Skip to content

Plan.md

Latest commit

 

History

History
221 lines (156 loc) · 13.5 KB

Plan.md

File metadata and controls

221 lines (156 loc) · 13.5 KB
Raw

Implementation Plan — Missing CAT Controls

Ordered by priority. Estimates are for backend + frontend together.

Phase 1 — Band & Step Navigation (Priority 2, ~1 day)

Commands: BU, BD, UP, DN, TS

All five belong together — UP/DN are useless without TS, and BU/BD live in the same "navigation" mental model.

Backend — CatController.cs

  • POST /api/cat/band/up → sends BU;, then re-reads FA;/FB; to update displayed frequency
  • POST /api/cat/band/down → same with BD;
  • POST /api/cat/vfo/step/up → sends UP;, then re-reads active VFO frequency
  • POST /api/cat/vfo/step/down → sends DN;
  • POST /api/cat/tuningstep/{receiver} body { "step": 4 } → sends TS{vfo}{step:D3}0;

Backend — RadioStateService.cs

  • Add TuningStepA and TuningStepB int properties (default 4 = 1 kHz)

Backend — CatMessageDispatcher.cs

  • Add case "TS": — parse TS{vfo}{step:D3}0 → update TuningStepA / TuningStepB

Backend — CatMultiplexerService.cs

  • Add TS00000; and TS10000; to GetInitialValues() to read both VFO steps on connect

Frontend — Index.cshtml

  • Band ◄ ► buttons in the toolbar alongside the existing VFO-B/Swap/Split buttons
  • Tuning step dropdown per VFO in the frequency display row (below the digit display): 1Hz 10Hz 100Hz 500Hz 1k 2.5k 5k 10k 100k 500k 1M — same row as the keyboard icon button
  • ▲ ▼ step buttons flanking the tuning step dropdown, calling step up/down

Frontend — site.js

  • Click handlers for band up/down that POST then re-render frequency
  • Click handlers for step up/down
  • SignalR handler for TuningStepA / TuningStepB updates (keeps step dropdown in sync if changed from front panel)

Phase 2 — Anti-VOX Wire-up (Priority 2, ~1 hour)

Command: VA

The state property, UI slider, and API endpoint already exist. The endpoint stores locally only and never sends a CAT command. This is a bug fix, not a new feature.

Backend — CatMessageDispatcher.cs

  • Add case "VA": → parse VA0{000-100} → update RadioStateService.AntiVoxGain

Backend — CatController.cs — SetAntiVoxGain

  • Replace _radioStateService.AntiVoxGain = request.Gain + comment with await SendCommand("VA0" + request.Gain.ToString("D3") + ";", true), then read-back VA; and update state from the response

Backend — CatMultiplexerService.cs

  • Add await SendCommand("VA;", true); to GetInitialValues() so the real radio value is read on connect

No UI changes needed — the slider is already there.

Phase 3 — Keyer Type + TX Narrow (Priority 3, ~half day)

Commands: KI, XN

Two independent small features, grouped because they touch the same files.

KI — Keyer Type

  • RadioStateService.cs: Add KeyerType int property (0–3)
  • CatMessageDispatcher.cs: Add case "KI": → parse KI{0-3} → update KeyerType
  • CatMultiplexerService.cs: Add KI; to GetInitialValues()
  • CatController.cs: Add POST /api/cat/cw/keyertype body { "type": 1 } → sends KI{type};
  • Index.cshtml: Add Keyer Type dropdown to the CW Keyer dialog, below the Break-in row: Single Paddle / Iambic A / Iambic B / Bug
  • site.js: Change handler + SignalR update on KeyerType

XN — TX Narrow

  • RadioStateService.cs: Add TxNarrow bool property
  • CatMessageDispatcher.cs: Add case "XN": → parse XN{0/1} → update TxNarrow
  • CatMultiplexerService.cs: Add XN; to GetInitialValues()
  • CatController.cs: Add POST /api/cat/txnarrow body { "narrow": true } → sends XN0; or XN1;
  • Index.cshtml: Add Narrow TX toggle button in the Transmit Controls row (§5.10), alongside the ATU and PROC buttons; show only on FTdx101MP/D (same @if gate as other dual-receiver-only controls)
  • site.js: Toggle handler + SignalR update on TxNarrow

Phase 4 — Scan (Priority 4, ~1 day)

Command: SC

No scan state currently exists. Scan is a self-contained mini-feature with 9 modes.

Backend — RadioStateService.cs

  • Add ScanActive bool property; the radio doesn't echo SC back so this is inferred from the last sent command

Backend — CatController.cs

  • POST /api/cat/scan/stop → sends SC0;, sets ScanActive = false
  • POST /api/cat/scan/start body { "mode": 7 } → sends SC{mode};, sets ScanActive = true
  • Modes: 1 = Up, 2 = Down, 3 = MHz Up, 4 = MHz Down, 5 = Prog Up, 6 = Prog Down, 7 = Memory, 8 = Select Memory, 9 = Memory Repeat

Frontend — Index.cshtml

  • Add a Scan button to the toolbar (next to the MEM button); clicking opens a small inline popout with scan mode options and a Stop button
  • Button turns amber while ScanActive is true

Frontend — site.js

  • Click handlers for start/stop
  • SignalR update on ScanActive to keep button state in sync

Phase 5 — Memory Channel Navigation (Priority 4, ~half day)

Commands: MC, MA, MB, CH

MR/MW (full channel read/write) are deferred — they may already be in the memory service and are a large, risky change. MC/MA/MB/CH are small and composable.

Backend — CatController.cs

  • POST /api/cat/memory/recall body { "channel": 5, "vfo": "a" } → sends MC{0/1}{channel:D3};
  • POST /api/cat/memory/tovfo body { "channel": 5, "vfo": "a" } → sends MA{channel:D3}; or MB{channel:D3};
  • POST /api/cat/memory/channel/up → sends CH1;
  • POST /api/cat/memory/channel/down → sends CH0;

Frontend — Index.cshtml

  • Add ◄ CH ► step buttons to the Memory Panel header (§5.15)
  • Add a → VFO A / → VFO B button next to each memory row in the panel

Frontend — site.js

  • Click handlers for the above; update frequency display after recall

Phase 6 — Advanced Odds and Ends (Priority 5, pick individually)

Item What to add Effort
VA already wired Covered in Phase 2 Done
BY — Busy detection Add GET /api/cat/busy → sends BY;, returns { busy: true/false }. Poll in MeterPollingService every 500 ms when squelch is active. Use in site.js to flash a "BUSY" badge. 2 h
PB — Playback recorder Add POST /api/cat/playback/{stop|play|record} → sends PB{0/1/2};. Add Rec / Play / Stop buttons to a small recorder popout (toolbar button). 2 h
KM — Read keyer memory Add GET /api/cat/cw/memory/{1-5} → sends KM{n};, returns text. Display read-back in the CW memory settings panel so user can verify what's stored. 1 h
KC — Clear keyer memory Add POST /api/cat/cw/memory/{1-5}/clear → sends KC{n};. Add a clear (✕) button next to each memory slot. 30 min
AB/BA — Atomic VFO copy Change copyVfo('ab') / copyVfo('ba') in site.js from the two-step FA/FB read-write approximation to a single AB; / BA; command. No state or UI changes. 30 min
PS write — Standby Add POST /api/cat/power/standby → sends PS0;. Add a Standby option to the app's tray menu or a confirmation dialog behind the Exit button. 1 h
BS — Band scope Low value — only relevant for the optional band-scope accessory. Skip unless a user reports it. —
EX remainder Tackle individual EX parameters on-demand as feature requests arrive rather than doing a mass mapping. —

Phase 7 — SP3L Alternative GUI Layout (Jacek SP3L — issue #48, PARKED)

Status: blocked — waiting for Jacek to answer the spectrum placement question.

Colin asked in the issue thread: "Where would the SDR spectrum displays go?" Jacek has not yet replied. Do not start implementation until that question has an answer, because the answer changes the layout fundamentally (spectrum below VFOs = tall scroll on 1080p; spectrum hidden behind toggle = layout only serves SDR-less users; spectrum replaces meter strip when enabled = different again).

What is agreed:

  • A second Razor page (e.g. IndexSp3l.cshtml) plus a companion CSS file. All existing JS and C# control logic reused as-is — no new backend work.
  • User picks the layout in Settings via a new "GUI Layout" dropdown. Options: Classic (current) / SP3L.
  • The Settings dropdown label for Jacek's layout: SP3L.
  • A feature/jacek-gui branch already exists and has the Settings dropdown skeleton in place.

What is agreed on the layout itself:

  • Two panels per radio: VFO-A and VFO-B (not MAIN/SUB/VFO-A/VFO-B — MAIN≡VFO-A and SUB≡VFO-B on the FTdx101 family; on single-receiver radios one panel goes grey as per R1–R12).
  • A global strip at the bottom: Frequency Lock, RF Gain, AF Gain, BK-IN. On the FTdx101 family RF/AF Gain are per-panel (inside each VFO panel), not in the global strip.
  • A TX Audio sub-menu button (next to VOX) grouping: MIC Gain, PROC, Monitor — same controls as the existing popouts, just regrouped.
  • Controls styled as: toggle button (ON/OFF state) + optional slider or dropdown alongside it when ON.

What still needs deciding before code starts:

  1. Spectrum placement — Jacek's answer needed (see above).
  2. "Button OFF" semantics — does pressing a control button to OFF mean "send the radio's default value" (active write) or "this YWC panel stops tracking / go configure from front panel" (read-only handoff)? Needs resolution for IF Width and similar always-on controls.
  3. Low Cut / Audio Filter — in the current layout these are per-mode parameters accessed via the Audio Filter popout (EX commands). Jacek's mockup removed Low Cut. Needs a placement decision for the SP3L layout.

When unblocked, implementation order:

  1. Finalise layout spec with Jacek (close open questions above).
  2. Create Pages/IndexSp3l.cshtml + wwwroot/css/sp3l.css — static HTML structure only, no JS yet.
  3. Wire Settings dropdown to switch the active layout (store choice in appsettings.user.json; serve IndexSp3l from / when SP3L is selected).
  4. Port all existing JS initialisation and SignalR handlers to work against the SP3L DOM.
  5. Test on FTdx101MP and FTdx10 (single-receiver grey-panel behaviour must match R1–R12 spec).

Phase 8 — Voice Control v2 (WAITING — no tester feedback on v1 yet)

v1 status: SHIPPED in v2.4.0-pre1 / pre2. System.Speech (SAPI 5), grammar-based, 6 intents, en-GB only, on-screen mic button. Functionally complete and verified end-to-end on Colin's FTdx101MP.

Trigger for v2 work: at least one named tester confirms v1 working. Do not start v2 until feedback arrives. The failure modes (accent, mic quality, background noise) need real-world data before extending the feature.

Named testers — neither has responded:

  • Yuri W4YSW — accessibility user, FTdx10. Original requestor of the feature.
  • Thomas OZ1JTE — partially sighted, detailed reporter, engaged on #20.
  • Not Bill W1WRH — feedback pace is too slow for a 1–2 week pre-release cycle.

Do not wide-announce on groups.io until v2.4.0 ships clean — voice recognition fails in subtle ways (accent, mic quality, room noise) and a broadcast would generate "doesn't recognise me" reports without clear fixes.

v1 known limitations (become v2 backlog when feedback arrives):

Limitation v2 fix
en-GB only Multi-language scaffold — let user pick recognition language in Settings
VFO A only for all commands Add "VFO A / VFO B" prefix intents: "set VFO B to 14 megahertz"
Restart required to enable/disable Live Settings toggle with no-restart grammar reload
No voice setup wizard Guided in-app setup (mic selection, test phrase, confidence calibration) — design doc at docs/design/voice-control-setup-wizard.md
User-defined commands not implemented Load SRGS files from %APPDATA%\…\Grammars\ at startup and merge into the live grammar. The folder and Settings button already exist as a v2 placeholder (§17.6). Needs: SRGS loader, merge logic, and either a hot-reload path or restart-to-apply. A simple table UI in Settings (phrase → intent mapping) is the minimum; a full SRGS editor is a stretch goal.

v2 implementation notes (when triggered):

  • Multi-language: CultureInfo is already a constructor parameter on SpeechRecognitionEngine. Wiring a Settings dropdown → service restart is the bulk of the work.
  • VFO B targeting: the intent dispatcher routes to CatController helpers that already accept a receiver parameter — adding "VFO A / VFO B" slots to the grammar is the main change.
  • Live toggle: the IHostedService lifecycle needs a ReloadAsync path that tears down and rebuilds the engine without a full app restart.
  • Setup wizard: design doc already accepted — docs/design/voice-control-setup-wizard.md is the authoritative spec. Read it end-to-end before implementing; don't re-derive choices.

Summary Table

Phase Commands New endpoints UI additions Estimate
1 BU, BD, UP, DN, TS 5 Band ◄/►, step ▲/▼, step dropdown ~1 day
2 VA (fix) 0 (existing fix) None ~1 hour
3 KI, XN 2 Keyer type dropdown, Narrow TX toggle ~half day
4 SC 2 Scan button + mode popout ~1 day
5 MC, MA, MB, CH 4 CH step buttons, →VFO buttons ~half day
6 BY, PB, KM, KC, AB/BA, PS 8 Minor per-feature additions ~pick & mix
7 (no CAT commands) 0 new SP3L alternative GUI layout ~1 week — PARKED pending Jacek's spectrum answer
8 (no CAT commands) 0 new Voice control v2 improvements WAITING — v1 shipped in pre-release, no tester feedback yet