Josh Pencheon - Software Engineer

Niri releases are built with a default log level of debug, which can spoil an otherwise silent boot.

Recently, I noted btop was showing some odd stats for a ZFS dataset; it was showing as 100% free, but with an unexpectedly small total capacity: ├─main-pool──────────────65,6─GiB─┤ │ IO% ........................... │ │ Used: 0% ■ 128 KiB │ │…

I’ve been using Ollama for the past year or so to experiment with open-source large language models, and have most recently been trying out OpenAI’s gpt-oss 20 billion parameter model.

The Arch User Repository (AUR) provides thousands of user-submitted packages, but each package in the AUR is just a git repository containing at minimum a PKGBUILD recipe.

Sanoid is a popular choice for maintaining a history of a ZFS dataset via snapshots.

For licencing reasons, ZFS is not able to integrated into the mainline linux kernel.

I’ve got an ancient Nvidia GT210, a GPU that was old even when it was new.

I previously wrote about enabling encryption for a ZFS dataset, but didn’t cover how to go about mounting such a dataset on boot.

I’ve got an SMB share that’s serving files stored in ZFS.

Suppose we have an existing ZFS dataset that was created without encryption, and we’d now like to encrypt it.