Hooptedoodle #57 - The Banks & the Krell

It is easy to use a blog as a dumping ground to share visions of gloom, or to have a sad little go at someone or something one has a personal grievance about. I try not to go overboard, and I try to keep it humorous, but I admit that I have a tendency to make a noise if I think something is sufficiently bloody awful to warrant it. I realise it may not achieve a great deal, but at least I’ve contributed my rather shrill tuppenceworth. This, I am led to understand, is one of the merits of the Internet.

Like most of the population of the UK, I’ve been a bit depressed by the state of our banking industry recently. I never worked for a bank, but I did work in the finance industry for my entire salaried life, and my views on this may be better informed than most.

There is a lot of virtuous outrage (partly fuelled, as ever, by public envy of personal wealth) over the situation of the gentleman at Barclays who may or may not be answerable for some stupefyingly corrupt behaviour in the Bad Old Days just before the world ended in 2008. I don’t wish to add any further silliness to what is already a hysterical issue, but I am mystified by his defence that it wasn’t him, that there were people who worked for him who were responsible, and he didn’t know.

OK – I’m not actively involved now, and maybe values have changed, but it seems to me that:

(1) If they pay you a lot of money to be in charge, then ultimately you are accountable for what happens in your area.

(2) Naturally you cannot know everything that goes on, but you are obliged to stay on top of things – to ensure that governance, procedures, rules and an ethical culture are in place to check that staff know what they can and cannot do, and to enforce correct behaviour. If you do not manage to do this – and it will not be easy – then you have failed in your job and you are answerable.

(3) Thus (in my extremely humble opinion) the man from Barclays either knew what was going on – in which case he is culpable – or else he was not in control – in which case he failed in his job and is therefore still culpable. It’s a tough life in a top job – that’s why they get all that money.

Anyway, that’s not what I wanted to write about. I wanted to write about the Krell.

If anything, I am more alarmed by the software failure at NatWest Bank last week – a system change caused problems which separated many of NatWest’s customers from their money for a period of days, and generated huge inconvenience and actual hardship. Scary.

On a smaller and less disastrous scale, I had a customer’s-eye view of the recent switch of Bank of Scotland to Lloyd’s Bank’s computer systems. It wasn’t good. For a period of some months, the bank was running with an interim system which had customers queuing up to fill in slips of paper in a manner such as I had not seen since the 1980s. Retro banking. I am confident that Lloyds and Bank of Scotland did their best in a difficult situation, but for a while their systems were really not fit for purpose. It’s a dangerous sign, but on a number of occasions I found myself thinking that this would not have been tolerated in my day.

My day? What the blazes has it got to do with me, then? Well, as it happens, I was a computer person for almost all my working life. By the late 1990s I was in charge of all business software development for a very large financial institution which shall forever remain nameless, and I knew the guys who had the equivalent roles in most of the UK banks – and they were good. They would not have rolled out any system which didn’t work, or which used customer inconvenience as a buffer to tide them over.

So what has changed?

Well, I was also around when the seeds of the coming storm were sown in the 1970s and 80s, so I have a very fair idea what I’m talking about. Which brings me to the Krell.

I don’t remember the details – maybe you do – but in the film The Forbidden Planet the Krell are a mighty, super-intelligent race of beings who have died out (for reasons I also cannot remember – perhaps they smoked), and part of their legacy is a collossal underground computer installation, which has been running for thousands of years and is still running – and nobody knows what it is doing.  That’s the important bit – since no-one knows what it is doing, no-one dares switch it off.

Right. Back to the 1980s.

In the 1980s I was a business systems analyst. We were hotshots. We would go into a traditional business department and we would ask them all about what they did and how it worked. We would capture the expertise of some very experienced and intelligent, professional people, and we would draw dataflow diagrams and build data models and we would automate their processes. It was brilliant. We used to make people happy – we took huge amounts of drudgery out of their jobs, we built in safeguards and automatic audit trails, and we saved them enormous amounts of money. When Jeannie who did the commission work became pregnant and left (and people did things like that in those days) then they didn’t need to replace her. Not only that, but any new staff who did come in required much less training, since a lot of the expertise and decision-making was now built into the computer system.

Fantastic. It was a wonderful job – people actually loved us. I have never been so happy at work.

Move the clock on 10 years. It is time to go back to one of our departments and their 1980s systems, and see what needs to be done to get things squeaky again – because there is now an accumulated tangle of 10 years of emergency fixes, rushed changes to support product launches and new regulations. Time for a detox.

Problems. If the analyst sat down with the new department manager in 1990, he might well be talking to someone who had no experience of this area before the systems were put in. It was almost certain that this manager would be unaware of some of the business rules, because they were now built into the desktop system – they just happened automatically. Similarly, the new model of the business process they agreed on might well omit a vital job which happened every night at 2am in the middle of a batch run which no-one understood any more.

Around this time, we used to talk a lot about system ownership. Business managers would laugh at this, and produce comic visions of putting their system in a bag and taking it home, but by and large they had washed their hands of understanding. When the computer systems arrived, responsibility for understanding the business shifted by default to the IT people. They had, after all, got everyone into this mess.

Well, the bad news is made even worse by the fact that the computer analysts had moved on as well, and the constant focus in the business world will always be to cope with new changes. Maintaining the old stuff is a lower priority – especially when it comes to allocating the budget. Yes, we know the roof is leaking and the foundations are sinking, but what we really want is a shiny new barbecue and some of that decking stuff. Great.

Everyone remembers that there was a huge panic prompted by fear that the year 2000 would cause disastrous software failures – a lot of money was spent and a lot of effort expended. Since we are all still here I guess it worked, but the thing I remember most is the effort that went into digging into that mysterious old software – paying over the odds for people who knew how it worked and – most scary of all – finding people who could still read the ancient languages it was written in. Looking, in fact, for surviving members of the Krell.

And still time passes, and still the software deteriorates, and still our understanding of what the great machine is doing becomes more hazy. Yes, business managers should have paid more attention and kept in touch with how things work, and – certainly – IT people should have spent less time obsessing about crap like Information Engineering and client-server and object-oriented and (who can forget...) artificial intelligence, but the fact is that they didn’t. Lots of money should have been devoted to keeping the old systems up-to-date and clearly understood and operationally viable, but it wasn’t.

The Krell are dead. The machines are still running. The problems at NatWest are just the beginning, gentlemen. The decay of old systems is exponential. Building replacement systems is not possible, because nobody any longer understands what it is they would be required to do.

Welcome to the beginning of the end. I must have a look on Amazon and see if I can get a cheap DVD of The Forbidden Planet.