• Num10ck@lemmy.world
    link
    fedilink
    English
    arrow-up
    0
    ·
    4 months ago

    theres countless major commercial, government, etc organizations that are totally dependent on systems that are 50+ years old. Like Marriott, Macy’s, Sears, Autozone, Transamerica, Air Traffic Controllers, Concrete mixer logistics, etc.

    It would be career suicide to truly tackle the problem, and previous major attempts gave been abandoned.

  • corsicanguppy@lemmy.ca
    link
    fedilink
    English
    arrow-up
    0
    ·
    4 months ago

    Hey, remember when we replaced only the broken stuff and let the other stuff keep working? Of course not: your entire country is barely over 200 years old.

  • Admiral Patrick@dubvee.org
    link
    fedilink
    English
    arrow-up
    0
    ·
    4 months ago

    Best guess would be banking systems and older defense systems. e.g. until 2019, the US nuclear command systems still used 8" floppy disks from the 70s.

  • deegeese@sopuli.xyz
    link
    fedilink
    arrow-up
    0
    ·
    4 months ago

    From a brief visit to the corporate website it looks like you can run MCP on a VM at a Cloud data center.

    This lets business keep running the ancient stuff while using modern hardware and devops practices.

    Those customers are never migrating off it.

  • sylver_dragon@lemmy.world
    link
    fedilink
    English
    arrow-up
    0
    ·
    4 months ago

    Probably running half the US Government’s systems.
    The US Navy famously paid Microsoft to keep supporting Windows XP well after it’s End of Life. There’s probably some highly critical mainframe running in a basement somewhere, with no backups, spare parts cobbled together from failing systems and some gray beard wizard keeping it all spinning.

  • Stern@lemmy.world
    link
    fedilink
    arrow-up
    0
    ·
    4 months ago

    Probably the same type of shit that cobol is. Some absolutely ancient system that they will never properly update because a 400k a year programmer is cheaper then doing so.

    • IMALlama@lemmy.world
      link
      fedilink
      arrow-up
      0
      ·
      4 months ago

      If it was only one shitty ancient system it would be one thing. For the company I work for it’s about 10 big interconnected mainframe systems with hundreds of non-mainframe systems cobbled together around them. They’ve been in place since the 80s, but you can trace their business logic back to the 50s and 60s. They start at cataloging all our parts and get into purchasing components from suppliers, describing the products we assemble, managing the supply chains for our factories, order management from our customers, etc.

      Replacing it all will be massive chore, but it’s becoming more and more clear that we need to. At the end of the day, capturing and understanding data in them takes so much skill that we have entire departments dedicated to being an interface between the actual users and the mainframe. The business rules might have worked before the products we build contained electronic controls, but everything is starting to implode now that “parts” also includes software. This has resulted in manual workaround on top of manual workaround.