There's a quiet rumor going around in the LTO community: LTFS is dying. Vendors are abandoning it. Maybe TAR isn't so bad after all.

It's worth unpacking, because the rumor is wrong — but the frustration behind it is real.

Where it comes from

The most visible piece of evidence is IBM's announcement that it'd sunset LTFS on Windows and Mac. IBM is the biggest player in the space, so that got people's attention back in 2023. Yes, that long ago. It was never officially communicated besides a mention in a changelog, and that type of underplaying tends to spread rumors: "IBM sunsets LTFS on Mac and Windows" became shorthand for "LTFS is on its way out." The thing is, there is not much to update about LTFS anyway - most of the work is on the firmware end of the equation, and that's where IBM is to this day very active.

HP hasn't helped matters. Their last two LTFS releases were in November 2023 and October 2025 — two years apart — and they don't sell or support LTO-10 yet. Quantum, meanwhile, stopped maintaining its own LTFS implementation entirely and consolidated onto IBM's.

So: one active vendor, one stagnant one, one that's outsourced the problem. Not exactly a picture of a thriving ecosystem.

But here's the thing — this is a vendor picture, not a format picture.

What LTFS actually is

LTFS is an open standard. It turns a tape into something that behaves like a filesystem: browse it, copy to it, read it back. More importantly, LTFS tapes are self-describing — the index lives on the tape itself. You don't need to know anything about the system that wrote it to read it. Mount a tape years from now with any LTFS-compatible software, and you'll see exactly what's on it.

That's not a small thing. For long-term archiving, it's the whole game.

The alternatives don't come close. Proprie_tar_y formats tie you to a vendor that may or may not exist in ten years. TAR has no index. To know what's on a TAR tape, you either need a just-as-proprietary database, or you have to read it from the start, sequentially, and if you don't know the block size it was written with, you can't read it at all. That's a real problem when the original machine is long gone.

The spec outlives the vendors

Because LTFS is open, no single company can kill it. IBM could stop shipping LTFS software tomorrow, and the format would survive. The specification is public. Anyone can implement it. Any tool that reads LTFS today will still be able to read LTFS tapes written years from now.

This is the core answer to the death narrative. "LTFS is dying" really means "some vendors are being flaky about their implementations." Fair point — but it's a different problem, and it has a different solution.

The real issue: vendor implementations don't play nicely together

Beyond stale updates, there's a more practical problem: IBM and HP LTFS implementations don't interoperate. Mix vendors, and you'll run into friction fast — effectively creating hardware lock-in almost by accident, which defeats the whole point of an open format.

We've addressed this head-on in Canister 26.1 for Mac. It ships with a new LTFS implementation that supports mixed-vendor setups — HP, Quantum, and IBM drives working side by side — stays close to the spec rather than any particular vendor's interpretation of it, and isn't dependent on any vendor's release schedule. No more waiting for HP to catch up, or worrying about what IBM decides to bundle or unbundle next.

Day to day, that means you can freely mix IBM, HP, and Quantum drives across LTO-5 to LTO-9, with IBM LTO-10 support included too.

Tandberg drives remain an outlier and still require the legacy HP LTFS driver. If you’re using Tandberg hardware, our support team can help you get set up.

It also resolves a stability issue between aging vendor LTFS implementations and macFUSE. Not the flashiest feature in the release, but one you’ll appreciate every day.

More to come

LTFS is the right answer for new tape archives — we're more convinced of that than ever. And we know that not everyone is starting from a clean slate. If you're on an older format and wondering how to get to somewhere more sustainable, that's something we're thinking about too.

We're not done yet. Watch this space - our engineering team is busy behind the curtain.