Reading Through the Logs

  • That is indeed an awesome tweet, thanks, hadn't seen it!

    I'm going to say that it looks like a right pickle. By which I mean that the subsequent account functioning and cross bleeding from account to account, and strange activity like the future payments, means it's impossible to just go back to where they started for many accounts. It's hard to see that the bank can ever fully recover - but I guess they can restore where needed and take the losses on the chin, against the future £100 million savings per annum - if things do get unravelled. Presumably many accounts have continued to operate OK.

  • I read through a lot of various transaction logs from various financial and other systems. It's really easy to make assumptions,mistakes and underestimate the complexity of business and real world environments.

    I think it was the hubris of the "modern" management of the bank to think they could replace a complex system that had grown organically over decades and had probably lost most of it's institutional memory in the form of retirements. It's the same mentality that brings us the insecure IoT and folks thinking that JavaScript programs with hundreds of externally managed dependencies can be "secure".

    We know how to build good, secure and real time software. We have quality modern languages and best practices that the industry ignores. Instead, we would rather have fast, insecure hardware and the latest fashionable framework. We are still tool using primates.

  • chrisn-585491 - Tuesday, May 8, 2018 6:59 AM

    ...
    We know how to build good, secure and real time software. We have quality modern languages and best practices that the industry ignores. Instead, we would rather have fast, insecure hardware and the latest fashionable framework. We are still tool using primates.

    Along the line of your comment, I place part of the problem back long before this project, when new features and functionality was apparently jury-rigged into the system with little thought to future issues. Every system should be maintained with the knowledge that it will one day need to be replaced.

    ...

    -- FORTRAN manual for Xerox Computers --

  • There is an amazing amount of slop in today's business world. I have no doubt the bank will continue, albeit with worse profits this year. Or maybe no profit, but they won't let it fail. I wouldn't be surprised if they renamed it in a few months.

  • jay-h - Tuesday, May 8, 2018 8:17 AM

    chrisn-585491 - Tuesday, May 8, 2018 6:59 AM

    ...
    We know how to build good, secure and real time software. We have quality modern languages and best practices that the industry ignores. Instead, we would rather have fast, insecure hardware and the latest fashionable framework. We are still tool using primates.

    Along the line of your comment, I place part of the problem back long before this project, when new features and functionality was apparently jury-rigged into the system with little thought to future issues. Every system should be maintained with the knowledge that it will one day need to be replaced.

    I'd agree we know, as an industry how to build good, secure, real time software. As many individuals, we don't. In fact, knowledge is very unevenly distributed. We don't have a formal set of things to learn like some other disciplines, but even they make mistakes. Witness the bridge in Florida, recalls by Toyota, and more. The best still make mistakes in complex endeavors.

    That being said , Toyota leans a lot,  as do others, and overall quality improves over time. Today our failures are more often because we place higher priorities on items outside of quality or durability.

    Should every system be maintained? I disagree. We may change how we do things and abandon some software, even software that seems really critical. The world changes. Banks have rules to operation and those ought to be documented, and a new system should be able to replace it. I'm not sure this one didn't. What I suspect is they skimped on testing, using happy paths, or limited load to prepare. They "thought" they knew banking software, but apparently didn't know it well enough for a UK system.

  • Steve Jones - SSC Editor - Tuesday, May 8, 2018 8:28 AM

    There is an amazing amount of slop in today's business world. I have no doubt the bank will continue, albeit with worse profits this year. Or maybe no profit, but they won't let it fail. I wouldn't be surprised if they renamed it in a few months.

    https://www.theonion.com/university-of-nevada-renames-vito-corleone-school-of-bu-1819579865

    ...

    -- FORTRAN manual for Xerox Computers --

  • Steve Jones - SSC Editor - Tuesday, May 8, 2018 8:37 AM

    jay-h - Tuesday, May 8, 2018 8:17 AM

    chrisn-585491 - Tuesday, May 8, 2018 6:59 AM

    ...
    We know how to build good, secure and real time software. We have quality modern languages and best practices that the industry ignores. Instead, we would rather have fast, insecure hardware and the latest fashionable framework. We are still tool using primates.

    Along the line of your comment, I place part of the problem back long before this project, when new features and functionality was apparently jury-rigged into the system with little thought to future issues. Every system should be maintained with the knowledge that it will one day need to be replaced.

    I'd agree we know, as an industry how to build good, secure, real time software. As many individuals, we don't. In fact, knowledge is very unevenly distributed. We don't have a formal set of things to learn like some other disciplines, but even they make mistakes. Witness the bridge in Florida, recalls by Toyota, and more. The best still make mistakes in complex endeavors.

    That being said , Toyota leans a lot,  as do others, and overall quality improves over time. Today our failures are more often because we place higher priorities on items outside of quality or durability.

    Should every system be maintained? I disagree. We may change how we do things and abandon some software, even software that seems really critical. The world changes. Banks have rules to operation and those ought to be documented, and a new system should be able to replace it. I'm not sure this one didn't. What I suspect is they skimped on testing, using happy paths, or limited load to prepare. They "thought" they knew banking software, but apparently didn't know it well enough for a UK system.

    I doubt the best or even the average were involved with the bridge in Florida. There's probably enough criminal neglect in the engineering, planning and construction of that boondoggle to hang a score of people. And Toyota got reamed in court about their code practices and management. And these are engineering fields with supposedly higher standards.

  • It seems that IT within banking, mortgage, healthcare, education, and government is far from being a model of excellence. I feel that the big name IT contracting, ISV, and implementation firms that cater to these industries are actually NOT among the best at what they do, and there are better options, like Microsoft for example.

    "Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic. Login to reply