• python@lemmy.world
    7 months

    Not to spread concern or anything, but the electrical grid is managed and controlled by software. And that software may or may not be very reliant on AWS. I’m probably not allowed to say more than that.

    • antimongo@lemmy.world
      7 months

      Power company engineer here, it’s true that a lot of our supporting and analytics software went down during the AWS event.

      However, most devices that actually control grid units (called bulk electric system cyber-assets) are air-gapped or utilize a data diode.

      FERC Reliability Standards and NERC CIP

      However-er, flipping through those standards just now, turns out it’s 100% permitted to connect your “bulk electric system cyber-asset” to a cloud integration if done compliantly.

      • marcos@lemmy.world
        7 months

        The process to decide to turn power plants on and off isn’t air-gaped.

    • lividweasel@lemmy.world
      7 months

      So somewhere in here we need some M. C. Escher stairs of AWS on the electrical grid on AWS on the electrical grid…

    • Knock_Knock_Lemmy_In@lemmy.world
      7 months

      And that software may or may not be very reliant on AWS

      Not. Electrical Scada systems are usually airgapped from the Internet.

    • Waraugh@lemmy.dbzer0.com
      7 months

      I love this because of how often a squirrel would take down our remote disaster recovery site.

      • Petter1@discuss.tchncs.de
        7 months

        Relax, it is USB C and only a small charger, this is only 20v max, and changes are high that CC lines are the ones severed first, resulting power supply to turn of Vbus or at least downgrade to V5SAFE.

  • Laser@feddit.org
    7 months

    In all seriousness though, the core of the technical stack has become very robust in my opinion (DNS being the exception). From a hobbyist’s perspective, things work much better than when the Web was still young. I can run multiple sites (some of them being what are today called apps) on a domain with subdomains, everything fast, HTTP3-capable, secured via valid free TLS certs, reverse proxied, all of that running on a system deployed in minutes…

    If you focus on the part of the Internet that you have control over, it’s a lot better than back in the simple days.

    • Petter1@discuss.tchncs.de
      7 months

      Imagine, we could kill all NAT/DNS/(reverse)proxy routing problems by adapting finally to IPv6

      • Laser@feddit.org
        7 months

        I don’t only run a reverse proxy because of having only a single public IPv4 address, but that probably is the best part

        In general, I’d say reverse proxies make things somewhat easier to manage, especially when it comes to TLS. No need for every service to integrate it.

  • Demdaru@lemmy.world
    7 months

    We arrivied thus at the funny moment where meme is accurate enough to be used for educational purposes.

    Look how little has to fail for whole web to decay, child xD

    • Galactose@sopuli.xyz
      7 months

      Or Fossil😅😅.

      For those people wondering, it’s an alternative to GIT created by SQLite devs. In fact their HomePage is actually a self-hosted Fossil repository

  • SCmSTR@lemmy.blahaj.zoneEnglish
    7 months

    Can someone please keep track of the evolutionary history of these? I wanna see a timeline.

  • Simulation6@sopuli.xyz
    7 months

    If you add infrastructure then you will need to add more transmission methods then a couple shark chewed undersea cables. Then you might as well add the millions of SAs, technicians, linemen (linepersons?), etc that install and maintain everything. Oh and I guess we would also need all the institutions and teachers that train all these techies.

  • manxu@piefed.socialEnglish
    7 months

    Can we please not make the layer above Electricity look like tombstones? I looked at “Linus Torvalds” and almost had a heart attack!