Show LOCKED feedback for blocked slice tuning

[rfoust]

Summary

• Add a slice-level signal for tune attempts blocked by the slice lock.
• Temporarily show LOCKED in the VFO and RX applet frequency displays when locked tuning is attempted.
• Route blocked wheel, direct-entry, keyboard, FlexControl, HID, MIDI, memory, and band-shortcut tuning paths through the same feedback path.

Testing

• git diff --check
• cmake --build build --clean-first -j8

[Copilot]

Pull request overview

This PR adds unified UI feedback when a user (or control surface) attempts to tune a slice while that slice is locked, by introducing a slice-level “tune blocked” signal and wiring multiple tuning entry points to it.

Changes:

• Add SliceModel::tuneBlockedByLock() signal (+ notifyTuneBlockedByLock() helper) and emit it when setFrequency() / tuneAndRecenter() are blocked by lock.
• Display a temporary LOCKED indicator in the VFO and RX applet frequency readouts when blocked tuning is attempted (with a short timer-based clear).
• Update several MainWindow tuning paths (FlexControl, HID, MIDI, keyboard/band shortcuts, memory recall, etc.) to trigger the same blocked-tune feedback mechanism.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
src/models/SliceModel.h	Adds the new blocked-tune notification API (method + signal).
src/models/SliceModel.cpp	Emits blocked-tune feedback when tuning methods are prevented by slice lock.
src/gui/VfoWidget.h	Adds timer/state needed to show and clear LOCKED feedback.
src/gui/VfoWidget.cpp	Routes blocked wheel/direct-entry tuning to slice notification and displays temporary LOCKED.
src/gui/RxApplet.h	Adds timer/state and helpers for LOCKED feedback in the RX applet frequency display.
src/gui/RxApplet.cpp	Displays temporary LOCKED on blocked tuning attempts and routes relevant UI events through the notifier.
src/gui/MainWindow.cpp	Ensures multiple tuning/control-surface paths trigger the same blocked-tune feedback behavior.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[aethersdr-agent]

Thanks @rfoust — nice clean approach to routing all the blocked tuning paths through a single tuneBlockedByLock() signal. The notifyTuneBlockedByLock() helper guarding the emit with if (m_locked) is a nice touch (safe to call from anywhere). The 500 ms feedback + cancel-direct-entry-on-lock interaction read well.

Blocking — accidental reversion of three merged PRs

The diff includes changes to three files that aren't in the stated scope and that revert commits already on main:

• src/gui/AetherialAudioStrip.cpp — reverts #2891 (#2902) "clarify BYPASS tooltip — Final Output Stage stays active". The diff rewinds the tooltip back to the old wording.
• src/gui/DxClusterDialog.cpp — reverts #2903 (which merged #2890) "raise DX Spots Levels slider cap 10 → 15". The diff puts the slider range back to 1..10.
• src/gui/WaveApplet.cpp — reverts #2827 "persist WaveApplet drawer collapsed state across restarts". The diff removes the WaveApplet_DrawerExpanded read/write and hardcodes setSettingsExpanded(true).

Looks like the branch is based on a main from before those three merges landed (all merged 2026-05-19/20). Could you rebase onto current main and force-push? That should drop these unrelated changes and leave only the 7 files listed in the PR description.

Minor observations on the intended scope (non-blocking)

• RxApplet::connectSlice lock handler (RxApplet.cpp:1670): on lock, it stuffs the current freq into m_freqEdit before switching the stack back to index 0. Since the edit isn't visible anymore, the setText is just bookkeeping for the next entry — fine, just worth a one-line comment if you respin.
• MainWindow::registerShortcutActions band-shortcut path: the de-nesting (early-return on !s) is a nice cleanup but slightly outside the lock-feedback scope. No issue, just noting.

Everything else in the intentional scope looks good. Ping me once it's rebased.

[aethersdr-agent]

Thanks @rfoust — nice, well-scoped change. The single tuneBlockedByLock signal on SliceModel with a notifyTuneBlockedByLock() helper that guards on m_locked is a clean design, and routing every tune entry point (FlexControl, MIDI, HID, keyboard step, band shortcut, memory recall, direct entry, wheel, applyTuneRequest, setFrequency/tuneAndRecenter) through it gives consistent UX. The 500ms single-shot timer + clearLockedFrequencyFeedback() on setSlice() and unlock both look correct, and updateFreqLabel() properly suppresses frequency redraws while LOCKED is on screen so a fresh frequencyChanged won't repaint over the feedback.

A couple of small things worth a look — non-blocking:

1. MainWindow.cpp — FlexControl frequency-mode init path (around if (s->isLocked()) near line ~3510): the new branch resets m_flexTargetMhz = -1.0 on lock, whereas the original if (!s || s->isLocked()) return; left it alone. Probably intentional (the coalesce timer would have cleared it on the next idle anyway), but if a user spins, locks mid-spin, then unlocks, the in-flight tune target is now thrown away rather than continuing. Confirm that's the behavior you want.

2. MainWindow.cpp — MIDI absolute fallback (around line ~14548): the lock check happens before steps = v - 64; if (steps == 0) return;. So a centered absolute controller value (steps=0, would have been a no-op) on a locked slice will still flash "LOCKED" briefly. Minor — moving the steps == 0 early-return ahead of the lock check would avoid the spurious feedback for absolute MIDI knobs that pass through center.

3. RxApplet.cpp — connectSlice() lock handler: the inner if (m_slice) guard inside the lambda is redundant (we're already in connectSlice(s) and the lambda captures this; m_slice == s here). Cosmetic.

Everything else looks good — RAII via QTimer member, no new persistence to worry about (so no AppSettings/QSettings concern), no obvious null-pointer or leak risks, and the file set matches the stated scope.

[rfoust]

Thanks for the review. I pushed an updated single-commit branch that addresses the feedback:

• Rebasing onto current main fixed the stale-base issue. The PR now only changes the intended seven files, so the apparent reversions of Aetherial channel strip BYPASS tooltip is misleading — Final Output Stage remains active #2891/fix(aetherial): clarify BYPASS tooltip — Final Output Stage stays active (#2891) #2902, feat(spothub): raise DX Spots Levels slider cap 10 → 15 (#2890) #2903/Remove limit of 10 lines of DX Spots on Panorama #2890, and fix(wave): persist WaveApplet drawer collapsed state across restarts #2827 are gone.
• The MIDI absolute fallback now returns before lock feedback when steps == 0, avoiding a spurious LOCKED flash for centered/no-op controller values.
• The RX lock handler now uses the connected slice directly instead of the redundant m_slice guard.
• I kept the FlexControl target reset on lock and added a short comment explaining that it prevents stale queued wheel input from replaying after unlock.

Validation:

• git diff --check
• cmake --build build -j8

The branch remains one signed commit, and GitHub shows it as Verified.