• Require SnapRAID 15.0rc2 or newer.
• Added a "degraded" health state for disks and the array when a
  filesystem UUID mismatch is detected against the content file.
  This distinguishes missing mountpoints or unmounted disks from
  hardware failures.
• Added a "check_updates" option to periodically check for new daemon
  releases on GitHub. The WebUI now displays the update status under
  the Configuration card on the Dashboard and shows a dedicated
  "NEW VERSION AVAILABLE" badge in the System Idle header when a new
  version is available.
• Introduced "Brutale", a high-contrast Neo-Brutalist WebUI. Loud,
  blocky, and completely unapologetic—sunglasses not included.
• Added a "net_auth_credential" option to protect the REST API and
  web dashboard via HTTP Basic Authentication. The credentials are
  securely verified using Argon2id ($m=65536, t=3, p=1$).
  Also added a -g, --gen-auth command-line utility flag to generate
  a valid configuration line from a username and password.
  Note: Since Basic Authentication transmits credentials in plaintext
  (Base64), it is not secure over unencrypted HTTP. Users must deploy
  an HTTPS encryption layer or restrict access to a trusted, secure
  network.
• Added support for managing multiple independent SnapRAID arrays:
  • On Linux/systemd, daemon instances can be managed individually
    using the template unit snapraidd@.service.
  • On Windows, multiple service instances can be registered and
    managed using the --service-install option with the -i flag.
• Schedule a report notification on degradations of counting/critical
  raw SMART attributes and tracked error counters. Normalized value
  degradations of critical prefail attributes also generate warning
  notifications, while normalized changes of non-critical prefail
  attributes are tracked as informational history.
  Historical tracking for normalized values (with highest/worst
  baselines and threshold values) is exported via REST API/YAML and
  integrated into status reports.
• Implemented a bracketed privileges system on Linux to run the daemon
  as an unprivileged user (e.g., "nobody") by default and temporarily
  escalate to root only when performing administrative operations.
• Added a "hook_docker_pause" option to specify a comma-separated list
  of Docker containers to pause during tasks that access the array.
  Containers are paused immediately before running the task (or the
  pre hook script) and unpaused immediately afterward.
• Added a "notify_start" configuration option to trigger a notification
  command at the start of high-level tasks like maintenance.
• Added a "sys_shutdown_on" option to automatically trigger a system
  shutdown based on a comma-separated list of events ("maintenance",
  "prefail", or "failing").
• Added a "smartignore" configuration option to allow ignoring specific
  SMART attributes (e.g., Spin_Up_Time, Raw_Read_Error_Rate) for one or
  more disks. Ignored attribute changes do not generate warning
  notifications and are omitted from status reports, while the attributes
  remain available and marked as ignored in the REST API.
• Export temperature warning/critical thresholds and status fields in
  the REST API, and integrate them into the web dashboards to
  automatically color-code drives (supporting higher limits for SSD,
  NVMe devices).
• Allow separate spindown idle timeouts for data and parity disks by
  specifying two comma-separated values in "spindown_idle_minutes"
  (e.g., 15, 30). If only one value is specified, it applies to both.
  This introduces the new REST API "/snapraid/v2/config" endpoint where
  the spindown configuration is returned and set as a string, while
  retaining backward compatibility with the legacy
  "/snapraid/v1/config" endpoint (which only returns the data disk
  timeout as an integer).
• Added non-resetting lifetime tracking for I/O and silent data errors
  on disks using snapraid_tracked structures ("error_io" and
  "error_data"), which include historical baselines and timestamps.
  This introduces the new REST API "/snapraid/v2/disks" endpoint where
  lifetime tracked error counters ("error_io" and "error_data") and
  active session counters ("transient_error_io" and
  "transient_error_data") are exported, while preserving backward
  compatibility with the "/snapraid/v1/disks" endpoint (which continues
  to export transient session errors under
  "error_io" and "error_data").
• Added the "/snapraid/v1/clear" REST API endpoint and a "Clear" button
  in the WebUI Tasks view to cancel all pending tasks waiting in the
  execution queue without affecting any currently running task.
• Added support for saving task log files in gzip (.gz) format,
  controlled by the "sys_log_compression" configuration option.
• Added %h (high-level task name) and %c (exit code of the task)
  placeholder tags to result notifications.
• Added environment variables to the hook script executions, exposing
  execution details (such as command name, task number, log file,
  config paths) for the task-begin event, and adding exit codes,
  signals, duration, error counters, array health, and diff statistics
  for the task-end and task-error events.
• Deprecated the "sys_engine" configuration option and replaced it with
  the new "./configure --with-snapraid=PATH" option to allow hardcoding
  the utility path at build time.
  Similarly added new "--with-docker=PATH" and "--with-curl=PATH"
  options to allow hardcoding the curl and docker utility paths at
  build time. The default runtime discovery behavior remains unchanged
  if these options are omitted.
  Primarily intended for NixOS and other non-FHS systems.
• Sort diff and fix lists by change type to prioritize the most
  relevant entries first.
• Include the system hostname at the beginning of the status report
  before the health status.