diff options
| author | Craig Jennings <c@cjennings.net> | 2026-05-19 13:20:05 -0500 |
|---|---|---|
| committer | Craig Jennings <c@cjennings.net> | 2026-05-19 13:20:05 -0500 |
| commit | 3044d4fbcf9077879784f433ae87ac452393e3d5 (patch) | |
| tree | daadeec38d42d055096a9a90d91d3894942f0859 /scripts/testing/lib | |
| parent | f91586248f467a2ba7a62f76caee1f40927655d9 (diff) | |
| download | archsetup-3044d4fbcf9077879784f433ae87ac452393e3d5.tar.gz archsetup-3044d4fbcf9077879784f433ae87ac452393e3d5.zip | |
chore: log AUR exit-0 fix verification and close the sub-task
Two updates in todo.org. I marked the two dconf entries in the `[#B] Fix install errors` parent body as RESOLVED in `dc06895`, since the system-wide dconf db migration verified clean in last night's VM run.
I converted the `*** TODO [#B] AUR exit-0 logged as error` sub-task into a dated event-log entry. The root cause was a bash if-compound semantics trap in `retry_install` (the if-compound exits 0 when no condition tested true, overwriting the eval's actual exit code that `last_exit_code=$?` was trying to capture). Today's VM run verified the fix: `mkinitcpio-firmware` and `tidaler` now report `error code: 1` instead of the misleading `error code: 0`.
Diffstat (limited to 'scripts/testing/lib')
0 files changed, 0 insertions, 0 deletions
