docs: sweep changelog framing out of current-state docs

[josephnef]

Follow-up from the PR #407 review: the two pre-existing "is now" lines flagged there (docs/rtl8733b.md fast-retune, src/rtl8733b/CLAUDE.md RF-domain), plus a repo-wide sweep for the same pattern. 13 spots across 7 files rephrase transition-narration into current-state — "is now ported" → "is ported", "the value the descriptors used to hardcode" → "the vendor descriptor default", "did historically" → "failure modes it avoids", the (the old text format used -) aside in logging.md, and the "earlier structural skip" narration in 8822e-quirks.md. Every fact survives; only the history framing goes — git is the changelog.

Deliberately untouched: article-style lesson narratives whose past-tense content is measured evidence (docs/fhss.md's SDR-matcher pitfalls with their regression guard, the −64 qdB clamp story and "earlier cut cost ~9 dB" adversarial notes in the rtl8733b docs), vendor-lineage references ("the old trees" = vendor source trees), and protocol-time uses ("the old permutation").

🤖 Generated with Claude Code

https://claude.ai/code/session_01RcEcJciyf9zQ9GLZJ7jYXx

[qodo-free-for-open-source-projects]

PR Summary by Qodo

Remove changelog framing from current-state documentation

📝 Documentation 🕐 Less than 10 minutes

AI Description

• Rephrases transition-oriented language as descriptions of current behavior.
• Preserves technical findings, measurements, caveats, and source references.
• Keeps historical context in Git while retaining evidence-based narratives.

High-Level Assessment

Directly rephrasing the affected passages is the optimal approach. It keeps current-state documentation concise while preserving technical facts and leaving transition history in version control; introducing a changelog section or retaining inline historical narration would work against the stated documentation convention.

Files changed (7) +20 / -20

Documentation (7) +20 / -20

8822e-quirks.mdDescribe path-B TXAGC findings without historical narration +7/-7

Describe path-B TXAGC findings without historical narration

• Reframes the apparent RX-deafening observation and invalid aggregate-frame checks as present-tense failure modes. The register exoneration, DPDT routing issue, and requirement for per-chain RSSI remain intact.

docs/8822e-quirks.md

adaptive-link-validation.mdRemove transition framing from residual-loss finding +1/-1

Remove transition framing from residual-loss finding

• States that the validation observes real end-to-end loss without characterizing the measurement as newly honest.

docs/adaptive-link-validation.md

logging.mdDocument unavailable fields using current JSON conventions +1/-1

Document unavailable fields using current JSON conventions

• Removes the comparison with the former text format while retaining JSON null as the convention for unavailable chip data.

docs/logging.md

pseudo-preamble-puncturing.mdPresent puncturing capabilities as current-state behavior +2/-2

Present puncturing capabilities as current-state behavior

• Removes “now” language from the implemented approximation and describes the RX CSI mask knob as exposed. Capability claims and measurement conclusions are unchanged.

docs/pseudo-preamble-puncturing.md

rtl8733b.mdState fast-retune support without temporal framing +1/-1

State fast-retune support without temporal framing

• Changes “is now ported” to “is ported” while preserving the independent validation results and measured contract.

docs/rtl8733b.md

scheduled-mac.mdReframe retry behavior and validation as current state +6/-6

Reframe retry behavior and validation as current state

• Replaces historical descriptor, retry-ladder, and check-script narration with current defaults, avoided failure modes, and present validation requirements. Retry semantics, examples, and warnings remain unchanged.

docs/scheduled-mac.md

CLAUDE.mdRemove historical qualifiers from RTL8733B guidance +2/-2

Remove historical qualifiers from RTL8733B guidance

• Describes Jaguar1 fallback and RF-domain measurements directly, without framing them as historical or newly available. Safety guidance and measurement provenance remain intact.

src/rtl8733b/CLAUDE.md

[qodo-free-for-open-source-projects]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0) 🎨 UX issues (0) 🔗 Cross-repo conflicts (0) 📜 Skill insights (0)

1. Two checks become singular ✓ Resolved 🐞 Bug ⚙ Maintainability

Description

The rewrite makes singular “a recheck … is blind” stand for two separate checks—the desense verdict
and full-duplex proof—creating a number mismatch and obscuring that both RX-health conclusions rely
on aggregate frame counts. This loses the original sentence’s explicit assertion that both checks
are blind to a dead chain B.

Code

docs/8822e-quirks.md[R70-72]

+and a recheck counting TOTAL frames, which chain A dominates —
`tests/eu_41e8_desense_recheck.sh`'s −2% "noise" verdict and the 24k-frame
-full-duplex proof (`tests/eu_fullduplex_pathb_check.sh`) were both blind to a
+full-duplex proof (`tests/eu_fullduplex_pathb_check.sh`) — is blind to a

Evidence

The rewritten documentation names the verdict from eu_41e8_desense_recheck.sh and the proof from
eu_fullduplex_pathb_check.sh inside an appositive for singular “a recheck,” followed by singular
“is blind.” The first script counts only total rx.frame events, and the second assesses aggregate
EU receive counts, proving there are two distinct validations to which the RX-chain limitation
applies.

docs/8822e-quirks.md[65-73]
tests/eu_41e8_desense_recheck.sh[40-41]
tests/eu_fullduplex_pathb_check.sh[39-47]
tests/eu_fullduplex_pathb_check.sh[53-59]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
Rewrite the sentence so it clearly states, in current-state language, that both referenced checks count aggregate frames and are blind to a dead RX chain B. Avoid the singular “a recheck … is blind” construction around two separate checks.
## Issue Context
`tests/eu_41e8_desense_recheck.sh` counts total `rx.frame` events, while `tests/eu_fullduplex_pathb_check.sh` evaluates aggregate EU receive counts. The documentation should explicitly apply the limitation to both checks without restoring historical narration.
## Fix Focus Areas
- docs/8822e-quirks.md[65-73]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

Tip of the day

💡 Did you know, you can describe a rule in plain language on the Rules page and Qodo drafts it for you

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗