v1.8.4: a broken cache store no longer breaks migrate

Fixed

• A cache store that was configured but not usable could fail php artisan migrate and take the dashboard down. CACHE_STORE=database is Laravel's own default and its cache table is not always there yet (a partial migrate --path= batch, a migrate:rollback past the table, a secondary connection), and a Redis or Memcached server can simply be unreachable. Two things went wrong. The rebuild that runs after a migration threw after every migration had already committed, so Artisan reported a failure on work that had succeeded, which can break a deploy step running migrate --force. And the schema endpoint returned HTTP 500, so a first visit to /truss on a fresh app showed "Could not load schema" over an empty diagram. Truss now treats an unreachable cache store the way it already treats an unreachable database or disk: the structure is read live and simply not cached, the dashboard says why, and the migration listener never throws. truss:show, truss:doctor, truss:diff and truss:export print a notice and keep working, with truss:export writing it to stderr so a piped export stays clean and --check still fails only on real drift. truss:rebuild is the deliberate exception: it reports a failed write and exits non-zero, because storing the snapshot is its only job. No configuration changes, and nothing to do on upgrade. Reported by @HafizMMoaz.

Upgrade: composer update albertoarena/laravel-truss. Nothing to change in your config, and no new keys.