicles.
The South fought until they ran out of ammunition. Resupply of which was denied by the US.
He's right. I listened to extracts from yesterday's HO Select Committee hearing. I normally defend Civil Servants, but the two who appeared yesterday (Rycroft + another) were embarrassingly hopeless - ill-prepared and incapable of answering basic questions. This led to the bizarre scenario of both the Chair (Diane Johnson, Lab) and Lee Anderson joining forces to tell them how useless they were.
It would appear that in their desperation to meet targets for removing the backlog, at least 17,000 asylum seekers have simply had their claims discontinued. When asked where they were, the response was simply "no idea".
I'm not sure on that one.
My MP Lee Anderson came across to me as a grandstanding barrack-room-lawyer - asking statistical questions based on multiple sub-categories, then having a go at Civil Servants for not having the answers immediately available in their heads.
Perhaps if he had given appropriate notice (eg 48-72 hours) of his complex questions Civil Servants would have been in a position to provide him with an answer.
I am not convinced that he actually wants answers, rather an opportunity to indulge in performative outrage on GB News / social media.
I'm open to correction by anyone with a more complete knowledge, but that is my impression from coverage I have seen.
@MattW. Quick follow-up: how is your chemo bearing up?
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
Not in the South, South Vietnam might be a prosperous, democratic version of South Korea if Congress hadn't give up and let North Vietnam capture Saigon
"A Handbook of British Cock-Ups" with on its cover the quote by @Leon: "a decent writer with interesting opinions". 😁
It'll catch fire, trust me.
The PO affair is an utter scandal, the absolute pits, but I wanted to ask you a more general question:
Say S69 comes back and there's a case where a vital piece of evidence for the prosecution is some output from a big company's IT system. So the Crown now have to 'prove' this piece of output is ok.
Can they rely on the Auditors having signed off on the key systems and controls as part of the statutory audit? I mean, would that normally be enough or does the legal process tend to need something more specific and rigorous?
The answer is no. You need to understand what the auditors are signing off on. What are they testing and what is the margin of error they are happy to accept? If they have tested the bit of the system which produces the data being relied on that might be fine.
But remember a system may be working fine but the data may still be unreliable eg if what has been inputted is wrong or if the assumptions underlying it are wrong etc.,. That is why you need to understand the whole process. And you generally need to dig deeper than a general sign off on key systems and controls.
Right. That makes sense given what I remember about what the Audit actually does - or more to the point doesn't do.
That's quite a burden on the Crown then, potentially. They'd need to be equal to it, as regards resource and expertise.
Are we going through this nonsense of everybody is off to Mastadon again...hows Threads doing these days, seems to have about as much popularity as Piers Morgan tv show.
Good point. Although I think Truth Social/Telegram provide the same functionality, unlike Bluesky/Mastodon. Neverthless a lot of people who leave Twitter come back. Zeihan's case is slightly different because he uses it to research and disseminate his stuff: the former is compromised by Muskery and the latter is not necessary
Enshittification is followed by high-value-users buggering off and founding their own communities. YouTube refuseniks are going to Nebula or Armchair History TV or that new gun one. Trump Twitniks went to Truth Social, Russians(?) went to Telegram. Threads/Mastodon didn't get traction because they were difficult to use(?). WhatsApp is its own thing.
We used to have megathings and now we don't: USENET went to newsgroups went to Geocities went to MySpace went to Facebook and fragmented to YouTube, Twitter, WhatsApp, Nebula, Truth Social, stuff, stuff, more stuff. Each year things become more difficult to analyse and summarise.
Are we going through this nonsense of everybody is off to Mastadon again...hows Threads doing these days, seems to have about as much popularity as Piers Morgan tv show.
Good point. Although I think Truth Social/Telegram provide the same functionality, unlike Bluesky/Mastodon. Neverthless a lot of people who leave Twitter come back. Zeihan's case is slightly different because he uses it to research and disseminate his stuff: the former is compromised by Muskery and the latter is not necessary
Enshittification is followed by high-value-users buggering off and founding their own communities. YouTube refuseniks are going to Nebula or Armchair History TV or that new gun one. Trump Twitniks went to Truth Social, Russians(?) went to Telegram. Threads/Mastodon didn't get traction because they were difficult to use(?). WhatsApp is its own thing.
We used to have megathings and now we don't: USENET went to newsgroups went to Geocities went to MySpace went to Facebook and fragmented to YouTube, Twitter, WhatsApp, Nebula, Truth Social, stuff, stuff, more stuff. Each year things become more difficult to analyse and summarise.
Most of the people on services like Nebula are on YouTube and use their YouTube videos to drive traffic to Nebula.
Thanks to Cyclefree for another challenging piece. As someone responded on a previous thread, there is a general problem in where to put the burden of proof when complex IT is involved. Clearly it's unreasonable to expect a subpostmaster to get to the bottom of a bug in the central software. At the same time, it's impossible to prove a negative, so requiring that the prosecution to prove that the software is bug-free really is 'impractical'.
Where something minor like a speeding fine is involved, I'd be inclined to trust the system, and wouldn't expect a court to assume that every speed camera is faulty unless proved sound. Where criminal prosecution is being undertaken, clearly a higher standard is needed. What would a fair system involve? Perhaps an *independent* IT analysis of whether there is a way that the apparent fraud could have been caused by a bug in the system?
The default defense for a speed camera was it was not calibrated which is why many police forces now employ a senior admin person to keep all that data up to date in 1 place.
When the police were first allowed to use laser guns and cameras as primary evidence of a criminal offence, many forces had a somewhat laissez-faire attitude to keeping the equipment correctly stored and regularly calibrated, and the officers formally trained on their use, as was required by law.
I've got off with approximately 50% of all of the speeding infringements I've ever had and was 100% guilty in all cases.
Only a small minority (two, I think) of those were based on a successful challenge of equipment or training. All of the other were because the case fell apart because the prosecution's paperwork was wrong or late or otherwise fucked up or the relevant copper was a no show or on the sick or fuck knows what.
The whole system is based on getting an incriminating statement (How fast do you think you were going?) and is very poorly optimised for getting a prosecution on an uncooperative and belligerent defendant.
Interesting stat.
In the other direction, witnesses who submit cam footage of people breaking motoring law via police portals can achieve success rates of towards 80%.
Or how "awkward" it is that South Africa is performing badly without white supremacy rule. Still some mileage there, I think.
I think I missed that one.
He has to be watched. If untrammelled it'd become Reactionaries Corner on here, all pink gins and revolving bow ties. As it is, we just about get a balance, but there's no room for complacency. You have to fight fight fight.
Are we going through this nonsense of everybody is off to Mastadon again...hows Threads doing these days, seems to have about as much popularity as Piers Morgan tv show.
Good point. Although I think Truth Social/Telegram provide the same functionality, unlike Bluesky/Mastodon. Neverthless a lot of people who leave Twitter come back. Zeihan's case is slightly different because he uses it to research and disseminate his stuff: the former is compromised by Muskery and the latter is not necessary
Enshittification is followed by high-value-users buggering off and founding their own communities. YouTube refuseniks are going to Nebula or Armchair History TV or that new gun one. Trump Twitniks went to Truth Social, Russians(?) went to Telegram. Threads/Mastodon didn't get traction because they were difficult to use(?). WhatsApp is its own thing.
We used to have megathings and now we don't: USENET went to newsgroups went to Geocities went to MySpace went to Facebook and fragmented to YouTube, Twitter, WhatsApp, Nebula, Truth Social, stuff, stuff, more stuff. Each year things become more difficult to analyse and summarise.
Something will replace Twitter, and will look like Twitter because that's the model that works. Either that or Musk will sell it, for one reason or another and its new owners will restore its previous form.
That 'something' hasn't been created yet and it may be some time before it is. Nonetheless, it's sufficiently broken that when a workable alternative arrives, X will mark the spot at which Twitter is buried.
Are we going through this nonsense of everybody is off to Mastadon again...hows Threads doing these days, seems to have about as much popularity as Piers Morgan tv show.
Good point. Although I think Truth Social/Telegram provide the same functionality, unlike Bluesky/Mastodon. Neverthless a lot of people who leave Twitter come back. Zeihan's case is slightly different because he uses it to research and disseminate his stuff: the former is compromised by Muskery and the latter is not necessary
Enshittification is followed by high-value-users buggering off and founding their own communities. YouTube refuseniks are going to Nebula or Armchair History TV or that new gun one. Trump Twitniks went to Truth Social, Russians(?) went to Telegram. Threads/Mastodon didn't get traction because they were difficult to use(?). WhatsApp is its own thing.
We used to have megathings and now we don't: USENET went to newsgroups went to Geocities went to MySpace went to Facebook and fragmented to YouTube, Twitter, WhatsApp, Nebula, Truth Social, stuff, stuff, more stuff. Each year things become more difficult to analyse and summarise.
Something will replace Twitter, and will look like Twitter because that's the model that works. Either that or Musk will sell it, for one reason or another and its new owners will restore its previous form.
That 'something' hasn't been created yet and it may be some time before it is. Nonetheless, it's sufficiently broken that when a workable alternative arrives, X will mark the spot at which Twitter is buried.
Bluesky at some point soon will be making posts visible to non members. At that point it may attract the next bit of twitter’s audience.
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
I'm not disputing the reasons *why* each war turned out differently, simply making the point that the morality of waging a war is so often judged by its success or failure.
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
Not in the South, South Vietnam might be a prosperous, democratic version of South Korea if Congress hadn't give up and let North Vietnam capture Saigon
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
Not in the South, South Vietnam might be a prosperous, democratic version of South Korea if Congress hadn't give up and let North Vietnam capture Saigon
Ever been there?
Vietnam has pretty well abandoned communism now, to the point that it's almost a US ally.
Jarnail Singh giving evidence to the Post Office Horizon IT inquiry. He's almost incoherent. It's a car crash, which we can all watch happening in real time here:
In terms of war, if it succeeds, it’s viewed as being morally justified. If it fails, it’s viewed as morally unjustified.
Had South Vietnam turned out like South Korea (and Rhee’s regime was considerably nastier than the South Vietnamese), the war would be viewed in the same light as the Korean War.
The USA is neither The Shining City on the Hill, nor The Great Satan. But, I sleep safely in my bed, due in no small part to US firepower, and for that, I am grateful.
Vietnam was never going to "turn out like South Korea". We've rehearsed why the two conflicts were entirely different quite recently, and at some length.
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Had Congress not withdrawn funding for the war effort, as the Democrat Congress did, it is possible the war would have ended in stalemate like Korea did rather than North Korea capturing the South
The North Vietnamese were 'winning' though. Much more popular support than the increasingly corrupt Southern government.
Not in the South, South Vietnam might be a prosperous, democratic version of South Korea if Congress hadn't give up and let North Vietnam capture Saigon
Ever been there?
Vietnam has pretty well abandoned communism now, to the point that it's almost a US ally.
Never seen so many people on motor-cycles (of various types) as in Ho Chi Minh City.
Listening to Jarnail Singh try and give an account of himself to Jason Beer felt rather like watching Baldrick explain a cunning plan to Edmund Blackadder. #PostOfficeScandal"
The only Christmas song I can stand, indeed quite like, is "Wombling Merry Christmas" by The Wombles. I think it might be drowned out by one of my least favourites this year.
Comments
My earlier comment was in jest.
https://www.lbc.co.uk/news/security-minister-home-office-missing-asylum-seekers/
Kissinger was also the last of the 'Rockefeller Republicans' and of course worked for President Ford too whose VP was Nelson Rockefeller.
If he was starting out in politics now Kissinger would probably be a Clinton Democrat rather than Republican, indeed he said Hillary was one of the best US Secretaries of State. He was too globalist for the Trumpite wing of the GOP which now dominates it
https://www.thenation.com/article/archive/henry-kissinger-hillary-clintons-tutor-in-war-and-peace/
If I give you £2, you have £2. If you computer says I gave you £20, you still have the correct £2, even if your boss’s lawyer looks at the computer and says you’re £18 short.
The other 2 points to remember are that:
1. the contracts with the PO were very one-sided and not given to SPM until after they'd signed up; and
2. The PO did not have the right to get information from Fujitsu and had to pay over a certain level so it was not even in control of its own system. It had to say that it was working fine because it had no way of checking this for itself and to admit this would have made it look stupid. Saving face is what mattered to it.
I daren't pick up a book on Trump.
It is wholly unfeasible to prove your case if you cannot rely upon the infallible nature of automated systems in a manner which admits vast amounts of hearsay evidence.
In finding a solution, the only one I can think of is to give the judge a guided discretion as to when to require strict proof of evidence X with equal access for both sides to expert evidence.
The PO case would be an obvious example, where (SFAICS) there was generally no corroboration, and the reliability of IT was effectively 100% of the case.
Of course as well as protecting the innocent, such a move would provide a field day for the guilty seeking to complexify basically simple evidence, like phone calls and messaging.
Simple stuff really. This wouldn't have got past a vaguely interested graduate internal auditor if they had access to all the data. I assume this is where the PO's prosecutors took over.
Like some others, I confess the PO scandal would barely be on my radar if not for your headers. Thank you.
Why are the news outlets not making more of it? Could it be that the arts graduates who run them are struggling to understand because it's, you know, techy stuff?
It also speaks (along with the 'reasonable doubt' rule) to the chestnut "it is better that X guilty men go free than one innocent man be wrongly convicted".
I put X since it's not about a precise number but the sentiment - that some of the guilty getting off is a price worth paying to minimise the chances of the innocent suffering miscarriages of justice. It's surely correct to have this sentiment baked into the system.
However the number (approx not precise) is important because neither extreme works. You don't want huge numbers of the guilty escaping justice in order to achieve zero risk of convicting the innocent. Neither do you want all of the guilty being convicted along with plenty of the innocent.
So even though it can only be estimated rather than measured at least the order of magnitude of X should be overt and a matter of consensus imo. Is it 10 or 100 or 1000? I'd be interested to hear which of those people think is about right. I bet there's quite a variation of opinion.
This post comes from a STEM type person rather than a waffly arts/humanities type obviously.
I agree with all that. But it’s necessary not sufficient for a criminal conviction in my view
And each system audits individually so you can check that the audit details matches across systems
But that said, like crypto, blockchain and various other tech fads it is also a magnet for Tate-osphere nonsense. And unlike e.g. NFTs, which I think to all clear-thinking folk was patently a scam, it is actually quite difficult for the inexpert to sort through the chaff.
With that in mind, I did find something genuinely useful from my own field (marketing & advertising), using LLM to build perceptual maps of perceptions of car brands. For an academic article, it's a fairly lucid read and quite an interesting case study of the sort of things LLM will soon be able to do for us, and will no doubt have equivalent examples in lots of other fields.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4241291
Excellent thread header @Cyclefree. Very thought provoking, although I do find the whole thing depressing. This stuff just repeats itself over and over again. Contrary to the statement always made, lessons are never learned.
It would appear that in their desperation to meet targets for removing the backlog, at least 17,000 asylum seekers have simply had their claims discontinued. When asked where they were, the response was simply "no idea".
https://twitter.com/PostOffInquiry
Mr. Ghedebrav, except for money laundering, where they're very effective, I remain astonished that anyone ever thought NFTs were a good idea.
I remember the case of a prisoner released 20 years after a wrongful conviction due to a botched lab test. Asked by a journalist eager for a scathing quote what he felt about the lab technician responsible, the ex-prisoner said mildly, "Well, we all make mistakes." The very best tradition of British understatement.
https://www.youtube.com/watch?v=YBK4g94PA5A
"It is better that X guilty go free than one innocent is wrongly convicted."
This is central to the CJS and the Order of Magnitude (not the precise number obviously) of that X should imo be thought about and generally agreed. 10 or 100 or 1000 are the realistic alternatives and you get a very different system depending on which one it is.
But tbh I'm more just interested in what people on here think. Eg you. 10 or 100 or 1000 iyo? Which feels most right to you?
It was only when defence lawyers (led by Nick Freeman) started questioning in court the reliability of such evidence, that the police forces actually paid attention to the issue. There will have been thousands of unsafe convictions for speeding, but most will have accepted the penalty without complaint or challenge.
What do you think the purpose of governments should be, if not to legislate wisely ?
😁
The problem is that this should have been sorted out at the investigation stage. Rather than assume that there was theft the PO should have looked at why discrepancies were arising and got those resolved. It didn't. It simply did not have the necessary IT people who saw it as their job to resolve issues in software and get Horizon working properly and reliably. As any IT person will tell you, no software is free of bugs.
They knew it didn't work. When discrepancies arose they jumped straight into prosecute mode with no intervening stage of inquiry into what was happening with Horizon and why. It is only once you have done that then you can even begin to contemplate prosecution. And when you do, precisely because you have investigated and audited your own system, the presumption that the discrepancies can only have been caused by theft is stronger because you can show how the system worked.
What was missing here was any sort of proper internal IT audit and then proper investigation.
Put it this way, if if some issue like this came to me in my professional life, the very first thing I would do would be to spend bloody ages understanding the systems before jumping to any sort of presumption that there must have been theft. You work by a process of elimination - you eliminate all other possibilities. None of that happened and that is why the prosecutions were fundamentally flawed. The investigations were fundamentally flawed. Indeed I wouldn't even call them investigations in any meaningful sense of the word.
A contract is not valid if signed with no sight, aiui.
KC says - odd, nothing in your book about that date.
Hmm....
Our criminal system has these checks in it: police, CPS, prosecutor, defence, judge, jury, the concept of 'beyond reasonable doubt' and the Court of Appeal. It is to that structure that one must look for reliability, while infallibility is beyond all that and all of us.
At each stage prosecutions fail. From police saying 'NFA' to the CA quashing a conviction. What cannot be quantified is this: How many people convicted were actually innocent; nor can you quantify How many not convicted were actually guilty.
There are some cases where there is abundant and clear proof of innocence is someone who has been convicted. The PO scandal unusually has brought to light a particular stream of them, but this may be exceptional.
So while your question is interesting, the data is unobtainable.
I don't blame anyone who had been subject of criticism already from trying to counter it, but the whole thing has been unedifying.
My only recourse has been to give Santander a wide berth, which is not difficult as the very name to my ears sounds like a dodgy suburb of Torremolinos. But there again, bank corporate IDs are a mystery to the uninitiated. Chase Manhattan? Why on earth would you need to chase them? The question answers itself. And whatever prompts the Hong Kong and Shanghai Banking Corporation to hide behind a cypher? Above my pay grade.
An excellent and pungent piece, @Cyclefree - demonstrating the huge problems that can arise from individuals making supericial assumptions, that go unchecked.
It's the first day of real winter around here, so happy #plonkatonka day to everyone, when anyone who has spent £40k on an SUV, or £25k on a fake SUV, suddenly imagines they have received the snow-driving skills of Stig Blomqvist by osmosis.
Remember Gloucestershire last year:
https://youtu.be/QPBsVEvpkk0?t=81
Only a small minority (two, I think) of those were based on a successful challenge of equipment or training. All of the other were because the case fell apart because the prosecution's paperwork was wrong or late or otherwise fucked up or the relevant copper was a no show or on the sick or fuck knows what.
The whole system is based on getting an incriminating statement (How fast do you think you were going?) and is very poorly optimised for getting a prosecution on an uncooperative and belligerent defendant.
And what was the PO's reaction? To try and get the judge thrown off the case for bias because if they lost the cases it would be disastrous for the Post Office's commercial viability. (That is what they actually wrote internally.) Shades of Denning's infamous judgment in the Birmingham Six case.
That attempt (in which, shamefully, Lord Neuberger, a former Supreme Court judge was involved) was thrown out and described by Court of Appeal judges as wholly without merit.
But it shows the mindset of the PO and of the government - which was using our money to fund this nonsense.
(A) It is better that 10 guilty men go free than 1 innocent man be convicted.
(B) It is better that 100 guilty men go free than 1 innocent man be convicted.
(C) It is better that 1000 guilty men go free than 1 innocent man be convicted.
I'm wondering which feels most right to people. Just instinctively.
I've ordered said item anyway and the last item I bought from Aliexpress (Which is a similar explicitly direct from China site) was fine so I think this one ought to be.
One has to wonder though how much of the current Amazon/Facebook (Or other social media ads)/Ebay goods/shops are simply marked up resale from Temu/AliBaba/Aliexpress. & How many people check the friendly Chinese suppliers directly ?
Whether the US was mistaken to get involved in the first pace is a matter of legitimate historical debate - but by the time of the '68 election, the war was unwinnable (absent a full scale US military invasion of the north, and possible nuclear conflict) - which was also the Pentagon analysis at the time. And that was nothing to do with the relative nastiness of the regimes in either north or south.
Kissinger's and Nixon's actions to prolong it were almost entirely for domestic political reasons, and Kissinger's post Vietnam justifications were the purest bullshit.
Frankly the whole enquiry is a massive 'gotcha' exercise when it should be a forensic investigation into lessons to be learnt
Furthermore, hindsight is a wonderful thing but those involved were dealing with a once in a generation pandemic which nobody had had any experience off
I give everyone involved, including Johnson, sympathy for the decisions they made of life and death with huge unknowns
The PO affair is an utter scandal, the absolute pits, but I wanted to ask you a more general question:
Say S69 comes back and there's a case where a vital piece of evidence for the prosecution is some output from a big company's IT system. So the Crown now have to 'prove' this piece of output is ok.
Can they rely on the Auditors having signed off on the key systems and controls as part of the statutory audit? I mean, would that normally be enough or does the legal process tend to need something more specific and rigorous?
I had to admire his dexterity.
I then cycled on and came upsides a police ERV and tapped on the window as I thought they might be vaguely interested. They completely ignored me. Some bloke tapping on the window of a police vehicle and they just looked at me and drove off.
I have some way to go with three different consultants but remain positive and keep taking my pills
But remember a system may be working fine but the data may still be unreliable eg if what has been inputted is wrong or if the assumptions underlying it are wrong etc.,. That is why you need to understand the whole process. And you generally need to dig deeper than a general sign off on key systems and controls.
Now we have some posts on other topics eg sports betting, Eurovision winner and occasionally broader political topics but this is quite a loose expansion of the heading of the site. Which is fine but that is what it is
My MP Lee Anderson came across to me as a grandstanding barrack-room-lawyer - asking statistical questions based on multiple sub-categories, then having a go at Civil Servants for not having the answers immediately available in their heads.
Perhaps if he had given appropriate notice (eg 48-72 hours) of his complex questions Civil Servants would have been in a position to provide him with an answer.
I am not convinced that he actually wants answers, rather an opportunity to indulge in performative outrage on GB News / social media.
I'm open to correction by anyone with a more complete knowledge, but that is my impression from coverage I have seen.
"Why I'm Done With Twitter (or 'X' or whatever you call it)", Peter Zeihan, see https://www.youtube.com/watch?v=0kwtVbM5o-o
The South fought until they ran out of ammunition. Resupply of which was denied by the US.
https://globalnews.ca/news/10116838/life-expectancy-decrease-canada-stat-can/
That's quite a burden on the Crown then, potentially. They'd need to be equal to it, as regards resource and expertise.
Enshittification is followed by high-value-users buggering off and founding their own communities. YouTube refuseniks are going to Nebula or Armchair History TV or that new gun one. Trump Twitniks went to Truth Social, Russians(?) went to Telegram. Threads/Mastodon didn't get traction because they were difficult to use(?). WhatsApp is its own thing.
We used to have megathings and now we don't: USENET went to newsgroups went to Geocities went to MySpace went to Facebook and fragmented to YouTube, Twitter, WhatsApp, Nebula, Truth Social, stuff, stuff, more stuff. Each year things become more difficult to analyse and summarise.
In the other direction, witnesses who submit cam footage of people breaking motoring law via police portals can achieve success rates of towards 80%.
Don't do drink and drugs kids.
That 'something' hasn't been created yet and it may be some time before it is. Nonetheless, it's sufficiently broken that when a workable alternative arrives, X will mark the spot at which Twitter is buried.
Mariah Carey or Wham most likely to be Christmas no 1.
@nickwallis
Jarnail Singh giving evidence to the Post Office Horizon IT inquiry. He's almost incoherent. It's a car crash, which we can all watch happening in real time here:
https://youtube.com/watch?v=JoSsalkkW98
#PostOfficeScandal #PostOfficeInquiry"
https://twitter.com/nickwallis/status/1730191835933393349
@Tucker5law
Listening to Jarnail Singh try and give an account of himself to Jason Beer felt rather like watching Baldrick explain a cunning plan to Edmund Blackadder. #PostOfficeScandal"
https://twitter.com/Tucker5law/status/1730196039934607561