v2.1.0 - Stability & UI Improvements

What's Changed

Some pretty big changes that should improve speed & stability across the board, and also adds the option for notifications if one of your devices goes offline/unresponsive. This can be disabled in the settings for the integration.

Full changes:

• Concurrent monitoring: devices are now checked in parallel each cycle - one unresponsive device no longer blocks the rest.
• Unreachable device handling: devices that fail to respond are correctly marked as disconnected rather than "other content", avoiding states that required manual restarts. An optional HA notification fires when a device goes unreachable and clears automatically when it recovers.
• Volume fix: the legacy 1-10 YAML scale multiplier has been removed. The UI slider is 0-100 and values are passed directly to catt — a value of 5 now sets 5%, not 50%.
• Subprocess cleanup: all catt processes are tracked and properly terminated on timeout or unload, preventing zombie processes over time.
• UI configuration: replaced YAML-only setup with a full config flow. Global settings and per-device dashboards are now managed through Settings → Devices & Services.
• Config validation: dashboard URLs, device names, and IP addresses are validated on save with clear per-field error messages.
• UI copy: time fields now show "Start casting at / Stop casting at" with 24-hour format hints. Logging level includes guidance on when to use Debug.
• Minimum HA version set to 2025.2.0 in manifest.json to match the subentry config flow API requirement.
• Code quality: f-string logging converted to %-style throughout, pathlib replaces os.path in stats, docstrings added across all modules.