
Don't forget NPfIT
Did we ever find out how much Fujitsu were paid when they lost the contract?
Since the broadcast of a television drama telling the story of the Post Office Horizon scandal — one of the most serious miscarriages of justice in British legal history — calls have mounted for those responsible to be held to account. Leaping into focus is Fujitsu, the supplier of a controversial £2.3 billion ($2.94 billion) …
Part of that, so I was told by the people in it, was because once the contract was in place, within the NHS each manager wanted more added, this added, that added, this added Fujtsu said "OK, pay us" they said no, and eventually decided they wanted out of the contract, at which point Fujitsu sued.
But that could all be inaccurate as we see how Fujitsu have behaved. Maybe the film Gung Ho was accurate after all.
In fairness, having done IT work in healthcare, I've never experienced a lack of control like it. I (genuinely) do get that the frontline staff are overworked, over-houred (?) and underpaid. Every NHS IT program however is doomed to fail as it quickly has to become all things to all people (wouldn't it be great if it could handle this process too), and there's no control over process alignment.
One notable implementation was a nightmare for me as two different teams (same local organisation) had been running the same process with the same input and output in two completely different ways. Both local managers were adamant that their process was unchangeable and that they couldn't possibly meet in the middle, often falling back to a "too busy to change" mentality. When escalated, this fell on deaf ears with a "the managers are right" response and the program nearly collapsed as the system very nearly couldn't be configured to run in two different ways at the same time. If this is just one tiny subset in one single organisation, when you face the considerable numbers of organisations that come together to form the NHS and that they all will not adapt their processes, it's nigh on impossible to implement a blanket platform.
Of course, those of us that have played this game for long enough know that the real pain is yet to come. When you've taken a largely off-the-peg solution and customised it to kingdom come, there is no simple update path, no option to upgrade without rewriting a lot of the customisation. Of course, even if you can, that leaves a platform that needs testing, and those that are "too busy to change" are certainly too busy to take part in testing. Those that are unwilling to leverage standardised processes are unwilling to enforce participation in a testing program.
I reiterate the love for the NHS, the people of it and notably those on the front line. However, someone who's been there recently and knows how it goes on the front line needs to get a proper hold of "what should be", and those above need to put their adult trousers on to enforce that all NHS processes are handled in one way, from the way the phone queues work, through the patient process (and staff workflow) for a specific diagnosis-procedure-aftercare lifecycle, to record keeping/interaction/notification and to the manner services are accessed. Until we stop this "but Jane needs to be notified **with a green banner** because she prints the paper letters when she's in on Wednesday because that's how we've always done it in our building and we won't rely on email even though everyone else does" nonsense, every single initiative is destined for immediate catastrophic overspend and ultimate failure.
Great final comment..."Until we stop this "but Jane needs to be notified **with a green banner** because she prints the paper letters when she's in on Wednesday because that's how we've always done it in our building and we won't rely on email even though everyone else does" nonsense, every single initiative is destined for immediate catastrophic overspend and ultimate failure." so true...
And when they remove jobs worth managers and heads of service from roles, especially in NHS IT or hire ones that listen to their staff.
We were looking to move to new laptops, away from Dell so HP was suggested. One Trust disagreed but was ignored and rightfully pissed off. When the execs came back from silicon valley in the old US of A paid for by, you guessed it, HP and then said "Look at these nice HP laptop tablets they've given us" it was obvious what was going on. Especially when we pointed out "You can't accept those laptops they are classed as a bribe" to which they replied they weren't and kept them. The HP contract was then signed.
We then had our director of IT, the biggest, corrupt cunt ever, get a contract with a provider for remote access for Windows XP (this was all many moons ago) and do a "testimony" review for them and their site. The office he was in, in the video wasn't his office. It was all staged at their gaff as their office was nicer. How much did he get for that deal or what discount did he get?
The amount of IT managers was/is in the NHS is such a waste. I even asked why don't we get our own HDD crusher instead of out sourcing. No was the answer. Then they got hit with a record fine when a few years later some drives sent for destruction where actually sold off. After that they got the crusher and were religious it be done properly. I was long gone then but funny how they now ran with my idea.
It was such a toxic environment to work in because of those IT managers and director (and it should never be this way at work) that I heard a few years back that that director had got cancer and had to take a year out. He was such an evil cunt that I blurted out "Shame he didn't die from it". No place of work should ever make you feel that way.
It doesn't take long to get "always done things this way" attitudes
In a public library the IT champion and I rolled out MSIE (yes, at the time it was better than Netscape). When Netscape improved beyond recognition a year later, we changed the default browser - only to have the librarians all held their hands up in horror and said they couldn't use it because it wasn't MSIE
It's a fricking browser FFS, with a nearly identical interface - and librarians are supposed to think critically (it's part of the job, filing books is for the assistants - who all embraced the browser change)
"the program nearly collapsed as the system very nearly couldn't be configured to run in two different ways at the same time."
Had that in spades. A shouting match in the general office between two directors, not about how two different teams should do things but how the same team should do the one thing. We offered to make everything they were arguing about configurable from the operator interface. That mollified them both so we implemented that.
Then we set up some reasonable-looking defaults for production which, ASFAIK were never touched thereafter.
Sometimes Just Working really is what's important.
uuuuuuuuuuuurrrrrrrggggggghhhhhhhhhhhhhhh. I have had the misfortune to work supporting a NHS CCG a few years ago, nightmare! And its not just IT. The wife works for a business who provide CPAPs and vents for the NHS. Its nearly broken her, the stories she can tell of utter housery that goes on would make your hair curl! The utter bonkers practices the half assed way of doing things. And the worse thing is the basic quality of admin staff within the NHS who are useless and make her job way more difficult than it should be. My particular favorite story is of 2 admin staff who required a A4 printed copy of the alphabet to be stuck to shelves as they couldn't quite grasp filing notes in alphabetical order!
What happens is that the vendor encourages feature creep requests by its customers knowing that the people requesting changes will be dimly aware at best of the cost implications. Moreover, it may be that the original system specification was incomplete or deficient in various ways and changes ($$$$$) need to be made to come up with something usable.
Clearly there needs to be a level headed manager who sits between these two entities to point out the cost to the department's budget for these added features will ultimately mean downsizing the departments workforce, "so do you really want to push ahead with your request?"
But as pointed out by others, that manager doesn't exists. They did once, but those with the skills to comprehend the outcomes were given every incentive to find a new role, with a new company, that might better appreciate their contribution to the bottom line.
It also needs someone who produces a working spec in the first place
I know of one project which 4 software houses rejected as it was unworkable as designed.
The CEO (who designed it) kept shopping around until he got a company who said "yes, we'll do it" - and they did, ontime, to spec & under budget. Of course, it was utterly unusable and they made serious bank on out-of-contract changes over the next 4 years.
$23 million, 2 lawsuits against increasingly loud critics and one fired CEO later, it was replaced with a $100,000 system (hardware and software) based on opensource - which was rolled out in less than 6 weeks worked perfectly on the first day
The funny thing is, when i studied systems analysis, one of the things they reminded us of constantly was the need to lock down the spec of the system early. To prevent feature creep, and the inevitable costs it generates.
I don't know whether the government employs it's own systems analysts (even contractors if not full time), or relies on the provider to provide systems analysis. Because, of course, if the contractor installing the system is the one providing the systems analysts, it's in their interest to alllow changes to the system.
There's a story of a manager who set a task for his team to create an emergency contact system for hikers stuck in mountains - the team went all out, creating what was effectively a smartphone and spent 5 hours aguing about the colour
It would have cost $2000+. This was the point that the manager retrieved the $100 EPIRB from his briefcase that he had in mind. These have a 10 year standby battery life, vs the 3 hours of the proposed device
It's not just about locking down the spec but not being afraid to call out the more outlandish stuff that happens along with knowing what's missing
They are not involved in this are they: The scandal of the settlement scheme for EU citizens?
Can we do anything right in this country?
FJ have been kicked out of france, germany etc. All those countries where they 'closed' their offices for financial reasons was they actually lost contracts because the software they made was incredibly poor, had backdoors that FJ could access etc.
They do this continually. They're now under investigation about backdoors into HMRC's tax system, where engineers can 'amend' how much tax someone owes.
To clarify, they didnt 'lose the contract'.
Fujitsu terminated the cpntract first, for breach of contract by HMG, for repeated and frequent failures to deliver on government obligations under the contract. HMG were utterly shocked and called Fujitsu in to 'discuss'. Fujitsu said there weas nothing to discuss as HMG had ben warned that they were in breach many many times, and that they (Fujitsu) had had enough.
Two days later HMg announcd they wer eterminating the contract with Fujitsu, blithely ignoring that it had already been terminated and HMG tried to pretend that it was HMG choice and to save face.
I know this as I know several people directly involved in the termination arrangements who were laughing at HMG panicing and flapping around as they had no experience of actually being terminated and were worrying about nothing except their jobs and HMG reputation.
It's not only "addiction" but also corruption as government has deliberately made it more difficult for other businesses to participate in tenders by enacting changes to IR35 that prevent smaller businesses from operating.
Small business effectively can't participate.
Medium business can no longer effective subcontract parts they lack expertise in.
Cowboys like Fujitsu then become a "safe bet".
I agree. Govt procurement frameworks are inherently geared to the big players, who, in my experience, couldn't come up with a new idea if you put it on a plate for them.
We responded to a RFP recently and were asked if we had made a mistake on our costing becuase we were considerably under the published budget. I'm a UK tax payer, saving the govt money is a good thing!!
Most of this stuff DOESN'T NEED new ideas, just the same thing - done well - in multiple countries
The problem is the massive amount of "extra coding" that goes into implementations at local level and isn't subjected to proper QA
It's better to have something which can do everything you need and just switch off the extra parts you don't need - something that British management don't understand when using Off-the-shelf vs having a team under them (prestige issues) all busily implementing tweakery
Do tell me how a small business could manage AND support a software system deployed to 1000s of terminals, including updates and maintenance. I'm sure we would all love to get involved in that.
These systems are complicated its why its the same half dozen supplier names because they are the only suppliers that can afford the huge costs involved
"1000s of terminals" sounds scary, but it really doesn't matter. Small teams can build robust complex systems.
Maintenance and updates are a separate thing. If systems required to have good documentation and interoperability, there wouldn't be a need for a single big company to service them.
they are the only suppliers that can afford the huge costs involved
This is another fallacy. In corrupt system of course the tenders will be structured in such a way that only the big players could realistically participate.
'"1000s of terminals" sounds scary, but it really doesn't matter. Small teams can build robust complex systems.'
Nobody of any competence in the subject would dispute that.
'Maintenance and updates are a separate thing.'
Utter nonsense, because in the real world it is a continuous program of installation, maintenance and update and that is what the big users are aware of and want. They have a business to run, are not stupid and understand that it all can't be magically installed in an instant without impacting their day to day operations. I guarantee you that the hardware / firmware levels of even dumb terminal No 1 and dumb terminal No 1000 will differ due to changing component availability during manufacture . If your terminal has any kind of OS I bet there will be revision differences there to.
'If systems required to have good documentation and interoperability, there wouldn't be a need for a single big company to service them.'
Again, drivel. PC's, networks and widely documented and interoperable OS components have been available since the Nineties and still there is a vast industry nailing them all together and struggling to keep it all working. If it's interoperability of the application you want, then the question is 'what do you want it to work with?' when that question is answered then the reality is that if it wasn't in the specification the next statement is 'I am sure we can, it will cost you this'. No business is prescient or a charity.
'This is another fallacy. In corrupt system of course the tenders will be structured in such a way that only the big players could realistically participate.'
This is where you finally display your tin foil hat.
You do understand that the customer sets the terms of the tender don't you? So if they set the terms such that only a large organisation can provide a feasible solution to meet their needs then that is their prerogative? Again, in the real world, it's someone doing their best not to get sacked for committing large amounts of their companies finances to some outfit with an unknown track record and unproven resources.
It's not corruption, it's something far simpler, it's human nature that when you don't really know what you are doing or are not interested in the technicalities or just have too many issues on your plate, you play it safe.
The old adage 'No one gets sacked for buying IBM' is as true now as it was when it was coined.
I don't know why your downvotes are far more than the upvotes.
I agree with you and unless & until we change from massive monolithic systems to smaller chunks with the ability to transfer the data between them (eg my local surgery does not need to use the same system as one in Cornwall) we are stuck with the mega firms getting the work.
Mind you I hate to think what a mess government would make of defining structures so data could be transferred.
Any large website has to cope with "thousands of terminals", all different, all configured by someone else and everybody just accepts that fact and gets on with it. And don't give me any crap about security either. Fujitsu et al are just repeddling the 1990s to dumb management fucks and coining it.
"Any large website has to cope with "thousands of terminals", all different, all configured by someone else..."
Realistically, they're running Chrome (or a derivative which for all intents and purposes behaves like Chrome) or Safari or Firefox. Apple can't keep Safari users on the same version, but the other two keep their users current. So you're looking at a handful of configurations - not thousands. The browser does all the work sorting out the user's configuration.
These days unless your browser isn't on their list you'll just be told to update it, even when it's up to date.
Reading the judgement on one of these cases it seems that they moved from specialist terminals to XP to W10. The problems were largely with the earlier system where they were, presumably, some sort of Fujitsu custom device.
To steal from someone else, the tenders are structured in a way that only very large organisations can qualify and respond. The insurance and indemnity requirements along can be very, very large which rules about absolutely anyone other than a Tier 1. I worked for an ambitious Tier 2 but we were still shut out of contracts as the bond requirements could be a few million quid which we just could not afford to fund and have in escrow.
Internal IT departments have done it since forever and on a pretty tiny budget too. The development (and thus support) team for a complex piece of software rolled out to thousands is often a couple of people and a coffee machine. I spent much of my career writing bits of software to do stuff that was apparently complicated and/or difficult, it's just what programmers do. Backed up by a competent ops team to do first-line support, that sort of development was unremarkable. If the software had problems that evaded testing, it was fed back to the team immediately and fixed, as was stuff that wasn't technically a problem but which annoyed users, hence the Problem Report/Change Request forms.
Okay, a lot of this is going back to the days of mainframes, minis and clusters but I honestly don't see what's changed (I mean other than the omni-menace that is ISO-9000, but even stuff like that is hardly a deal-breaker).
Right. And in that case, the original team remain on hand to maintain and support the system while they do other tasks within the internal IT dept. There is therefore an incentive to make a simple, robust and maintainable system given that it is they who will be supporting it. Whereas if you sub it out then the support bods will be low-paid plebs with no understanding of the system, long after the dev team have moved on.
The problem is the post office was being prepared for privatisation. The UK government wanted to reduce headcount, increase profits (not to mention cover up any scandalous IT disasters) so that it would look more valuable to a potential buyer.
No, it was only Royal Mail that was suggested be privatised: Green Paper on Postal Reform. The Post Office was to remain untouched.
For small values of "untouched".
The plan was/is to cut funding and make them dependent on partnerships with other businesses instead, and to dress them up as a private company to make them look more credible to other businesses as a potential business partner.
All that's happened is they've sold off their own buildings, become more dependent on sub-post offices which they've alienated by treating them like crap, and they get bailed out from time to time.
Yes and no.
If the post offices were meaningfully separate from the Royal Mail then the post offices could also act as collection and delivery hubs for parcel companies in competition with Royal Mail, which would make sense from the customers perspective. Want to send a parcel? Take it to the post office and pick who you want to send it with.
However, as the Post Offices are basically more or less exclusively providing Royal Mail services the separation is more on paper than it is in reality.
"Want to send a parcel? Take it to the post office and pick who you want to send it with.
However, as the Post Offices are basically more or less exclusively providing Royal Mail services the separation is more on paper than it is in reality."
Are readers aware of the announcement in Nov 2023 that (many? all?) Post Offices will now be offering Evri (previously known as Hermes) and DPD parcel services?
https://www.bbc.co.uk/news/business-67336778
"Post Office customers will soon be able to send parcels over the counter using DPD and Evri delivery services.
It is the first time in the Post Office's 360-year history that people can choose to use companies other than Royal Mail.
The new service will begin in selected stores in time for the peak Christmas delivery season, said the Post Office.
But shops in smaller, more rural areas of the UK may not benefit immediately from the new agreement.
Post Office chief executive, Nick Read, denied that offering deliveries with more carriers was a slap in the face for Royal Mail.
It was, he told the BBC's Today programme, a "natural evolution" for postal services.
(continues)"
"the separation is more on paper than it is in reality.!
I wish it was. Our local RM sorting office is round the back of the PO with a minute office at the side with its own door for "customer service". The sorting office hours are essentially those needed to get stuff out for delivery and not much more. If you're not in when a parcel's delivered you have to take your card down the sorting office during their opening hours rather than the more reasonable ones of the PO. It was not a split made with customer service in mind.
As to that, it's probably the inevitable end of any "Tell Sid" sell off. The rosy ideal of the Great British Public all owning shares in a sold off company- particularly one sold off at a significant discount- was always a fairy tale.
Because why would they not take the money and run when it's offered.
Well one of my family members built a very successful deliberately kept very small company where a 10K to 20K seat project was just a typical bed and butter project. No big deal. Design. Install. Roll out. Support. With clients you have all heard of. Fortune 100 clients etc. If you know what you are doing and a have a small core team of technically first rate people there is little difference between a 100 seat one location rollout and 10K's seat multi location, multi country roll out. If you know what you are doing. Good design and good management scale nicely. Bad design and bad management does n't.
The Big Guys are always hired by technically incompetent customers. Thats a given. Because apart from a few (very few) juniors I've yet to see any technical employees on a Big Guys team that was worth hiring as anything other than maybe a junior QA guy. Maybe. This is especially true of the senior / lead titles.
Governments and technically incompetent big businesses are paying for gold plated peanuts for not very bright monkeys,.
Just want to vehemently agree on everything you said.
It's insane how badly this area of government and state-funded IT is run compared with high quality private sector firms, even like bigger ones like Endava, or Epam.
Just so sad to think of the effect all this had on people's lives.
"Do tell me how a small business could manage AND support a software system deployed to 1000s of terminals, including updates and maintenance."
Do you actually work in IT? This sort of thing is routinely done with retail EPOS systems, and I assure you they can be every bit and more complex than the various services the Post Office provide. I know of a system used by several of the UK's largest retailers that was designed and built by six blokes in a shed, and they did the maintenance for the next few years until a private equity investor bought the business. One of those retailers has around 3,000 terminals, but there's multiple customers, all with 100s or 1,000s of EPOS terminals.
What on earth makes you think that properly designed, coded and documented software needs a small army to create and maintain it?
'I know of a system used by several of the UK's largest retailers that was designed and built by six blokes in a shed, and they did the maintenance for the next few years until a private equity investor bought the business.'
I have some experience in EPOS systems from back in the day and can tell you that the frontend hardware became increasingly noddy and disposable such that the majority of the maintenance became basically a shipping operation to and from the end user. I can also tell you that there is significant difference in system / software complexity depending on a specific retailers product file size and warehousing operational requirement.
It's no surprise that six guys could design and build a system (even in a shed) along with the maintenance if the complexity isn't there, then it's doable.
What is a surprise is that they would sell such a supposedly stellar business? Was it perhaps that it was too much work for them therefore not such a great business model? Yes a few years would cover the first iteration but at the speed hardware, software and customer requirements change it seems it wasn't a sustainable model either as in reality all systems require ongoing development.
Private equity money is unlikely to pile in unless it thinks it can make a shed full more money than the incumbent owners so looking from outside it appears they got lucky and decided to cash out when a decent offer came along. Nothing wrong with that but it's not a compelling argument that it was a sustainable business.
Yeah, yeah. Until someone moves on or gets hit by a bus. The reason they use the big boys is because they know when the shit hits the fan they have the resources to step up to the plate and fix things.
What on earth makes you think that properly designed, coded and documented software needs a small army to create and maintain it?
It doesn't, but that is quite a challenge if your customer base grows or changes, and with the inevitable corporate actions then to keep on top of. Easy when they all want the same thing. But if your customer has 1,000 EPOS terminals their particular needs will become important. Not all customer needs are the same. And then you need people with customer-specific knowledge.
“ Yeah, yeah. Until someone moves on or gets hit by a bus. The reason they use the big boys is because they know when the shit hits the fan they have the resources to step up to the plate and fix things.”
This one is not true. Large companies don’t have lots of idle highly skilled resources to jump in when it all goes tits up. The only thing they have is the ability to take the hit when it goes to shit and provide a barely alive level of service to try to unfuck stuff. So they won’t go under but your project is still gonna die. Smaller companies often have more skin in the game and can’t afford to just write off their million pound customer like your Deloitte, ibm, Microsoft’s can…or just tell them to sit in line.
The reason they use the big boys is because they know when the shit hits the fan they have the resources to step up to the plate and fix things.
The Post Office have renamed the compensation scheme from Historical Shortfall Scheme to Horizon Shortfall Scheme. Because over 20 years later after the initial rollout, shortfalls are still happening on Horizon. So much for the big boys stepping up to the plate and fixing things.
"Stepping up to the plate and fixing things" really means the customer spending millions on a big new shiny system from the same big boy.
The only difference is that the Post Office have (allegedly) given up prosecuting, but they still require an investigation and may still require the sub-postmaster to obtain legal advice. Also refunds to sub-postmasters under the scheme are taxable income because of course it's the Post Office and fucking people over is in their nature.
"1000s of terminals" are no longer a thing. These days nearly every large system actually runs over the web (not necessarily the public one), meaning that there are no terminals or dedicated network to maintain. Your delivery channel is just part of the regular IT infrastructure for any business. This usually applies even when the backend is IBM and CICS. You can deliver a green screen on a PC, but usually you've replaced it wth a prettified front end delivered over https.
I knew of a (logistics) customer that had an AS/400 system (IBM). The hardware was long out of support but they could go to eBay to buy replacement components. The system was so reliable they just kept using it. It worked, so don't touch it. I've seen many established businesses, normally SMEs, not change a system because it still works.
I started my career as software developer for a company designing multimedia servers, the first generation of totally-software based IPBX, but also ACD, vocal servers, and CASE tools so the customers could build their own addons... we had hundred of customers worldwide, in Europe, the US, in Japan, with thousands of servers. We had regularly new versions released. And for all this, we were 4 developers and 6 people for the support. And this was working fine.
SMB can be much, much more efficient than big companies. The differences are SMB have to pay their fare share of taxes and don't have the good connections in ministries.
These systems are complicated its why its the same half dozen supplier names because they are the only suppliers that can afford the huge costs involved
The half-dozen suppliers have huge costs due to their huge overheads precisely because they're huge. I work in such a corporation. Processes are long, complicated, hindered by too many people who have to contribute despite not knowing what they're contributing to. Bad developers who shouldn't be there, good developers who don't want to be there, not enough resources allocated where they're necessary, too many resources sitting doing nothing between projects but won't be allocated to other projects where they're needed because accountancy reasons.
Nowadays the hardware, network infrastructure, and network protocols are all off-the-shelf, so much so the big names would use the same as the SME. The software is the only difference, the big name would probably use a stack from 10-20 years ago, the SME would probably be up to date, the big name would probably use a mallet to hit every peg into a Java hole, the SME would use the right languages for the job.
It's the "Nobody ever got sacked for buying Microsoft" phenomenon.
It was not ever thus though. I think the nails in the coffin for small biz involvement in public sector IT were the introduction of national framework contracts and the EU public procurement rules. Both of these made it prohibitively expensive for smaller suppliers to even tender for public sector work.
There's also issues with undue personal influence. The head of UK Fujitsu during this scandal, Michael Keegan, is currently employed as "a crown representative to the Cabinet Office, managing cross-government relationships with BAE Systems as a strategic supplier to the Government, He is a non-executive director of the technology company Centerprise, "
He also happens to be married to the current Minister for Education, Gillian Keegan.
It's cronyism all the way down...
The "nobody got fired for buying IBM" mentality.
Speaking of IBM, they were responsible for the Canadian government's Phoenix pay system fiasco. My understanding is that they were the sole bidder and were able to insist on the waiver of penalty clauses in the contract documentation. If you are a small supplier, you definitely won't get that courtesy i.e., there are perverse incentives that favor the big suppliers. At the same time, many of those suppliers, IBM again being a notable example, have gone through bouts of downsizing and outsourcing with the result that the project team has a lot less experience and capability than the contract managers believed they were getting.
The backwards bit is that it's the need for fair competition that led to the current situation.
If I want to outsource some work I have to run a fair competition, put the work to tender via something like the Digital Outcomes framework. It takes weeks to get listed, a month or two to compete, you get 50+ submissions, 100 or so clarification questions and endless lawyering. Most of the time you can get most of the task done for the effort it takes to outsource.
So the frameworks were born, spend 6 months competing a single framework which lasts 3 years - and now it takes 2 weeks to get that work on contract.
From that we get strategic partnerships and then the safe bet.
"corruption as government has deliberately made it more difficult for other businesses to participate in tenders by enacting changes to IR35 that prevent smaller businesses from operating."
I'm as much against IR35 as you. But
1. it was a Labour government who introduced it. They would inherently be against what they saw as individual workers not being unionised. Ironically of course, it led to us forming a trade association.
2. The limit is, IIRC, a 20% share holding - a small business with 5 owners is outside.
3. A 5-person business is not going to be able to tender for anything of this scale. Even if it tried it wouldn't be able to get the finance to buy the kit.
Somewhere in the various threads someone posted a link to the technical appendix of one of the trials. It seems to have been written by the trial judge. It's worth a read, partly to get an idea of just how big the project was and also to see just how well a judge can get to understand the vast welter of technical evidence the case raised.
Many years ago, in another life, my employer was invited to "tender" (i.e. waste a fucktonne of money) on a contract that was clearly going to these scumbags. They quite rightly declined.
As did every other bugger. Fujitsu won by default.
Then there were EU regulations about state tenders requiring more than one candidate. Which led to the Alice in Wonderland world of the UK government paying companies to produce tenders that were never going to be accepted. I believe HMRC did that.
However, as with all the other "crises" facing the UK, this won't be fixed until there is a political will to do so (see also: US school shootings).
"(see also: US school shootings)"
You were doing so well until this, which as written is meaningless but I'm going to assume you favour increased gun control of some sort. Which isn't the problem, never has been and will not be fixed by 'political will.' The US constitution will put paid to that.
The real issue is the mental health and routine doping of ten's of thousands of kids.
There's a very strong case for legislation to prevent vendor lock-in on government projects and the development of minimum interoperability standards. Quite apart from over-runs and dodgy practices, when systems eventually require replacement it's currently necessary either to rip out the entire extant system and start again from scratch with a new vendor or be locked in to the existing supplier regardless of their performance.
Others have been saying this for years. See the file format wars of the early 2010s. Cabinet Office produced some half decent guidelines which were mysteriously withdrawn and nothing has happend since. At the time someone said "it would be nice not to be here in 10 years time" and now the CMA has opened an inquiry into the problems of lock-in with cloud.
It's always been about interoperability and for some reason (I think we can guess) government of any persuasion has always listened to those with a vested interest in avoiding it.
Let's not forget departments taking money for fighting corruption, being corrupt themselves.
If this behaviour was policed, we might not have this mess today where we pay highest taxes in living memory with nothing to show for it.
So the problem I have observed is that UK Gov depts don’t write good requirements. No end of times, big name company wins contract only to find requirements are missing, so contract change kicks in and value goes up. Contract also goes long. Big name company doesn’t know how to do some things, so subcontracts to SME or even the previous incumbent, who got squeezed out because big name company bid low to win. Big name company staff are poorly paid and under skilled, so client get rubbish service or late delivery. Nobody else then wants to touch it with a barge pole so big name company gets repeat business until client finally wants them out and refuses to allow them to tender.
It's not just the UK Gov nor is it just in IT, you could use the ongoing and never-ending Ferries Fiasco in Scotland as a prime example of this. Contract awarded and payments given out based on milestones being hit on ships that don't have an agreed upon design and which only recently had to modified again due to not meeting safety standards.
See the file format wars of the early 2010s
Yes, and see what efforts MS went to to derail "thou shalt support open standards" requirements that were starting to appear in public procurements before then - mostly across Europe. To the extent that they clearly stuffed national standards bodies to ratify their "has open in the name but isn't actually open"* proprietary standard as an international standard. That allowed them to tick the box while carrying on with their "lock out competition by making office document formats a moving target" policy.
* If you look, the standard is actually incomplete, has things like "binary blob in X format", and is generally a mess and an example of how not to write a standard.
We still seem to be locked in to the same assumptions about “vendor lock in”. How about, not have a vendor at all? This sort of governmental/institutional software should just be a function of a Gov IT department.
We’ve tried “competition”. It clearly doesn’t do what the ideologues claim it does, in practice. We’ve tried it dozens of times, and it always results in the same mess. What the current system *does* achieve is a large amount of resourcing on spurious bid phases (the general public would be shocked I think to hear that 10% of the entire budget is going on documentation which is thrown away as soon as the contract is won). The current system entrenches “the usual suspect”, so it certainly doesn’t achieve competition. And most of all, it achieves slopy-shouldered “not in the contract guv”, which is the death of these projects. The PO couldn’t get their hands on the error logs of the project they’d paid for, WTF.
We all know this doesn’t work. And yet when I point out the bleeding obvious, that if you want national-scale IT which is tailored to govt thinking, natiOnalised implementation would be more effective….out comes the kneejerk “but government is inefficient”’.
Can I remind people that “Fujitsu” in this case is just the old ICL, which was for all practical purposes by the 80s just a spun-out government dept anyway. So we’ve already got the slow inefficient bit, without any of the benefits of a tighter integration and throwing the contracts and marketing shmoozers out on their arses.
'Can I remind people that “Fujitsu” in this case is just the old ICL, which was for all practical purposes by the 80s just a spun-out government dept anyway.'
Citation?
I know this is so far from the truth you may want to review your handle.
Justhepureconjecture would appear more appropriate.
ICL was formed by the Wilson government in ‘68, by Tony Benn. Uk govt had a 10% direct golden share, nominally +10% by other ownership vehicles, but the board was govt appointed.
Major significant mergers/acquisitions like Ferranti were all decided by govt for both parties. For a good couple of decades, govt official policy was single-source ICL unless ICL couldn’t do it.
In my view, that’s a nationalised company. I am aware that in the European model, they still like to maintain the fiction with eg Airbus; and it’s considered bad manners to mention it. Apologies for offending you.
You didn't offend me, I am just pointing out to the wider audience that you are expressing an opinion as opposed to presenting fact.
A reference to the position in the sixties does not represent the situation in the company two decades later.
The fact that they had preferred supplier status was probably more to do with their position in the Military installations ( this changed ) and their supply to the less publicised departments in the Government/Military area.
Fujitsu came in and took a 50 percent chunk but they were also an amalgam of Canadian, American and European companies around that time.
Far from the picture you paint.
I would suggest starting at this wikipedia page and following the references as necessary.
"International Computers Limited was formed in 1968 as a part of the Industrial Expansion Act of the Wilson Labour Government. ... Upon its creation, the British government held a 10% stake in the company and provided a $32.4 million research-and-development grant spread across four years."
"ICL initially thrived, but relied almost wholly on supplying the UK public sector with computers. ... Until the 1970s launch of the 2900 Series, the UK government had a single-tender preferential purchase agreement wherever ICL could meet the requirements."
"The company had an increasingly close relationship with Fujitsu from the early 1980s, culminating in Fujitsu becoming sole shareholder in 1998. ICL was rebranded as Fujitsu in April 2002."
This bit should also have been quoted:
The main portions of ICL were formed by merging International Computers and Tabulators (ICT) with English Electric Computers, the latter a recent merger of Elliott Automation with English Electric Leo Marconi computers, which itself had been a merger of the computer divisions of English Electric, LEO and Marconi.[3] Upon its creation, the British government held a 10% stake in the company and provided a $32.4 million research-and-development grant spread across four years.[4]
So ICL was an amalgamation of private companies which in turn were amalgamations of other private companies or divisions (follow the definitions in the article). Sometimes the companies themselves instigated the mergers, sometimes the government did in return for cash.
> "you may want to review your handle."
We typically don't chose our handles according to who we really are but rather how we would like to be perceived. Possibly attempting to fend off in advance, what we deem were misperceptions already experienced in the past.
It is therefore logical, for somebody finding themselves often contradicted, to present themselves has "just right" (as in "you can't contradict me: I'm just stating facts" [my own facts is implicit]).
To the rest of us, readers at large, that looks more like a big flashy warning sign (not unlike dendrobats, mushrooms or ophidians sporting toxicity warnings by means of their distinctive colourful liveries).
I’ve recently left government IT and its pension after 3 years of waiting for leadership. The paralysis in the civil service is not down to poor staff or skills or ability, it’s down to political leadership and utter and complete inertia and paralysis. Also a lack of money to do what needs to be done. I am so much happier back in the private sector where money matters, projects have priority and there is adequate leadership from service provider and customer. So much happier.
I worked at a network / telecoms outfit once that will remain unnamed, they were small you probably wouldn’t have heard of them, it was so awful in terms of incompetency and unprofessionalism that I left after 2 weeks, no way am I allowing that company to tarnish my name by association.
If they’re still in business it will be a miracle
I worked for Fujitsu in Australia, their stuff was excellent and they supported it well. Unfortunately that bought ICL and things went downhill fast, essentially the ICL management took over. I took the opportunity to slide sideways into a spin-off where I was insulated from the semi-competent arseholes who took over.
I was seconded to the Revenue's internal IT team way back to help develop something. Same story, the project was delivered on time and under-budget, and also turned out to have spare capacity to do various other interesting things. Some of the internal politics and gnashing of teeth weren't always fun but most were a good bunch and stuff just worked. They did have a bit of a love-affair with ICL (and hence Fujitsu's entrance, sadly) but AIUI the computers and software that were supplied in the ICL years worked well; about the only thing I heard anyone complaining about was a mainframe-based text editor that was apparently a bit shit.
Horizon was very much developed in the ICL years. And to be honest, the reason why Fujitsu kept getting government contracts is because it was ICL with a new letterhead. Same customer reps, same salesdroids, the departments just kept going to who they were comfortable with, competence or no.
Was it really that far back? Blimey. I suppose my observation is that I knew some people who worked for ICL around the time Fujitsu took over way back and their morale absolutely plummeted; and while I'm not going to claim ICL was without its flaws, it seemed to rapidly get very much worse, at least from a bystander's perspective.
Before Horizon there was Pathway - reported on extensively by Vulture Central - one of the most expensive software fails in history up to that point (£1.5billion pounds). It was so awful that the DWP pulled out of its partnership with Post Office/RoyalMail
Pathway morphed into Horizon but many bugs remained - the accounting issue amongst them
I read that the origins of Horizon go back to ICL Pathway (a tripartite agreement between ICL, the Post Office, and the Department of Social Security).
They won a contract for an electronic way of paying benefits, the Benefits Payment Card, in May 1996. The Horizon Pilot was introduced in a small number of branches in 1996. This tri-partite scheme was abandoned in 1999. Pathway cited “greater than expected complexity” and “…major implications for the degree of difficulty of the project”
Then in July 1999 Post Office and Pathway agreed to utilise the project to automate branch Post Offices. This is what is now called the Horizon System and it was rolled-out from late 1999 onwards.
So they struggled to get a system designed to address one requirement working so they then used it as a basis for an all-singing all dancing accounting system!!!!!!
The Post Office fiasco has all the hallmarks of an incompetent attempt to take a monolithic legacy system and tart it up to meet requirements for massively distributed transactions. Confirmed when I read that Horizon is basically an old ICL system. Lipstick on a pig.
It says a lot about Fujitsu that they claimed there were no error logs*. Whoever heard of such a thing? And they apparently allowed support staff unrestricted access to live financial data without logging that. Has anyone here ever worked on a serious system where that was allowed?
No. In fact my experiences with the then Inland Revenue were quite the opposite: all staff who had access to the test systems (which had a small subset of live data) required positive vetting, which AIUI was an ongoing process which could see staff permanently turfed out at any moment if something was flagged up. Physical security was significant (keycard-controlled entry booths right in front of the permanently manned security offices, patrols, video surveillance, all areas were behind locked doors with keycards and/or combination locks, underground walkways so staff, documents and data didn't have to leave the premises to get to other buildings and so on) and there was definitely no remote access. They used me to physically move some data to another site as they didn't trust any couriers with it, even though I was all "but I have a hangover" (though more seriously, it went straight there and didn't leave my possession until it was at its destination; compared to, say, giving it to an MP who'd probably have printed it off and left it on the train). Live systems were even more stringently guarded, by actual security personnel who constantly hovered around the one time I was briefly allowed near it. I don't know the exact details about logging but there's no way it wasn't a thing, and it would've been pretty comprehensive because their internal security people were actively all over that stuff.
It's quite astonishing to hear of a system where general staff were allowed unrestricted access to probably similarly sensitive information without even logging stuff. Well, "astonishing" implies surprise and at this point I'm not, I'm just depressed about the state of things.
Sad thing is that the government keeps claiming that they couldn't possibly block Fujitsu from tendering in future is because they have the massive mainframes at HMRC etc.
It seems said ministers fail to realise that one *can* split services from actual hardware, so tendering for new mainframe type hardware is possible without tying oneself to having to allow them to tender for software projects like Horizon (or whatever the next project is).
Whether Fujitsu et al would go for that is another question, but money is money and money talks, and if you are given a choice of earning another 23.6 million on another mainframe upgrade or nada, I suspect the account manager(s) in question would say "yeah, we'll take those 23.6 million".
Yeah. IIRC that was the whole point of the shake-up with IBM back in the '70s when they started charging for their operating systems, because IBM's prior insistence on being the sole provider of the entire package was problematic and anticompetitive. I still think a major part of the problem is the outsourcing of development: government used to do their own and would take on various contractors (whether from the hardware and/or software providers or independents) to help out with their own development rather than to farm out the entire project. Though not without its problems (hard to say if government was actually more bureaucratic than DEC or just a different sort of bureaucracy) it worked pretty well, and certainly worked a lot better than the innumerable failures of the past 20-odd years courtesy of the usual culprits. You'd think by now someone would be doing something about it instead of "well maybe that sucks but we're still just going to keep on doing the same thing indefinitely".
The old online web portal for lodging and paying VAT worked just fine, and it wasn't even a bit clunky.
But they felt the need to change (ie. get rid of it) and make you run software that connects to their server and makes the data transfer "digitally" cos that's new and trendy and a buzzword.
I worked for one of the named suppliers - IBM - for more than 20 years, then worked for the UK public sector and ran a "small" £37.5m procurement as part of my job. I and my team were highly technically skilled and we had excellent support from the people used to running procurements who advised us on the process but left us to make the decisions. Without both of these aspects, procurements are often doomed, as the article says. On the other hand, if it's done properly with the right people and without too much senior management interference, it makes sense. We completed on time and on budget. As I expected we would.
Big IT projects shouldn't be outsourced. If HMRC is spending £250m on a project, they should be employing the engineers, project managers and architects directly. Pay them a good salary and recruit well so you get the cream of the crop. Once the system is created, they should own all of the IP of the system, and it will be fully documented. They can use contractors if needed here and there, but they should be only used as specific parts of the system, not to deliver the whole system.
Otherwise, this gravy train of incompetent contractors delivering failure and vendor lock-in will continue, resulting in poorer services for the rest of us poor fools.
This is another sign of corruption in the public sector. The wage scales are artificially kept below the market rates, to so that departments cannot actually find anyone competent to do anything in-house, but it is totally fine to hire a supplier and pay them over the market for the same labour.
It's gotten worse after IR35 changes that limited competition.
Ideally it should be as you say - things should be developed in house and cowboys like Fujitsu should be looking for work elsewhere.
“ This is another sign of corruption in the public sector. The wage scales are artificially kept below the market rates, to so that departments cannot actually find anyone competent to do anything in-house”
Just not true. Plenty of highly skilled civil servants. You sound like a bitter white knight IR35 sad boy.
We're (at my place of public sector employment) stuck in a never ending loop of job adverts, zero decent applicants, rinse and repeat because as has been pointed out repeatedly the scale we're advertising on is "wouldn't get out of bed" money for people skilled in the art. There's a hell of a crunch coming but meanwhile we seem to be staggering on with a "I see no ships" attitude to it all.
That would be the way to go, except for civil service politics and wage scales which would undoubtedly get in the way
A while ago there was a bit in the CS wide new email asking for feedback. Being autistic and prone to "saying it how it is" I did provide feedback - it was ... rather uncomplimentary ... and I made exactly the same points - the Civil Service (generally) just doesn't pay to keep the right people in the right places. That's not to say there aren't good people who accept the poor pay for their own reasons, but they are outnumbered when it comes to dealing with "the usual suspect" when it comes to procuring big projects.
I didn't get any feedback on that, but I did get an invitation for a chat with Simon Case to feature in one of the "5 minutes with ..." slots these newsletters contain. Being rather out in the sticks, he's not up this way very often (and I think it was the Russia-Ukraine business that first got his trip cancelled), so it could be a while before I get to meet him. First item will be to have a gentle dig about wearing the wrong blue - Oxbridge people will get that immediately ;-)
Of course, since then we've had political announcements like "massive cull of numbers" (dressed up in different words, but same intent), more years of below inflation constraints on pay, ...
I know more than one colleague who's "fed up" and thinking about jacking it in.
but this (while absolutely no excuse) is the case in any business and public sector that's consolidated enough, the internets, railways, energy supplies, etc. Now we're in the process of oursourcing health services and the education is the next juicy bit. Ironically (?) the public (tax payers) ultimately pay through the nose, either because it's run by a horrible wasteful and inefficient 'public' institutions, or because it's run by a less (?) wasteful and less (?) inefficient corporations that screw you how they want and any way they want, once you let those pimps into your parlour.
it's run by a horrible wasteful and inefficient 'public' institutions,
This was a carefully constructed myth created for the purpose of undermining and privatising public services./ Give a dog a bad name etc.
It's not public bodies that are dumping sewage into our rivers.
Many publicly owned and local authority service were run pretty well for the time.
Council housing was far better maintained than many of the current "Housing Association" properties seem to be.
There were, undoubtedly, good reasons for wanting the privatising of some services, like say British Gas, the GPO Telecoms or - topically-the Post Office, that were underfunded, underperforming and poorly managed. But no real evidence to say that the privatisation would actually solve the problems that they had..Rather than simply shaking them up from what was simply a rather Edwardian Age bureaucracy..
It certainly doesn't seem to have much good for the railways, or, as noted above, water.
It's not public bodies that are dumping sewage into our rivers.
Not now they've been privatised, no. However they were however doing exactly this in most cases from the same facilities before they were privatised; during which it was seen as being perfectly acceptable at the time. Which is good; private companies are being held to a higher standard than the same facilities were under public ownership.
There were, undoubtedly, good reasons for wanting the privatising of some services,
If you look into it you'll discover that it was a requirement to obtain the 1976 IMF loan to bail out Britain's economy. Loss making entities which had bankrupted the government were privatised and allowed to sink or swim. Some swum, many more sank.
It certainly doesn't seem to have much good for the railways
The railways were entirely built from the 1840's as private for profit businesses with private finance, which were only taken into public ownership after they were all loss making after WW2 when everybody had switched to new transport options like buses and cars. Notably when it was semi privatised this was done so with a huge maintenance backlog that built up while in public ownership, because given a choice of money going to the NHS or maintaining track etc guess who lost out?
Council housing was far better maintained than many of the current "Housing Association" properties seem to be.
It was. However a huge reason for selling it off cheaply when they did was that a huge amount of council housing stock was prefabs designed to last for 25 years when built post WW2, and so by the 1980's significant amounts of housing stock was coming to end of design life and so likely to start needing expensive repair work or replacement in the near future, which was avoided by selling them off to the owners; this neatly avoiding any government liability for repairs or rebuilding. If you think about it you'll no doubt remember entire council housing estates being bulldozed from 1990-2010 and either rebuilt or the land sold for redevelopment? That's why.
Honestly, it makes sense for the government to own some things. However, the problem is that if a private business is sufficiently crap compared to it's competitors then it goes bankrupt, and the customers go to the competition. This ensures a basic performance floor for business which tends over time to trend upwards in average service performance and downwards in average cost due to competition.
If a government run monopoly is sufficiently crap then you just have to take the shit service and like it. When well run it's excellent, but it's not always. Personally i'd be quite happy with government running some things, but there needs to be some mechanism for ensuring a similar performance floor to private business, and one hasn't been found and implemented in ~500 years of government owned infrastructure so far (Portsmouth Royal Dockyard was founded 1496) beyond periodically closing things and then starting again from scratch.
Surprisingly, one of the biggest problems in goverment operation of essential services is that they AREN'T a nationwide entity but run as regional fiefdoms
Education, police, fire, water - all need to be part of a national body, not at the mercy of council budgets - this means that poor areas are guaranteed to get worse service, etc. With national funding the wealthy can no longer ignore shortfunding in those deprived areas which has become self-sustaining deprivation
"But no real evidence to say that the privatisation would actually solve the problems that they had"
The New Zealand experience shows that turning such things into State Owned Entities works pretty well, however selling those SOEs off to private interests immediately undoes any good that might have been achieved
I think it is far too easy to claim that because in-house government projects worked well in the past, they would work well now. I suspect this wouldn't be the case. Back then civil service was seen by many to be an honourable duty; people tried to do the best they could. I don't think this is the case any more, in the private or public sector.
"Clark Vasey, who has been Fujitsu UK’s head of corporate affairs since 2014, is the director and founder of the Blue Collar Conservatives group, which he set up with the now “minister for common sense” Esther McVey in 2012.
Mr Vasey stood for election in Workington in Cumbria in 2017 as a Conservative Party candidate, but lost"
It's past time for legislation to effectively control the almost incestuous conduct so frequently seen around government IT contracts.
Budgets should be baked into the contract and legally enforceable (if you SAY you'll do it for £x then that is what you get and if you got your costs wrong, tough luck!)
Likewise timescales (with a reasonable allowance for unexpected issues) should be legally binding with significant penalties for failure.
Contingency funds should likewise be baked into the contract and similarly legally enforced, to a maximum of 20% of the total contract value.
Any company which fails to deliver on time & on budget should be legally barred from tendering for all future government contracts for a minimum of 12 months.
Repeated failures to deliver should see companies legally barred from tendering for at least 10 years.
Any company who's system is shown to be unreliable/inaccurate/unfit for purpose should be banned permanently from all government work and required to pay compensation, in full, to everyone adversely impacted by the system concerned.
Only when there are real, enforceable penalties for such unacceptable behavior which directly hit the bottom line of the culprits will we see an improvement.
Budgets should be baked into the contract and legally enforceable (if you SAY you'll do it for £x then that is what you get and if you got your costs wrong, tough luck!)
That only works if the other party to the contract agrees never to make any changes to the software. They'll never agree to that. Unlike, say, a bridge or a skyscraper, there's a general mindset that software is easy to change, and so additional requirements can be incorporated without any significant extra costs. That's what needs to change.
There's a bigger problem that (as with the case of POL), whilst the management may know the generalities of their operation and the intricacies of their particular role, they have little appreciation of the total process and software required, meaning they're easily fed lines of BS by sales reps and happily lap it up
It's not just Fujitsu. Oracle may have a good database, but when they sell software wrapped around the database, it rapidly turns into a pile of fetid dingo kidneys (They never did explain to me why they need huge java engines for simple accounting entry functions, etc)
All utter rubbish.
Jujitsu is staking all the flak here but most of the issues are on the buyers side:
"Budgets should be baked into the contract and legally enforceable" - legally unenforceable if the buyer changes the specification constantly, which always happens.
"Likewise timescales" - see above.
"...to a maximum of 20% of the total contract value." And if the buyer demands changes costing +100% of the original contract?
Until government/civil service learns to nail requirements down up front and stick to them, this is going to keep on happening.
In case it isn't clear, Fujitsu is getting the flak because it (through its employees) deliberately lied under oath in court, in numerous separate court hearings, in support of Post Office procecutions that resulted in the conviction and often financial ruin of hundreds of innocent people.
I've been listening to some of the inquiry session, in detail (must have too much time on hands these days), and there were Fujitsu technical people saying that there were problems with the systems early on, (data reliability is endemic) and that the Post Office were warned about it, but continued to prosecute anyway. No project of that complexity is ever bug free, especially early on and the Post Office should have been aware of that and and dealt accordingly, especially the trainwreck at the changeover from the early, to later horizon system. Corporate group think and a culture of deference also seems to have payed its part at Fujitsu, but a lack of true leadership and transparency are the real culprits in this story...
Nonetheless, those concerns were covered up by POL _and_ pliant FJ "expert witnesses" found to parrot the party line in court
POL _and_ Fujitsoc management colluded "to perpetrate a fraud upon the courts" in order to protect profits - something which is likely to eventually cost very dearly and all those who testified for POL in those cases should be looking over their shoulders
Most of this happened around the time that POL was separated from RM and there was pressure from "on high" to keep everything under wraps in order to not scare off potential buyers
I know POL was never sold, but it was being prepared for sale and whilst the mindset of manglement in such situations is to cover up stuff which might affect the sale price, I think the eventual realisation that serious bugs would have come out during due diligence (or given cause for massive legal claims by the buyers if discovered after purchase) were a large factor in why that didn't happen - especially when the liabilities of having the coverups exposed started to sink in - making them double down on covering up after the sale in order to avoid personal legal liability
Let's not forget: Whilst "limited liability" protects SHAREHOLDERS from financial fallout of bad decisions, it in no way shape or form legally immunises the MANAGEMENT or board from liabilities accruing from unlawful actions
"Let's not forget: Whilst "limited liability" protects SHAREHOLDERS from financial fallout of bad decisions, it in no way shape or form legally immunises the MANAGEMENT or board from liabilities accruing from unlawful actions"
So you are the person I can finally sell this bridge to!
The problem is a government, a political cast, a generation of managers who believe privatisation in general and outsourcing in particular is a good thing.
As fir privatisation that went well sewage in the rivers, monumental electricity bills and buying all your gas from Putin.
Outsourcing having Benn on both sides of the deal Nealy always ends in a long drawn out expensive mess.
The whole country needs re-education to try and forget this failed ideology.
"How governments become addicted to suppliers like Fujitsu"
Its laziness and "the boys club". So they'll pick the person that bids the lowest sometimes, if not most times, they'll ignore people who are in the tender process who point out flaws because "We just want to get this done" or "Its fine, we know them, its fine". Then once in the door and in fact before that, they visit site in their expensive cars, brief cases and firm (why is that needed) handshakes with the fake laughing, much like the character Phil in the IT crowd, series 4, episode 1, Jen The Fredo.
They'll all pat each other on the back, make sexiest jokes in the process and be awarded a contract. During their contract, that will be shit and issues will arise. The engineers will point all these out then a mass e-mail will go out by the exec team and specifically the person who "project managed" the contract stating "There are NO issues with the current software/application/hardware. I politely request there is no discussion of any issues in front of customers/staff and that any tickets raised are processed in a timely manor, again highlight to the end user that this is a one off issue". Rinse and repeat until the engineers dealing with it get fed up a leave. Looking back a year later finding out said "project manager" was promoted and that everyone still hates the new app.
Then a new contract will come up and "the boys club" will just fudge it to allow the existing company to get it so they don't have to bother with all the tender shit again.
Been there, walked away. We tendered for a local government service and it was made clear that we wouldn't get the gig unless we outsourced what we were fully capable of doing in-house to their nominated sub-contractor which individuals in the selection process had significant financial interest in. We'd have been allowed to cream off a little of the contract money for ourselves, most of the money we received would have gone to this sub-contractor who always subbed it out to the cheapest knuckle draggers they could barely employ. The end result is that a piss poor job would have been done, costing more than we quoted for doing it properly, the required sub-contractor would have made a tidy profit and our company would have been blamed for a crap end result.
The government (central and local) won't pay for skills they need, but for some reason will pay these large companies more to do the same jobs they won't pay their in house staff to do. After a couple decades I gave in and took the money with the big boys (not Fujitsu), doing the same job for the same sorts of public sector clients but for double the money.
I've seen some poor practices but nothing quite so dastardly as shown in the TV show.
Let's all repeat: "Scale is not complexity".
This is the core myth on which multi billion pound contracts are built. It's not just in government (though they are top of the list) but any large organisation that hasn't actually got the skill to assess what it is that they're trying to achieve (apart from in nice safe corporate speak, laced with a dash of lawyer)
They say: "It must be expensive because it's complex"
You ask: "Why is it complex?"
They say: "Because it's big"
What they mean is they can't actually describe the functions they wish to manage, do not know how to achieve separation of concerns, and cannot comprehend abstraction of their business processes. Add in a large estate to manage and suddenly the only "safe" option is to call in a consultancy firm who speak nice safe corporate speak, laced with a dash of lawyer.
And whilst there are certainly areas where being in the right crowd benefits companies (corruption), it is certainly the case that many organisations willingly sign up to incredibly inefficient and expensive projects because they cannot concieve of any other option. We see this happen in public services because they cannot hide it (forever), but it's also the case in private organisations - who may or may not hush up the hundreds of millions they've lost on a failed modernisation program, or new ERP system.
And in turn that failure of a big project is used to justify the belief that it must be complex - rather than the more depressing admission that a lot of the people commissioning such projects haven't got a clue what they're actually trying to implement.
At first, reading people went to jail, this is impressive!
Wait, that was a mistake? And nobody in charge is going to jail? Seen that in Canada with the Phenix payroll system, at the federal government. Criminal recklessness, zero consequences.
This is what happens when you have your technology stack entirely made of proprietary solutions. You "buy" a product which you never truly own. You can't hire someone else to do anything about it, it's not their secret sauce.
A free software mandate across government is a necessary condition to break such self imposed corporate capture. Long lived systems meant to last generations should never be subject to the whims of a single provider.
They didn't just not got to jail many of the senior executives went on to lead major public sector organisations and Paula Vennells was rewarded with a CBE for "services to the Post Office". So they actually benefitted from their time in charge. She's just in the last few days voluntarily handed this back with the current outcry casued by the tv show.
Why not. The ex-ICL bits seem to deserve special attention e.g. this from
(wait for it)
2003 in Computer Weekly.
Another PFI project, another disaster. Twenty years later, lessons are still queueueing up to be learned.
I haven't seen it mentioned in the recent posts and discussions.
https://www.computerweekly.com/news/2240049341/Courts-Libra-system-is-one-of-the-worst-IT-projects-ever-seen
"The cost of the Libra project to provide a national system for 385 magistrates courts soared from £146m to £390m, and the main supplier Fujitsu twice threatened to withdraw unless it was paid more money.
When seeking extra money from the Lord Chancellor's Department, which oversees magistrates courts, Fujitsu also issued deadlines for officials to make a decision on whether to renegotiate the contract.
Unless the department met Fujitsu's deadlines, the company might have withdrawn from the contract, said a report published yesterday (Wednesday) by the public spending watchdog the National Audit Office.
(continues)..."
"Although the principle of PFI was that the main risks were transferred to the supplier, the report by the audit office revealed that the Lord Chancellor's Department felt it could not afford to allow Fujitsu to withdraw from the contract. Magistrates courts could have been left with only minimal support for national systems that were still being installed."
(continues)...
"After nearly 10 years of disclosures in Computer Weekly about a series of failed IT projects to build a national system for magistrates courts, the Commons Public Accounts Committee asked the NAO to investigate. "
Around the same time, maybe a litle earlier, there was also this:
https://www.theregister.com/2002/07/25/govt_sidelines_fujitsu_in_courts/
and in 2003 the National Audit Office report was published: HC 327 Session 2002-2003: 29 January 2003 "New IT systems for Magistrates' Courts: the Libra project"
https://www.nao.org.uk/wp-content/uploads/2003/01/0203327.pdf
which in due course was covered elsewhere
https://www.theguardian.com/politics/2003/nov/11/uk.society3
which includes these words
"The MPs point out that Fujitsu was also behind the benefits payment card project, which was cancelled with losses of £1.1bn.
"This is not the first time that Fujitsu has let taxpayers down so badly," Mr Bacon said. "The time has come to consider whether the government should be awarding further contracts to this company."
The MPs' report found: "The department... ran a poor competition, attracting only one bidder, and it failed to take decisive action when ICL did not deliver what was required. For its part, ICL did not understand the department's requirements, took on excessive risk and underpriced its bid."
The MPs conclude that the computer firm performed poorly throughout. "As a result of these failures the cost of the project has more than doubled in just four years to almost £400m and magistrates courts still do not have the IT systems they need to manage their workload properly"
...
The whole outsourcing model has come full circle.
The tender of these contracts are way too detail and typically are geared for big companies that have big pockets to spend on fancy presentations to backup their bid.
Its then a race within the company to maximum the amount of profit against each contract.
There is an argument that this core system should have been developed within the Post Office itself and had the necessary resource to acquire the correct staff/resource.
Horizon was a custom solution for the Post Office so its not like Fujitsu was reselling a solution used by other companies already.
Its a wake up call for the rest of the world to understand the finer details of IT systems.
"Minister, you asked me to produce a paper on reducing the burden of Civil Service pensions."
"Yes Sir Humphrey, and you are a year late."
"I much regret that Minister, but we had to weigh the pro's and con's to produce a balanced solution. It entailed much research."
"So, you had lunch at Club, Sir Humphey?"
"Quite Minister, but the answer is brilliantly simple. Even you could grasp it. We merely transfer the Civil Service personnel to an external organisation. They'll take on the pensions responsibility."
"Won't that be expensive Sir Humphrey?"
"Oh no Minister, my contacts have assured me that it will not increase costs of current services."
".and nothing can go wrong?"
"Oh no Minister. We retain the services of those transferred out, and the Supplier provides additional technical experience."
And following links we can see it was incorporated into the Office of Government Commerce in 2000, which then was incorporated into the Efficiency and Reform Group in 2010, which was then closed down.
Result... government is even more at the mercy of big consultancies than ever before, but that's probably the way they want it.
I think it probably works something like this.
A lot of problem is because companies like this have become very skilled at winning such work. The issue is that a government department is obliged to compete such work. They put out an ITT, and get responses. Some of those responses may be reassuringly expensive. Some of them will be worryingly cheap. The treasury rules mean one has to seriously consider the cheapest "compliant" one.
This is where the problems probably set in. "Compliance" is essentially self-asserted by the bidding company - they say they'll do everything that's being asked for. The government department's staff might consider this to be total bullshit, especially if they know how much the last one cost, and can probably opt to do some due-diligence on this bid. The problem with due diligence will be that, essentially, that will come down to one person's word against another's and is something that might get tested in court if the bidding company were to choose to make an issue of it. No one would want to get involved in that, and so the company view is likely to prevail.
So even if the government department is well staffed with knowledgable people who can smell the bullshit a mile off, it's probably extremely difficult for their views to override the company's assertion. So they end up getting the job, with the key staff in the department knowing that it's going to go wrong. Having a contract is no use if it starts unravelling - enforcement can be very difficult, involves lawyers, takes time, all the while a public service is not being delivered.
A classic example is the security mess for the Olympics; G4 (was it them? Apologies if not - it's a while ago) won the contract, had an absurd scheme in place to call up the staff resources for the event, and in the end couldn't deliver. The Armed Forces had to step in and fill the gap (and did so marvellously and - as it turned out - for a whole lot less money than G4 were going to charge). There must have been people in the relevant government department who were deeply suspicious of the staff-resource raising scheme that underpinned the winning bid, but their concerns were insufficient to deny the company winning the contract and had to be proved correct by means of it actually all going wrong.
And if the government department does not have the knowledgeable people, they're defenceless. There is certainly an element of "you get what you asked for", and many a government contract has been beset by problems of contracts not really containing everything that is being asked for.
There was a big project between the Met Police and IBM once for a finger print recognition database thingy. IBM successfully delivered something that, by all accounts, was pretty good at fetching out a subset of dabs that matched to those from a crime scene, requiring a whole lot less expert interpretation to find the most likely one. The Met thought they were getting something that would give them a single, perfect match. Now, even in this day and age of AI, I think many here would recognise the difficulty in making a system match patterns with complete and reliable exactitude. No one in the Met understood that, and hadn't understood IBM's bid's detail. They sued IBM, lost, and had to pay them the £500million anyway.
Similarly with the recent Ministry of Justice courts system fiasco. They managed to blow £500million on an Agile development, when it was clear that no one in the Ministry really understood what it takes to be the customer of such a program (proactive involvement and comprehension of what's going on).
> Some of them will be worryingly cheap. The treasury rules mean one has to seriously consider the cheapest "compliant" one.
The requirements for mandatory liability cover, etc usually result in everyone except the "preferred few" big companies steering clear as the cost/benefit isn't there - even when the second tier companies are _the_ experts in the fields with many deployed systems
You could do with something to manage this and promote government tech savvy in Computers, software, communications, emergency and military services comms and provide hosting.
A Central Computer and Telecommunications Agency for want of a better name ……
… oh dear torn apart by Government interference - read the history.
https://en.m.wikipedia.org/wiki/Central_Computer_and_Telecommunications_Agency
Many government contracts concentrate on getting lots of "features" the money, include contradictive requirements and requirements that if fully met would not provide the optimal performance-value balance but does tick boxes. On the vendor side there can be extreme pressure to win on the commercial and sales people. The procurement process doesn't engender good dialogue. The technical people play second fiddle which is crazy because the systems need to work reliably. Even if a consultancy is engaged to guide and advise the process it is usually commercially and process focused because those are the people client side engaging the consultancy. Personally, I think tech, experience and critical thinking is undervalued and under-represented in the decision process.
But ... on the whole businesses seem to procure more successfully so it can be improved. No doubt their process would be considered to lack transparency or be unfair because it requires fewer to manage. In business it is simpler as a large damaging procurement results in severe financial strain because shareholders and banks, unlike taxpayers, do not keep providing money for repeated failure. So that system is self-cleansing.
I don't see an easy answer. It's not as if you can let a department collapse. Maybe outsourcing more is the actually better and have the departments become small but very savvy management teams, paid highly and fired for non-performance along with the outsourcers. Transparency is key but frankly our media is so unbalanced that needs fixing first.
Ian Dunt explains the underlying (and systemic) causes of the lack of technical expertise in the higher level of the Civil Service in chapter 6 of his recent book "How Westminster Works ... and Why It Doesn't". I highly recommend Dunt's book. It will depress the hell out of you, but at least you'll understand why big government IT projects always end in failure.
I'm convinced that, in any subcontracting process, be it commercial or government, you run an almighty risk of it going wrong if you / your organisation is itself incapable of doing the thing that it is contracting for. If you can't build it yourself, you're not likely to have the expertise to supervise some other organisation building it for you. It's even worse if the contractor hasn't built one themselves before.
With government projects, this is quite often the case. No one really knows how to buy something, doesn't really understand what it is that should be built, doesn't understand the consequences of this lack of knowledge. It's a serious problem.
The best answer for government is that government should do more things for itself, in house. The costs of outsourcing badly (i.e. wastefully) would more than pay for the dev teams to be able to do things for themselves. In particular, MoD contracts these days are a nightmare because MoD itself doesn't really know. Go back 40 years to when MoD had it's own research establishments and knowledgable people, things went a lot better I'm sure.
They ordered a Oracle system that was designed to do some other function and then ordered that it was altered to do something completely different. If you put the non-techys in charge of large IT procurement then this is what will happen. Where was the UAT (Users acceptance testing) for the Horizon system ?
Not unusual for the existing supplier to win the renewed contract especially if they have been around for Along time as they have the knowledge and experience that is hard to put a price on, especially if operationally, nothing went wrong, as in the case for HMRC. Another example is Fujitsu looked after MODs legacy systems until replaced by new platform 15+ years back and HP won, yet Fujitsu UK was contracted to provide field engineers globally for maintenance and hardware rollout, as everybody else had made their engineers redundant.
Some of us are old and ugly enough to remember the nineties when EDS, the former incarnation of HP, had so much government IT business - well over half - that questions were asked in Parliament and a policy of positive discrimination was put on the table. At the time I was an embedded contractor in a certain government client of Guess Who. Our programme board, in charge of implementing such policies, had a majority of EDS contractors on it.
That was over a quarter of a century ago. AFAICT, fsck all has changed.
Such monopolies have impinged on politics repeatedly throughout history in other quarters; others of note during my lifetime include service industries, defence hardware and the education of future Cabinet Ministers.
Ultimately, it costs governments time, money, persistence and ideology to maintain both the scale and competition to avoid such monopolies forming. But a) they are incapable of agreeing to that price, and b) is it a good idea anyway? Case in point: Rolls-Royce and the PV-12 aero engine. A pile o' shite when first plonked on the test bench, HMG nevertheless put all their faith in the company and dropped the competition and eight years later, renamed the Merlin, it won the Battle of Britain, went on to offer a supreme power-to-weight performance that even its successor the Griffon could not match, and the competitor to that - the Napier Sabre - cost four times the price. Hitler and his high command used to hold meetings whining about it. The lesson here is that neither R-R nor HMG ever gave up on the crappy PV-12. They just kept on re-engineering it until they got it right. Sending pilots to prison because the engine had stalled in a dogfight was never on the agenda. Instead, they listened: first came Mrs. Shilling's orifice, then the pressure carburettor. All these miserable bureaucrats need is being forced to use their ears. Statutory declaration of criticisms received, verifications and and remedial actions taken, along with stronger immunity for whistleblowers, would be a good start.
P.S. Another highly successful ploy was to embed a Civil Servant with the contractor, as well as the other way round.
Case in point: Rolls-Royce and the PV-12 aero engine. A pile o' shite when first plonked on the test bench, HMG nevertheless put all their faith in the company and dropped the competition
Did they?
Rolls-Royce continued development of their Kestral, Vulture and Griffon.
Napier produced their Saber engine all the way through the war. (despite early versions having problems; an engine failure killing Captain Martin, of Martin-Baker which led to Baker becoming obsessed with pilot safety and ultimately the Martin-Baker name being synonymous with "working ejection seat").
Bristol kicked out the Taurus, Hercules and Centaurus engines.
And this overlooks many other engines, including Whittle's jet engines.
You have your aero history wrong there, brother. Most of those were in different power classes.
Suffice to say that the Air Ministry made a very definite policy decision to put all their front-line fighter eggs in one basket, carried out a careful evaluation of all the various projects of the day, and turned the money-hose on the Merlin.
You just need to read a little deeper than Internet scrapings.
Because if you blacklisted all of the failures one of two things would happen...
Firstly any of the suppliers would sue for unfair barring.
Secondly if you got away with barring the f ups, you'd have none left to bid for anything.
It's almost like there's a common denominator in government IT projects...
The Reg should write a proper piece on the technical details of Horizon. There's plenty on www.postofficetrial.com
For the version in use till c2010 it seems the PO branches wrote to an xml file for each transaction - but a single free text field for date, time, amount, item detail - seemingly not node based.
At the end of each working day that lot of files got parsed and sent over to a central server - a system called Riposte was the intermediary, supplied by another company, Escher Group not Fushitsu.
Who needs experts.
Fuji are in the doghouse in Japan for exactly the same filings that they are in Britain
It makes you wonder who was influencing whom
Incidentally, ICL lost a lot of antipodean computer business in the late 70s-early 80s to Fuji (then Facom) due to their rotten performance in the 1970s
Many of these contracts have invoicing happening 3-6 months after work has happened with a further 3 month payment terms. How many suppliers can afford to be paid in time frames of 6-9 months, while still paying their staff in that time?
Then we've got the proposal phase which can be another 3-12 months before even starting work
The Inquiry Videos, though very slow to watch, have revealed lots of facts about how this debacle came about..
At the bidding stage, the ICL solution was deemed to be the weakest, with IBM stronger. However, IBM was ruled out for two reasons firstly, its reluctance to accept the open-ended liability (risk)
of future benefit fraud. This condition was imposed by the DSS/BA who were obsessed with it. and secondly,. IBM added sums to its bid in an attempt to protect itself from such a ridiculous risk, which made it more expensive. So, in effect, ICL won on price. Ironically, after only a few years, BA pulled out and the fraud risk went away. as did the PFI structure for the contract, and the price went through the roof as the deal became more like a normal contract.
It also appears that the initial Horizon version (subsequently called Legacy) was based on a demo model created by ICL in the bid phase. The barrister asking the questions likened this approach to that of an architect who shows the client a model built with balsa wood rather than a real building. The counter element was a "masterpiece" written in Visual Basic.
Astonishingly, the PO accepted the system although it was known that it had errors in its cash accounts modules. They agreed that it was acceptable so long as the errors occurred in less than 0.6% of transactions. Not such a small error rate when millions of transactions are considered!
The revealed documents show minutes of meetings at the Ministerial level of HMG. Most interesting are the minutes of a meeting with the Japanese CEO of Fujitsu. This took place when the Treasury was pressing to cancel and start again. The Fujitsu boss said that if Horizon was canceled, there might be a bad reaction from other Japanese companies considering their investments in the UK. Though not explicit in the paperwork, the car industry springs to mind
Also revealed is the fact that around 2015 IBM worked, at the PO's request, on a replacement for Horizon. Again, Fujitsu "threw Teddy from the pram" and the work was canceled. It appears that IBM sued or threatened to, and was given millions for its nugatory work.
While much of the media is enjoying the witch-hunt of a few of the seniors, the reality lies with the incompetence of the original ICL solution, the failure of the Post Office to manage it & the sinister attitude of Fujitsu.
Nice precis but not convinced regarding your conclusions.
The initial version of Horizon was substandard. Barrister used a nice metaphor. (No mention of later versions.)
'the PO accepted the system although it was known that it had errors in its cash accounts modules. They agreed that it was acceptable so long as the errors occurred in less than 0.6% of transactions.'
So the customer accepted the offering despite acknowledged issues. For their own reasons they chose to go forward with it. (Other sources appear to suggest this would be due to their own timeframe requirements.)
At the start of the process Fujitsu accepted high risk and potentially punitive terms from the customer, but IBM wouldn't,
Having awarded the contract, the customer then changed the terms again and these changes removed these risks and happened to favour Fujitsu.
Then, later in the product lifecycle they engaged IBM again, (It appears to provide contract leverage) but when Fujitsu appear to apply their own business leverage and the customer concedes, IBM apply their leverage and receive 'millions' from the customer.
You conclude Fujitsu have a 'sinister attitude' but as there is no mention of IBM's behaviour presumably you think that their attitude is above reproach?
Not sure I agree with your conclusion regarding Fujitsu on this info, I see two companies doing business, (no one said multi million £ business is pretty) but hey, that's me. Still, thanks for distilling it down.
Sunak and Co are finding out what happens when people in this country get seriously pissed off.
They tend to think they can do whatever they want and all that will happen is a few grumbles here and there.
The problem with politicians is that when they get into power they think they rule the country.
In fact, they're civil servants elected to manage (or usually mis-manage) UK Plc. They are not kings or queens, they're employees.
And we are their employers.
If there's any good to be had from the PO scandal it's giving the government a serious wake up call that they don't have carte blanche.
“ Sunak and Co are finding out what happens when people in this country get seriously pissed off.”
They continue insisting white is black (and vice versa), banking the cash, making contacts, and adding largely unearned credibility and gravitas[1] to their resume until their positions become so completely and utterly untenable that no amount of bare faced arrogance can survive, continue in lucrative careers as non-exec board members, lobbyists, consultants, and in high profile lecturer/speaker gigs, then eventually retire as highly respected “elder statesmen” with a coterie of protégés and acolytes poised to take up their mantle next time around the cycle
Yeah. That’ll totally teach them…
[1] Note that Sunak has now managed to tick off the “ordered a military operation” box.
a good idea will be to hold a post mortum into the whole horizons system when everything is over, the postmasters compensated and senior PO execs in prison etc etc.
So that us mere mortal programmers/developers/project manglers and the students that are taking computer science 101 can see where the failures were, what caused them, and how we can avoid such failures in the future because those who do not learn from history are doomed to repeat it.
One of the positive things my boss does is not to yell and scream at failures, but to look and see where we went wrong, and to learn from that and implement solutions that avoid the problem ever happening again, and it varies from training operators, to ensuring documentation is written... or correctly written, or QA procedures are improved, or a final visual check before we box up the stuff and send it (thats saved us several times after painters applied the wrong colours), or whatever is needed although making the same mistake twice does involve the meeting room with the loose window and cunningly positioned woodchipper ominously running below.
But the other thing this shows is the lack of technical knowhow at the top... again (I'm thinking boeing here too from another of this week's stories), I dont expect a C-level exec to understand the ins and outs of concurrent programming, but I do expect him/her to be able to smell out bullshit supplied by whatever sub-contractor hes talking to.
But we all know what will happen
Lessons will be learned
Postmasters will be cleared/compensated
Senior level execs will retire with a full pension
And the low level sods at ICL/Fushitshow who created horizons will be thrown under a bus.
And 5 years down the line ANOTHER major government IT project will go titsup costing billions in losses
I was wondering about the source of the image for this article - I was thinking one of the statues of the Monks from the "Lie of the Land" - rather appropriate I thought. Alas no, cranking over the evil machinery of the internet I find its a cropped image of a sculpture outside of a library of Belfast's Queens University.
This is hardly just as British problem. American governmental units of all levels are notorious for poorly managing IT projects. Oracle in particular is notable for walking away from a project that is deemed unfixable afte,,r say $150 million has been spent. Then they just get another contract with the same entity.
Republican politicians in the US are always talking about waste, fraud, and abuse, but they ignore stuff like this which happens all the time. Larry Ellison of Oracle is a notable Republican donor.
one of the single biggest issues with software creation is that unlike any other aspect of business or manufacturing, the actual needs, processes and steps aren't fully worked out before the code starts being committed
In the case of these bespoke systems, nothing is modular so code blocks can't simply be improved and swapped out, which in turn makes it a cash cow for the contractor
The bottom line is that the management of POL weren't competent to write requirements, nor cognisant of it to hire in specialists who could actually create the requirements and contracts
As far as using WinNT is concerned... It's not exactly the first fustercluck I've seen and it's always championed by idiot managers who think that they know better than the experts
It's not just computers. Certain large companies, in a variety of sectors, have become very closely intertwined with government at all levels. If you want better value for money, for large IT, infrastructure, education and health projects, then a way needs to be found to break that. We have a lot of very good SMEs who are more than up to delivering an awful lot of government-funded work and who, at the moment, would never stand a chance.
I was responsible for the tech response on a government pitch a few years ago and the admin required for the response was so enormous that smaller suppliers wouldn't stand a chance. Unless you could take a whole senior layer of people off their day-to-day work and have them do nothing but write the response to brief, you wouldn't even be able to write a response, never mind actually win the thing. Work on one and it's really obvious how the pitching requirements drive massive companies to specialise in winning government contracts.
That's not totally unreasonable - big companies deliver big projects. Except for problem two.
Problem two is that I couldn't see anywhere in the brief where capability to deliver the service was being assessed. You were marked only on what you said you would do and not on whether you were likely, based on the staff at your company and past performance, to actually be able to do it. The whole process was geared towards creating a beautiful fantasy technical specification, regardless whether your company was capable of building what you'd described. Most beautiful fantasy wins.
My guess is that there's an underlying assumption that if you don't deliver the project as described, you won't get paid, which is supposed to be the disincentive to submitting unlikely fantasies. But politics doesn't work like that. Once you're in, you're in, and everyone involved on both sides has a huge incentive to make it look like the project is going well.
"Unless you could take a whole senior layer of people off their day-to-day work and have them do nothing but write the response to brief, you wouldn't even be able to write a response"
Probably a good task for one of those AI thingies - and you can even tell it to throw in a few buzzwords, etc
Some of the reason for Fujitsu getting these contracts is because many of the Government systems still run on ICL 2900 VME - yes, I know it is now called something different like DME or other names, and I know it now runs in emulation or something on top of generic Fujitsu Linux systems, but the code and the middleware like the database and the transaction manager is based on ICL VME. A specific example is the Customs Handling of Import and Export Freight (CHIEF), but others are I think the VAT system and DHSS systems. Government is trying hard to get off these old systems, but the length of time it has taken to replace CHIEF just shows how hard it is. Meanwhile, if Fujitsu decided to pull out of UK and pull out of supporting all that middleware, I wonder how long it would be before government systems collapsed.