# Changelog

## Score

- CAI 73 → 74 (+1.6)

## Lenses

- Code Health 82 → 82 (+0.0)
- Architecture 92 → 91 (-0.7)
- Maturity 67 → 70 (+3.5)
- Readiness 72 → 72 (+0.0)
- Security 83 → 83 (+0.0)

## Resolved (4)

- Inconsistent naming for identical state update operations. While `force_data_update` is used in Cpu, Disk, and Temp widgets, the Memory and Network widgets use `init` for their lifecycle setup, and the Process widget uses `set_table_data` or `force_update_data` (note the typo/variation). Specifically, `CpuWidgetState.force_data_update` vs `DiskTableWidget.force_data_update` is consistent, but `ProcWidgetState.force_update_data` (missing 'd' in data) is a typo/inconsistency.
- Inconsistent naming for toggle operations. Some use `toggle_` (e.g., `toggle_mem_percentage`), while others might be expected to follow a stricter pattern like `set_mem_percentage` or `enable_mem_percentage` depending on whether they are state flips or setters. More critically, `ProcWidgetState.is_mem_percent()` and `ProcWidgetState.is_using_command()` are getters, but the corresponding setters/toggles are named `toggle_mem_percentage` and `toggle_command`. This mixes imperative toggle verbs with boolean state queries in a way that suggests `toggle` is the action, but `is_` is the check. This is minor but inconsistent with the `set_` pattern often seen in other widgets (e.g., `set_table_data`).
- While the method name `get_schema_names` is consistent across Disk, Process, and Temp columns, the return type and usage context differ slightly. `ProcColumn` also has `parse_column_name`, which is unique to it. This is not a direct inconsistency but highlights that `ProcColumn` has more functionality (parsing) than the others, which might be a design inconsistency if all column types should share a common interface.
- Workflow token grant is wider than its jobs use

## New (2)

- Orphaned file (src/collection/disks/zfs_io_counters.rs)
- Workflow token grant is wider than its jobs use (.github/workflows/build_releases.yml)