hfsfuse

AUR package: FUSE filesystem for HFS+ (upstream: https://github.com/0x09/hfsfuse).

Quirk: this AUR git repo is a reused/adopted slot

The AUR git repo ssh://aur@aur.archlinux.org/hfsfuse.git previously hosted a different package, bindfs (see git log in this submodule – history before commit c7b2a37 "initial revision" is entirely bindfs commits up to 1.18.4). On 2026-07-31 the maintainer replaced the content wholesale with the real hfsfuse PKGBUILD, starting fresh at pkgver=0.425, but left .nvchecker.toml untouched – it still pointed at bindfs.org and checked bindfs’s version.

Effect: check_for_updates_maintained.sh flagged UPD maintained/hfsfuse 0.425 1.18.4, which looked like a real (large) version bump but was actually nvchecker silently checking the wrong upstream. Always cross-check a suspiciously large/unexpected version jump against the PKGBUILD’s actual url= before trusting the STATES.txt line – see also [[feedback_verify_aur_flags_before_acting]] for the same class of “don’t blindly trust the automated signal” issue.

Fixed .nvchecker.toml to:

[hfsfuse]
source = "github"
github = "0x09/hfsfuse"
use_latest_release = true

2026-08-15: 0.425 -> 0.435

Real upstream update once nvchecker was pointed correctly. Plain tarball source from GitHub tags (refs/tags/${pkgver}.tar.gz), no patches needed. Build via make WITH_UTF8PROC=local WITH_ZLIB=local worked without changes. namcap notes (pre-existing, not addressed): BSD isn’t a strict SPDX identifier, and libutf8proc/glibc dependency notes – cosmetic, left as maintainer already had them.

2026-08-21: 0.435 -> 0.444

Routine tag bump (tarball confirmed available before touching PKGBUILD). updpkgsums + .SRCINFO regen, built clean via extra-x86_64-build, same pre-existing namcap notes as before (not addressed, cosmetic). Signed + published via repo_release.sh maintained/hfsfuse; old 0.435 build archived automatically.