• Zachariah@lemmy.world
    link
    fedilink
    English
    arrow-up
    7
    ·
    1 day ago

    yes, please

    90% of my large data needs are for long-term storage, and I’d prefer fast retrieval, but if it was just not too slow, I could put a caching system in place between the data and where it’s used.

    • MalReynolds@slrpnk.net
      link
      fedilink
      English
      arrow-up
      6
      ·
      edit-2
      1 day ago

      For consumer accessible long term storage…

      Large amounts (many TB), look at LTO tape, 10-30+ yrs depending on storage, but you have to keep drives alive. Second hand LTO5 (1.5 TB per tape) run around $300 per drive (you’ll need an SAS card as well), tapes are pretty cheap and last many rewrites. You’ll want 2 drives of the same manufacturer and type (that read each others tapes, finicky) and perhaps a third that doesn’t work for spares.

      Smaller (100GB) amounts Metal, probably good for 100 yrs, not organic dye, that shit self destructs in 5 years, Bluray-XL disks. Good for photos, text and probably audio depending on how much. Unfortunately there’s only one disk manufacturer left (so monopolistically expensive) and no drive manufacturers currently. So 2+1 drive protocol applies.

      Long term data storage sucks (yes I did the research). There’s a couple of techs like the article describes, I like the quartz one best, millions of years, but the chance of seeing them mass produced is low, built in obsolescence baby, it’s an income stream.

      For mere mortals a 3-2-1 backup strategy using spinning rust HDD is still most practical, and still expensive and cumbersome. If you’re willing to trust online backups (encrypt at home with like cryptomator) are not uneconomic, given the easiness, but you may want two for safety.

      Editorial : You’ll own nothing and hate it.

      • Zachariah@lemmy.world
        link
        fedilink
        English
        arrow-up
        5
        ·
        24 hours ago

        As long as there was a small business solution, I’d be okay. Maybe something like sending the archive to the writing company who sends back the written crystals, and then buying/leasing reader hardware for onsite data access. Probably also another copy of the crystals off-site and offline.

        Doesn’t need to be a consumer product, as long as it’s not only for enterprise.

        • MalReynolds@slrpnk.net
          link
          fedilink
          English
          arrow-up
          2
          ·
          edit-2
          23 hours ago

          Yeah, that’d be grand, and probably profitable. If you hear about one, let me know. Won’t hold my breath, late stage capitalism, line must go up this quarter…

          Small business is consumer to them, and they’re the only ones who can make devices like this.