aurupdater

botan2

Category: local-install, plain unmodified AUR rebuild (no local edits). Rebuilt via the standard yay -Qu-triggered workflow — found outdated on euronuc, not tracked under arch/, rebuilt+published for archlinuxaba per [[feedback_local_install_rebuilds]].

VCS-style pkgver

Not a real upstream release cadence: pkgver() runs git describe on a pinned _commit= of github.com/randombit/botan.git, giving versions like 2.19.5.r22.g409ab81. AUR’s PKGBUILD simply bumps _commit periodically to a newer upstream commit — “outdated” here just means the AUR maintainer advanced the pin, not that there’s a tagged release. Expect this package to show up in yay -Qu again whenever the AUR maintainer next moves _commit forward; that’s normal and each rebuild is a fresh git clone per the standard workflow (step 2), not something to special-case.

2026-09-21 build

First publish. x86_64 only (arch=(x86_64)), no validpgpkeys. Build went smoothly — the only friction was operational: the orchestrating repo_release.sh process got killed by an earlier (mistaken) kill during this session while its child repo_build.sh kept running as an orphan to completion, so steps 2/3 (sign) and 3/3 (publish) never ran automatically. Recovered by running repo_sign.sh and repo_publish.sh directly by hand against the already-built .pkg.tar.zst files once the build finished, then cleaning up local artifacts manually (same end state as if repo_release.sh had completed uninterrupted). No package-level issue here — just don’t kill a repo_release.sh orchestrator PID once its child build has started; if you need to stop a run, kill the whole process tree, not just the parent.

namcap flagged (informational, not blocking): BSD isn’t a valid SPDX identifier, docs are 69% of package size, and unreproducible sphinx .doctree cache files got packaged. None addressed — this is a plain rebuild of AUR’s own PKGBUILD, no local patching intended.