Organize documentation by audience: packaged user guides in docs/operations
(including Devices, Tape scripts, Auto read order, Jobs, offline migration and
tar recovery), contributor procedures in docs/development, maintainer release
procedures in docs/release, current behavior in docs/architecture (now with
separate Tape and read-planning documents) and evidence in docs/research. Each
rule keeps one owner and other pages link to it; the document checker, the
packaged guide allowlist and their tests follow the new layout. The documents
land together because the new guides and index link to each other.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Compose tar FORMAT/APPEND, Restore, Scan and Verify through complete Media sessions and the configured encrypt adapter. Publish only complete members at normal or capacity/EW completion, maintain position maps, apply Logical or Auto read order, record passive read learning observations in each Job and refit learned read parameters after drive release. LTFS keeps its adapters and gains the shared read order.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Preserve the existing Job lifecycle while keeping timing, cancellation and independent pipeline errors observable.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Use the common temporary root for Library, Restore probes, Preview and script output. Preserve active ownership, ordinary outputs and retryable cleanup failures.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Upgrade ACP to v0.2.6 for continuous linear-target copying and ordered native filename completion.
Preserve Archive capacity diagnostics, use indexed Restore checks, and maintain real Archive/Restore physical acceptance with reproducible capacity research.
Samuel CuiandClaude Opus 5.5