New Mod Proposal: Numeric Control Graph (NCG) — Numeric Signals + Math Nodes + Splitter Control

[tajemniktv]

Summary

Introduce a new automation-style system (likely a standalone mod): Numeric Control Graph (NCG) — a set of nodes that generate, transform, and propagate numeric signals to automatically drive management decisions, primarily splitter ratios/priorities.

Motivation / Problem

Managing splitter ratios becomes tedious because any change requires recalculating ratios from scratch across the graph (e.g., prioritizing Research over AI Trainer, or AGI Core over AI Trainer).

Players want a way to set high-level intent once and have the system rebalance automatically.

High-level Concept

NCG is a parallel “control plane” for the factory:

• Nodes output numbers (signals).
• Other nodes perform math/logic on those numbers.
• Splitters (and potentially other nodes) accept numeric inputs to drive their ratios dynamically.

This turns micromanagement into a design problem: build a control graph, then tune one or two parameters.

Scope (MVP)

1) Numeric Source Nodes

Nodes that output numeric values based on player input or system state.

Examples:

• Constant Number (manual input; supports decimals)
• Resource Rate Monitor (outputs current production/consumption rate)
• Throughput Monitor (outputs actual throughput over a time window)

2) Math Nodes (Transform)

Nodes that process numeric signals.

MVP set:

• Add / Subtract / Multiply / Divide
• Clamp(min,max)
• Normalize (optional)

3) Splitter Integration (Control Input)

Extend existing splitter nodes with numeric control inputs:

• e.g., a control input sets “desired ratio weight” per output
• ratios update automatically based on incoming numeric signals

Optional: skip advanced conditionals for MVP if too complex.

Non-goals (initially)

• Full boolean logic / complex conditional graphs (can be Phase 2)
• Full scripting language
• Overriding core gameplay constraints (heat, power, capacity, etc.)

Proposed UX / UI

• New node category/tab: NCG
• Numeric edges visually distinct from resource edges (color/style)
• Nodes show current output value (live)
• Tooltips explain units (%, weight, rate/min, etc.)

Example Use Cases

• “Always prioritize Research at 60%, AI Trainer at 40%”
• “Scale Research priority up when a certain resource buffer is high”
• “Throttle AI Trainer when heat/power is near limit” (if monitors exist)
• “One knob control”: a single Constant node changes the whole factory policy

Technical / Implementation Notes (suggested)

• Numeric updates should be throttled (e.g., 5–10 Hz or per tick batch) to avoid perf issues.
• Values should be well-defined (float) and clamped to sane ranges.
• Splitter ratio mapping should be deterministic and stable (avoid oscillation):
  • smoothing / moving average
  • hysteresis / minimum change threshold
• Save/load: NCG nodes and connections must serialize cleanly.

Acceptance Criteria (MVP)

• Player can place Numeric Source nodes and set/read numeric values.
• Player can connect numeric signals through basic math nodes.
• At least one splitter type accepts numeric control input(s) and updates ratios accordingly.
• Changes propagate automatically without manual re-tuning across the system.
• System is stable (no rapid oscillation) and performance remains acceptable.

Future Enhancements (Phase 2+)

• Comparison nodes (>, <, ==), thresholding
• Min/Max selectors, conditional routing
• PID-ish controllers / smoothing nodes
• UI panel to inspect/control the whole control graph
• “Policy presets” (research-focused, upload-focused, etc.)

References

Community suggestion: “Numeric Control Graph (NCG)” — numeric nodes to drive management decisions and automate splitter ratio changes.