Context
In high-frequency update scenarios (e.g., Grid scrolling), ScrollManager currently uses a fixed throttle (e.g., 48ms) to prevent flooding the VDOM worker with updates the Main Thread cannot consume fast enough ("Death Spiral").

While effective, a fixed timer is not adaptive. It caps performance on high-end devices (which could handle >20fps) and might still be too aggressive for very low-end devices.

The Goal
Implement Adaptive Backpressure where the producer (ScrollManager) automatically scales its update rate based on the consumer's (Main Thread) capacity.

The Mechanism

1. Backpressure Check: The producer checks component.isVdomUpdating.
2. Opt-Out: If true, the producer skips the heavy calculation (e.g., createViewData) and registers a Pre-Update Hook.
3. Re-Entry: When the component finishes its current update (resolveVdomUpdate), it executes the registered Pre-Update Hook before checking needsVdomUpdate for the next cycle.

Proposed Changes

1. src/manager/VDomUpdate.mjs:

   • Add preUpdateCallbackMap (Map<String, Function[]>).
   • Implement registerPreUpdate(ownerId, callback).
   • Implement executePreUpdates(ownerId): Executes callbacks and clears the list.

2. src/mixin/VdomLifecycle.mjs:

   • Update resolveVdomUpdate:

     // ... existing logic ...
     VDomUpdate.unregisterInFlightUpdate(me.id);
     VDomUpdate.triggerPostUpdates(me.id);
     
     // NEW: Execute Pre-Update Hooks (e.g. re-run createViewData)
     VDomUpdate.executePreUpdates(me.id);
     
     if (me.needsVdomUpdate) {
         me.update();
     }

3. Usage (Example):

   // ScrollManager.mjs
   if (gridBody.isVdomUpdating) {
       // Queue the work for immediately after the current frame clears
       Neo.manager.VDomUpdate.registerPreUpdate(gridBody.id, () => {
            // This sets needsVdomUpdate = true internally
            gridBody.createViewData();
       });
       return;
   }

Benefits

• Zero Latency: No fixed timers. Updates happen as fast as the hardware allows.
• Stability: Impossible to flood the queue, as new work is only scheduled when the previous work completes.
• Efficiency: Skips redundant calculations during blocked periods.