Fediverse is great, I can post stuff here now as it’s down. xD

Edit: damn, they’re already back up this quick

Here was the archive: https://archive.ph/RAZ5Q

    • Rai@lemmy.dbzer0.com
      32·
      20 hours ago

      I had BOTH of my memory modules die simultaneously in my first beefy computer I built. I was very lucky that OCZ had a lifetime warranty—they didn’t even ask for the memtest results, they just sent me two new sticks.

      Then both of my (SLI) video cards died… those too were in warranty, but I began to wonder if my computer was cursed.

        • Rai@lemmy.dbzer0.com
          5·
          18 hours ago

          Yknow that’s what I thought! But after replacing memory and GPU, everything worked great for a year or two until I sold it! I did bring it into my work (I used to work computer repair) and used the shop’s PSU tester which passed it…. I think it was just a fluke! I do know the cards I bought (2 x 7900GT) had a bit of a notoriety for failing, but I do not think the RAM had the same issue.

            • Rai@lemmy.dbzer0.com
              2·
              17 hours ago

              The friend I sold it to used the PSU and mobo for another couple years before upgrading, so I think you’re right! Just bad luck on my part hahaha. Thank goodness for good RMA policies!

    • sobchak@programming.dev
      3·
      13 hours ago

      Never had one fail. Bought a faulty one once. Took me months to figure out why I was having data corruption issues. Thought it was one of the old HDDs I was using in a ZRAID array, so I would swap one out and try again until I’d eventually get a corruption error again. Finally found the issue after about 10 minutes of running a memory test of a bootable USB.

      • tempest@lemmy.ca
        3·
        15 hours ago

        It’s still not that common to have one die. Servers have ECC memory which can get you by longer even as a DIMM starts to die. It’s also rarely run at speeds consumer RAM is run at.

        The thing with server memory is when you get to point where you have hundreds of servers and each one has 12 to 24 DIMMs the chance you have a bad one somewhere increases.

        • dan@upvote.au
          4·
          13 hours ago

          Some servers have ECC. If you get a cheap one (like a Hetzner auction server), it’s less likely to have ECC. ECC protects against bitflips, but it won’t help if the RAM is starting to die. ECC isn’t magic - it just has an extra 8 bits of parity data per 64 bits of data. It still uses the same type of RAM chips.

        • The Quuuuuill@slrpnk.netEnglish
          2·
          15 hours ago

          nah for sure, but i’d give a bet ruud was running on consumer hardware which has far less fault tolerance built in

    • bonenode@piefed.socialEnglish
      10·
      20 hours ago

      Worst time for faulty memory. Hope someone had a stick lying around so they don’t have to pay outrageuous prices.

    • Egonallanon@feddit.uk
      6·
      20 hours ago

      At least anecdotally I find them to be the third most common thing to fail after storage drives and fans. Though yeah still super rare.

    • dan@upvote.au
      3·
      14 hours ago

      If you want to try alternate UIs, you might be interested in trying Photon and Alexandrite.

      • FartsWithAnAccent@fedia.io
        1·
        11 hours ago

        I’ve tried a few variants out and they all have their own benefits and drawbacks, but I kinda settled on Mbin with Interstellar on mobile (though the website works fine in Firefox mobile too).

    • tempest@lemmy.ca
      2·
      16 hours ago

      So does .CA

      It’s one of the advantages of the being a lot of sys admins and Linux users in this community.

      You can run Lemmy pretty comfortably on used servers.

      • dan@upvote.au
        1·
        13 hours ago

        I’m running mine on a ~$33/year VPS at GreenCloudVPS. It’s a small instance (just me) but it federates with all the major instances which means it still does a bunch of work (since it has to handle incoming posts and comments from federated servers). It’s a decently powerful VPS with an AMD EPYC Milan CPU, 10GB RAM, and 100GB NVMe storage.

        For a medium-sized instance, I imagine you could get pretty far with a single <$100/month dedicated server from Hetzner or a similar provider.