Skip to content

Commit 186a8c5

Browse files
RFingAdamclaude
andcommitted
docs: add CHANGELOG + finish reference glossary (v5.1)
- Add CHANGELOG.md (Keep-a-Changelog) seeded from RELEASE_NOTES.md, surfaced in the docs nav via docs/changelog.md (include-markdown). Closes #21. - Glossary: add the remaining emitted-metric terms (EIRP, Sidelobe, Group delay, Cal drift). The glossary/measurement-types/mcp-overview pages were substantially rebuilt in v5.0.0; this finishes the metric coverage. Closes #20. - mkdocs --strict builds clean. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
1 parent 3493c3b commit 186a8c5

4 files changed

Lines changed: 66 additions & 0 deletions

File tree

CHANGELOG.md

Lines changed: 58 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,58 @@
1+
# Changelog
2+
3+
All notable changes to RFlect are documented here. The format follows
4+
[Keep a Changelog](https://keepachangelog.com/en/1.1.0/) and the project adheres
5+
to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
6+
7+
Full, detailed per-release notes live in
8+
[RELEASE_NOTES.md](https://github.com/RFingAdam/RFlect/blob/main/RELEASE_NOTES.md); this file is the concise summary.
9+
10+
## [Unreleased] — v5.1 (Correctness & quick wins)
11+
12+
### Added
13+
- Golden-reference regression tests locking the TRP/gain integration core
14+
(isotropic EIRP → itself; upper hemisphere → −3.01 dB).
15+
- `compute_group_delay_dispersion()` — per-frequency group-delay variance/std
16+
across theta cuts (replaces the old dead TODO block).
17+
- Test coverage for previously-untested modules: `plotting.py`, `save.py`,
18+
`groupdelay.py`, `plot_group_delay_cst.py`, `uwb_plotting.py`.
19+
20+
### Changed
21+
- `cal_drift_report` MCP tool now returns `{output_path, format}` / `{error}`
22+
and never raises to the client.
23+
- Slimmer install: `openai` moved to the `[ai]` extra; `keyring`/`cryptography`
24+
dropped from core deps (unused since the v5.0.0 AI removal); `pyinstaller`
25+
removed from the runtime requirements; numpy ceiling raised to `<3.0.0`.
26+
- GUI settings save/load and data-processing failures are now surfaced
27+
(stderr / log) instead of being silently swallowed.
28+
29+
### Docs
30+
- This changelog; expanded reference docs.
31+
32+
## [5.0.0] — 2026-05-28
33+
34+
**Zero-dependency, MCP-first relaunch.** Removed the entire in-app AI/LLM stack
35+
(chat, AI report generation, provider abstraction, API-key store) — RFlect makes
36+
no LLM/API calls and needs no key or subscription. `generate_report` is
37+
deterministic by default with an optional agent-authored `narrative`. Added 6
38+
MCP tools (compare_antennas, analyze_s11, analyze_group_delay,
39+
estimate_link_budget, analyze_mimo_diversity, generate_active_cal); MCP tool
40+
count 35 → 41. Renamed `ai_analysis.py``analysis_engine.py`.
41+
42+
## [4.3.0] — 2026-05-28
43+
44+
Multi-angle iperf throughput-validation MCP tool (`analyze_iperf_angle_sweep`).
45+
46+
## [4.2.0] — 2026-05-12
47+
48+
`process_folder` single-call folder-orchestration MCP tool; MkDocs Material docs
49+
site.
50+
51+
## [4.1.9] — 2026-04-14
52+
53+
Patch: restored the Active Chamber Calibration routine.
54+
55+
---
56+
57+
Earlier releases (4.1.8 and prior) are detailed in
58+
[RELEASE_NOTES.md](https://github.com/RFingAdam/RFlect/blob/main/RELEASE_NOTES.md).

docs/changelog.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
{%
2+
include-markdown "../CHANGELOG.md"
3+
%}

docs/reference/glossary.md

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -25,3 +25,7 @@
2525
| **VSWR** | Voltage Standing Wave Ratio. Derived from S11 magnitude. 1.0 = perfect match; 2.0 ≈ -9.5 dB return loss. |
2626
| **WTL** | Wireless Telecom Lab — the chamber-output format RFlect's parsers were built against (V5.02 / V5.03). |
2727
| **XPD** | Cross-Polarization Discrimination — co-pol field divided by cross-pol field, in dB. Uses 20 log, not 10 log. |
28+
| **EIRP** | Effective Isotropic Radiated Power — power radiated in a given direction relative to an isotropic source. RFlect's per-angle chamber input; TRP is its sphere average (the 1/4π in `calculate_trp`). |
29+
| **Sidelobe** | A radiation lobe other than the main beam; sidelobe level (SLL) is its peak relative to the main-lobe peak, in dB. |
30+
| **Group delay** | −dφ/dω, the negative derivative of transmission phase w.r.t. angular frequency (s); flatness across the band matters for wideband/UWB pulse fidelity. |
31+
| **Cal drift** | Change in a chamber's path-loss calibration factors over time; RFlect tracks runs across epochs and flags drift beyond a threshold. |

mkdocs.yml

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -110,4 +110,5 @@ nav:
110110
- CLI: reference/cli.md
111111
- API: reference/api.md
112112
- Glossary: reference/glossary.md
113+
- Changelog: changelog.md
113114
- Contributing: contributing.md

0 commit comments

Comments
 (0)