Two targets have shipped. The goal is a family of
pure-Go, cgo-free reimplementations of well-known bootloaders and boot managers —
readable, testable Go in place of the usual C and assembly.
grub and
systemd-boot are real,
tested modules in production use (both consumed by
go-diskimages/diskimage) โ
neither has cut a tagged release yet. The remaining four targets below are still a roadmap of
intent, and the status badges are honest about where each one stands.
Targets
pure-Go reimplementation of GNU GRUB
GRUB administration toolkit: opens a GPT disk image, mounts the ESP and a Linux /boot (ext4/btrfs) read-write, reads/generates/patches grub.cfg, registers a UEFI boot entry, and optionally measures boot into TPM PCRs. 93.9% test coverage; no tagged release yet.
pure-Go reimplementation of systemd-boot (sd-boot)
Boot Loader Specification reader: GPT-locates the ESP, parses loader.conf + BLS entries, reads/writes UEFI BootOrder/BootNext, and measures the selected kernel/initrd/cmdline into TPM PCRs. 92.9% test coverage; no tagged release yet.
pure-Go reimplementation of Limine
Modern x86-64 / aarch64 / riscv64 boot protocol. Compact spec makes it a natural early pure-Go implementation.
pure-Go reimplementation of Syslinux / ISOLINUX / PXELINUX
Lightweight loaders for FAT, ISO9660 and PXE network boot.
pure-Go reimplementation of rEFInd
Graphical UEFI boot manager that auto-discovers EFI boot loaders.
pure-Go reimplementation of Das U-Boot
Embedded / SoC boot firmware across many architectures.
Every module is (or will be) pure Go with CGO_ENABLED=0 and held to
the same engineering bar as its sibling organisations: tested, honest about what works, and
BSD-3-Clause throughout. For the four planned targets, treat the cards above as direction
rather than capability. The brand repo holds
the org's logos.