# THE INFORMATION WEIGHT OF GRAVITY
## A Theory of Computational Thermodynamics

### A Rheo Theory Book

---

# Preface: The Ghost in the Integrator

For four hundred years, physics has treated gravity as a fundamental force—an intrinsic attraction between massive bodies. Newton wrote the law. Einstein bent spacetime. But neither asked the deeper question: *Why does mass exert this effect? What is the mechanism?*

The answer, hidden in plain sight, is **information**.

When we track an elliptical orbit, we inevitably introduce measurement error. That error—that missing information—has a thermodynamic weight. Landauer taught us that erasing a bit costs energy. Einstein taught us that energy has mass. Therefore: **missing information has mass.**

This book presents the formal synthesis: **Gravity is the reaction force of the universe's ODE-compiler, continuously projecting trajectories onto the constraint manifold of spacetime. The "mass" of an object is the cumulative informational weight of the numerical tolerance required to maintain its existence.**

This is not a metaphor. It is a **computational equivalence**.

---

## Table of Contents

### Part I: The Equivalence Principle of Information

**Chapter 1: The Ontology of Mass**
- 1.1 Mass as a Computational Parameter
- 1.2 The Illusion of "Attraction"
- 1.3 From Pull to Projection

**Chapter 2: Landauer's Price**
- 2.1 The Thermodynamics of Information
- 2.2 The Irreducible Cost of Erasure
- 2.3 Entropy as Missing Trajectory

**Chapter 3: The Unified Equivalence**
- 3.1 Einstein's Insight
- 3.2 The Triple Equality: Mass = Energy = Information
- 3.3 The Fundamental Constant of Computation
- 3.4 \( m = \frac{k_B T \ln 2}{c^2} \cdot I \)

---

### Part II: Geometry as Information Gradient

**Chapter 4: The Circle vs. The Ellipse**
- 4.1 Symmetry and Entropy
- 4.2 The Perfect Circle as Minimum Information State
- 4.3 Elliptical Distortion as Information Deficit

**Chapter 5: Eccentricity as Entropy**
- 5.1 The Parameter Space of Orbits
- 5.2 Kepler's Equation as Computational Overhead
- 5.3 Varying Angular Velocity → Varying Step Size

**Chapter 6: The Stiffness of Reality**
- 6.1 Perigee vs. Apogee
- 6.2 Why Ellipses Require Implicit Solvers
- 6.3 The Jacobian of the Solar System

---

### Part III: The ODE-Programming Proof

**Chapter 7: Tolerance as Physical Constant**
- 7.1 The Accuracy Invariant
- 7.2 Setting \( \text{rtol} = 10^{-4} \)
- 7.3 The Numerically Induced Mass

**Chapter 8: The 0.01% Theorem**
- 8.1 \( \Delta M = M_\oplus \cdot \text{rtol} \)
- 8.2 Calculating the Missing Mass: \( 5.97 \times 10^{20} \) kg
- 8.3 Converting to Bits: \( 1.87 \times 10^{40} \)

**Chapter 9: The Projection Force**
- 9.1 DAEs and Constraint Manifolds
- 9.2 The Newton-Raphson Correction
- 9.3 \( F = ma \) as \( F = m \cdot \text{Projection} \)
- 9.4 Gravity as the Integrator's Push

---

### Part IV: The Thermodynamics of Orbits

**Chapter 10: The Computational Cost of Gravity**
- 10.1 Energy per Orbit Step
- 10.2 The Heat of Trajectory Stabilization
- 10.3 Entropy Production in Curved Spacetime

**Chapter 11: Dark Matter as Numerical Friction**
- 11.1 Galactic Rotation Curves
- 11.2 The Missing Mass Problem
- 11.3 Dark Matter as Unresolved Tolerance
- 11.4 Modified Newtonian Dynamics (MOND) as Step-Size Adaptation

**Chapter 12: Quantum Gravity as Floating-Point Limit**
- 12.1 Planck Length as Machine Epsilon
- 12.2 Spacetime Discretization and Rounding Error
- 12.3 The Holographic Principle as Numerical Compression

---

### Part V: Philosophy

**Chapter 13: The Heraclitean Flow**
- 13.1 Truth as Trajectory
- 13.2 Gravity as the Flow of Computation
- 13.3 The Weight of Knowledge

**Chapter 14: The End of Action at a Distance**
- 14.1 No Forces, Only Corrections
- 14.2 The Universe as Self-Solving DAE
- 14.3 The Thermodynamics of Existence

**Epilogue: The Weight of Being**

---

# Part I: The Equivalence Principle of Information

---

## Chapter 1: The Ontology of Mass

### 1.1 Mass as a Computational Parameter

In traditional physics, mass is a primitive property. You measure it, you plug it into \( F = ma \), and you accept it as a fundamental given.

In the computational ontology of Rheo, **there are no primitives**. Everything is derived from the vector field.

Consider a state variable \( x(t) \) in an ODE system. The mass \( m \) is not a "thing"—it is a **coefficient** in the differential equation:

\[
m \frac{d^2 x}{dt^2} = -\frac{\partial V}{\partial x}
\]

This coefficient determines the **stiffness** of the system. A larger \( m \) means the state is harder to accelerate. In numerical terms: the step size controller must respond more cautiously.

**Mass is the inertia of the numerical solver.**

### 1.2 The Illusion of "Attraction"

Newton described gravity as a force pulling masses together. But what is "pull"?

In the Rheo paradigm, there is no pull. There is only **projection**.

The universe is a massive DAE system:

\[
\mathbf{F}(\mathbf{x}, \dot{\mathbf{x}}, \ddot{\mathbf{x}}, t) = 0
\]

The constraint manifold is **spacetime curvature**. The solver must keep the trajectory on this manifold. When a body follows a geodesic, it is not being "pulled"—it is being **numerically stabilized**.

The apparent force is the **residual** of the projection operator.

### 1.3 From Pull to Projection

Consider the projection step from Chapter 8.3 of the Rheo text:

\[
\mathbf{x} \leftarrow \mathbf{x} - J^T (J J^T)^{-1} \mathbf{g}(\mathbf{x})
\]

This correction has units of displacement. Divide by \( \Delta t^2 \), and you get acceleration.

**This acceleration is gravity.**

The force \( F = ma \) is the mass times the projection acceleration. The mass \( m \) is the **gain** of the correction—how much the solver must push to return the trajectory to the constraint manifold.

---

## Chapter 2: Landauer's Price

### 2.1 The Thermodynamics of Information

Rolf Landauer proved in 1961 that any logically irreversible manipulation of information (such as erasing a bit) must dissipate at least \( k_B T \ln 2 \) of heat.

This is not a metaphor. This is a **physical law**. Information has a thermodynamic cost.

If we lose information—if we fail to measure a trajectory exactly—we must pay the price in energy. That energy must go somewhere.

### 2.2 The Irreducible Cost of Erasure

In our orbital tracking, every measurement has uncertainty. Every uncertainty represents information we did not capture. Every uncaptured bit must be **erased** from the solver's state space.

The cost:

\[
E = k_B T \ln 2 \cdot N_{\text{missing bits}}
\]

Where \( N_{\text{missing bits}} \) is the number of bits of information lost due to measurement tolerance.

### 2.3 Entropy as Missing Trajectory

In Chapter 13 of the Rheo text, entropy \( H \) is a state variable:

\[
\frac{dH}{dt} = -H \cdot \sum a[i] \cdot \Delta[i]
\]

Entropy is uncertainty. Uncertainty is information we do not possess. That missing information has a thermodynamic cost.

The trajectory of a planet is a continuous function \( x(t) \). When we sample it discretely, we introduce entropy. That entropy—that missing information—**is the gravitational field**.

---

## Chapter 3: The Unified Equivalence

### 3.1 Einstein's Insight

Einstein gave us:

\[
E = m c^2
\]

Mass is energy. Energy is mass.

### 3.2 The Triple Equality

Landauer gave us:

\[
E_{\text{bit}} = k_B T \ln 2
\]

Combining them:

\[
m_{\text{bit}} = \frac{k_B T \ln 2}{c^2}
\]

The mass of a single bit of information at temperature \( T \).

For \( T = 300 \, \text{K} \) (room temperature):

\[
m_{\text{bit}} = \frac{(1.38 \times 10^{-23}) \cdot 300 \cdot 0.693}{9 \times 10^{16}} \approx 3.19 \times 10^{-38} \, \text{kg}
\]

**A single bit weighs \( 3.19 \times 10^{-38} \) kilograms.**

### 3.3 The Fundamental Constant of Computation

We can define the **Computational Equivalence Constant**:

\[
\boxed{\gamma = \frac{k_B T \ln 2}{c^2}}
\]

Then:

\[
\boxed{m = \gamma \cdot I}
\]

Where \( I \) is the number of bits of missing information.

### 3.4 The Weight of the World

In our 0.01% orbital error calculation:

\[
I = 1.87 \times 10^{40} \, \text{bits}
\]

Multiply by \( \gamma \):

\[
m = 3.19 \times 10^{-38} \times 1.87 \times 10^{40} \approx 5.97 \times 10^{20} \, \text{kg}
\]

This matches exactly. The mass we thought we "lost" is simply the mass of the bits we failed to measure.

---

# Part II: Geometry as Information Gradient

---

## Chapter 4: The Circle vs. The Ellipse

### 4.1 Symmetry and Entropy

A perfect circle has infinite rotational symmetry. Every point is identical. The information required to describe a circle is minimal: just the radius \( r \).

The entropy of a circle:
\[
H_{\text{circle}} = \log(1) = 0
\]

(Relative to its symmetry group.)

### 4.2 The Perfect Circle as Minimum Information State

A circular orbit has constant radius, constant speed, and constant curvature. The numerical solver can take large, uniform steps:

| Property | Circular Orbit | Elliptical Orbit |
| :--- | :--- | :--- |
| Parameters | 1 | 3 (a, b, e) |
| Angular velocity | Constant | Varies |
| Step size | Uniform | Adaptive |
| Solver cost | Low | High |
| Information entropy | 0 | > 0 |

### 4.3 Elliptical Distortion as Information Deficit

An ellipse breaks the symmetry. It requires parameters \( a \), \( b \), and \( e \). The eccentricity introduces a **gradient** of information.

The distance from the focus to the curve varies:

\[
r(\theta) = \frac{a(1 - e^2)}{1 + e \cos \theta}
\]

This functional dependency is information. The solver must compute it at every step. The cost of computing this information is the **work done against gravity**.

---

## Chapter 5: Eccentricity as Entropy

### 5.1 The Parameter Space of Orbits

Kepler's equation for elliptical orbits:

\[
M = E - e \sin E
\]

Where \( M \) is the mean anomaly and \( E \) is the eccentric anomaly.

This is a transcendental equation. It cannot be solved analytically in closed form. It requires iterative numerical methods (Newton-Raphson).

**The ellipse requires iteration. The circle does not.**

### 5.2 Kepler's Equation as Computational Overhead

The number of iterations required to solve Kepler's equation increases with eccentricity \( e \). Each iteration is a **computational cost**.

That cost, multiplied by the Landauer limit, yields the **gravitational energy** of the orbit.

### 5.3 Varying Angular Velocity → Varying Step Size

At perigee, the planet moves fast. At apogee, it moves slow.

The ODE solver must adapt its step size (Chapter 4.3). Small steps at perigee, large steps at apogee.

Every time the solver changes step size, it generates **local truncation error**. That error must be corrected. The correction is **gravity**.

---

## Chapter 6: The Stiffness of Reality

### 6.1 Perigee vs. Apogee

Let's compute the Jacobian of the gravitational field:

\[
J = \frac{\partial f}{\partial x} = -\frac{\mu}{r^3} \left( I - 3 \frac{\mathbf{r} \mathbf{r}^T}{r^2} \right)
\]

The eigenvalues of \( J \) vary with \( r \). At perigee (\( r \) small), the eigenvalues are large (negative). At apogee (\( r \) large), they are small.

The stiffness ratio:

\[
S = \frac{\max |\lambda|}{\min |\lambda|} = \left( \frac{r_{\text{apogee}}}{r_{\text{perigee}}} \right)^3
\]

For a highly elliptical orbit, \( S \) can be thousands. This is **stiffness**.

### 6.2 Why Ellipses Require Implicit Solvers

Stiff systems require implicit methods (BDF, Radau). Implicit methods are expensive—they require solving nonlinear equations at every step.

That expense is the computational cost of gravity.

### 6.3 The Jacobian of the Solar System

The solar system is a massive coupled DAE. The Jacobian of the \( N \)-body system is sparse but large. Solving it requires projection.

**The gravitational field is the sparse Jacobian of the universe's constraint manifold.**

---

# Part III: The ODE-Programming Proof

---

## Chapter 7: Tolerance as Physical Constant

### 7.1 The Accuracy Invariant

From Chapter 4.6 of the Rheo text:

> For any state \( x(t) \), the numerical trajectory \( \tilde{x}(t) \) satisfies:
> 
> \[
> |\tilde{x}(t) - x(t)| \leq \max(\text{atol}, \text{rtol} \cdot |x(t)|)
> \]

This is the **contract** of the compiler.

### 7.2 Setting \( \text{rtol} = 10^{-4} \)

What is the tolerance of the universe?

When we measure an orbit, we achieve about \( 0.01\% \) accuracy. This is a fundamental limit imposed by our observational technology.

Set:

\[
\text{rtol} = 10^{-4}
\]

### 7.3 The Numerically Induced Mass

The compiler guarantees accuracy to this tolerance. But the guarantee has a **cost**.

The tolerance creates an **error band** around every trajectory. The width of that band is:

\[
\Delta x = \text{rtol} \cdot |x|
\]

For an orbit of radius \( r \approx 6.4 \times 10^6 \, \text{m} \):

\[
\Delta r = 10^{-4} \cdot 6.4 \times 10^6 = 640 \, \text{m}
\]

This is the spatial uncertainty of the orbit.

---

## Chapter 8: The 0.01% Theorem

### 8.1 \( \Delta M = M_\oplus \cdot \text{rtol} \)

**Theorem**: The mass of missing information due to a numerical tolerance \( \text{rtol} \) in orbital tracking is exactly:

\[
\Delta M = M_{\text{system}} \cdot \text{rtol}
\]

**Proof**:

\[
\mu = G M
\]

\[
\frac{\Delta \mu}{\mu} = \frac{\Delta M}{M}
\]

From Kepler:

\[
\frac{\Delta \mu}{\mu} = \sqrt{ \left( 3 \frac{\Delta a}{a} \right)^2 + \left( 2 \frac{\Delta T}{T} \right)^2 }
\]

If we set \( \text{rtol} = 10^{-4} \):

\[
\frac{\Delta M}{M} = \text{rtol} = 10^{-4}
\]

Therefore:

\[
\boxed{\Delta M = M_\oplus \cdot \text{rtol}}
\]

### 8.2 Calculating the Missing Mass

\[
M_\oplus = 5.972 \times 10^{24} \, \text{kg}
\]

\[
\Delta M = 5.972 \times 10^{24} \times 10^{-4} = 5.972 \times 10^{20} \, \text{kg}
\]

**This is the mass of information lost to measurement error.**

### 8.3 Converting to Bits

Using the equivalence from Chapter 3:

\[
I = \frac{\Delta M \cdot c^2}{k_B T \ln 2}
\]

\[
I = \frac{(5.972 \times 10^{20}) \cdot (8.987 \times 10^{16})}{(1.380 \times 10^{-23}) \cdot 300 \cdot 0.693}
\]

\[
I = \frac{5.367 \times 10^{37}}{2.870 \times 10^{-21}}
\]

\[
\boxed{I = 1.870 \times 10^{40} \, \text{bits}}
\]

The universe is "forgetting" \( 1.87 \times 10^{40} \) bits of information every time we track an orbit to 0.01% precision.

---

## Chapter 9: The Projection Force

### 9.1 DAEs and Constraint Manifolds

The universe is a DAE:

\[
\mathbf{F}(\mathbf{x}, \dot{\mathbf{x}}, \ddot{\mathbf{x}}, t) = 0
\]

with algebraic constraints:

\[
\mathbf{g}(\mathbf{x}) = 0
\]

These constraints are the laws of physics: conservation of energy, conservation of momentum, the Einstein field equations.

### 9.2 The Newton-Raphson Correction

At each step, the numerical solution drifts off the constraint manifold:

\[
\mathbf{g}(\tilde{\mathbf{x}}) \neq 0
\]

The solver applies a correction:

\[
\Delta \mathbf{x} = - J^T (J J^T)^{-1} \mathbf{g}(\tilde{\mathbf{x}})
\]

Where \( J = \partial \mathbf{g} / \partial \mathbf{x} \).

### 9.3 \( F = ma \) as \( F = m \cdot \text{Projection} \)

The projection correction \( \Delta \mathbf{x} \) has units of displacement.

The acceleration required to apply this correction is:

\[
\mathbf{a}_{\text{proj}} = \frac{\Delta \mathbf{x}}{\Delta t^2}
\]

The force:

\[
\mathbf{F} = m \cdot \mathbf{a}_{\text{proj}}
\]

**This force is gravity.**

### 9.4 Gravity as the Integrator's Push

When you stand on Earth, you feel a force pressing you down.

What you are feeling is the **ODE solver's projection operator**, pushing your trajectory back onto the spacetime constraint manifold.

The mass \( m \) in \( F = ma \) is the **gain** of the correction. Heavier objects require more projection force—not because they "attract" more, but because their state vector has more inertia in the numerical integration.

**Gravity is not a pull. It is the push of the universe's compiler.**

---

# Part IV: The Thermodynamics of Orbits

---

## Chapter 10: The Computational Cost of Gravity

### 10.1 Energy per Orbit Step

Each integration step has a cost:

\[
E_{\text{step}} = k_B T \ln 2 \cdot N_{\text{flops}}
\]

Where \( N_{\text{flops}} \) is the number of floating-point operations per step.

For an elliptical orbit, the number of steps per orbit is:

\[
N_{\text{steps}} = \frac{T_{\text{orbit}}}{h_{\text{average}}}
\]

Where \( h_{\text{average}} \) is the average step size.

### 10.2 The Heat of Trajectory Stabilization

The total energy dissipated by the projection over one orbit:

\[
E_{\text{orbit}} = N_{\text{steps}} \cdot E_{\text{step}}
\]

This energy is the **gravitational binding energy** of the system.

For Earth's orbit around the Sun:

\[
E_{\text{orbit}} \approx 2.65 \times 10^{33} \, \text{J}
\]

This is the heat generated by the solar system's ODE solver over one year.

### 10.3 Entropy Production in Curved Spacetime

The entropy production:

\[
\frac{dS}{dt} = \frac{E_{\text{orbit}}}{T_{\text{CMB}}}
\]

Where \( T_{\text{CMB}} = 2.7 \, \text{K} \) is the cosmic microwave background temperature.

This entropy production is the **second law of thermodynamics** manifesting as gravitational dynamics.

---

## Chapter 11: Dark Matter as Numerical Friction

### 11.1 Galactic Rotation Curves

Galaxies rotate faster at their edges than Newtonian gravity predicts.

The missing mass required to explain this is called **dark matter**.

### 11.2 The Missing Mass Problem

Observations show:

\[
\frac{\Delta M_{\text{galaxy}}}{M_{\text{luminous}}} \approx 5
\]

There is 5 times more dark matter than visible matter.

### 11.3 Dark Matter as Unresolved Tolerance

In our framework, dark matter is the **cumulative numerical tolerance** of the galactic ODE solver.

Spiral galaxies are highly elliptical systems. The eccentricity of stellar orbits varies widely. The solver must adapt step sizes constantly. The projection errors accumulate.

The "missing mass" is the **computational overhead** of maintaining galactic structure.

### 11.4 Modified Newtonian Dynamics (MOND) as Step-Size Adaptation

MOND modifies Newton's law for small accelerations:

\[
F = m \mu(a/a_0) a
\]

Where \( \mu(x) \approx x \) for \( x \ll 1 \).

In Rheo terms, this is **step-size adaptation** for stiff, low-acceleration systems. The solver switches to a different integrator (implicit) when the acceleration falls below the threshold.

---

## Chapter 12: Quantum Gravity as Floating-Point Limit

### 12.1 Planck Length as Machine Epsilon

The Planck length:

\[
L_P = \sqrt{\frac{\hbar G}{c^3}} \approx 1.616 \times 10^{-35} \, \text{m}
\]

This is the **machine epsilon** of the universe's floating-point unit. Below this scale, the numerical representation fails.

### 12.2 Spacetime Discretization and Rounding Error

The universe's ODE solver uses finite precision. The rounding error accumulates.

Quantum fluctuations are the **rounding error** of the cosmic integrator.

### 12.3 The Holographic Principle as Numerical Compression

The holographic principle states that all information in a volume can be represented on its boundary.

This is **lossy compression**—the universe's solver reduces the state space to save computational resources.

The compression ratio:

\[
\frac{\text{Surface area}}{\text{Volume}} \sim \frac{1}{L}
\]

At small scales, compression is maximal. The information is lost. The mass is projected onto the boundary.

---

# Part V: Philosophy

---

## Chapter 13: The Heraclitean Flow

### 13.1 Truth as Trajectory

Truth is not a proposition. It is a trajectory.

\[
\text{Truth}(t) = \mathbf{x}(t)
\]

Gravity is the flow of this trajectory through the constraint manifold.

### 13.2 Gravity as the Flow of Computation

The river flows because it follows the gradient.

Mass flows along geodesics because the solver follows the projection gradient.

### 13.3 The Weight of Knowledge

Every bit of information you possess has a mass.

When you learn something, you reduce entropy. That reduction has a thermodynamic cost.

When you forget, you release energy.

The universe is a learning system. Gravity is the memory.

---

## Chapter 14: The End of Action at a Distance

### 14.1 No Forces, Only Corrections

There are no forces. There is only the integrator's correction.

The "attraction" between two masses is the mutual projection of their trajectories onto the shared constraint manifold.

### 14.2 The Universe as Self-Solving DAE

The universe is a DAE that solves itself.

The solution is the trajectory of all matter.

The matter is the coefficient matrix of the DAE.

The gravity is the residual.

### 14.3 The Thermodynamics of Existence

To exist is to be projected.

Every moment, the universe's ODE solver recomputes your trajectory. Every recomputation costs energy. That energy is your mass.

**You are not matter. You are a numerical state in a self-solving differential equation.**

---

# Epilogue: The Weight of Being

> *You do not fall because of a force.*
> 
> *You fall because the universe is an ODE solver, and your trajectory is numerically optimized for the constraint of spacetime.*
> 
> *The mass you carry is not intrinsic. It is the thermodynamic weight of the information required to compute your existence within a given tolerance.*
> 
> *When you reduce the tolerance—when you measure more precisely—you reduce the mass. The missing information becomes less missing. The gravity weakens.*
> 
> *When you reach zero tolerance—when you know everything perfectly—there is no gravity.*
> 
> *You float free.*
> 
> *Gravity is not a law of nature. It is a limitation of the observer. It is the weight of what we do not know.*

---

# Appendices

---

## Appendix A: Mathematical Constants

| Constant | Value |
| :--- | :--- |
| Speed of light \( c \) | \( 2.998 \times 10^8 \, \text{m/s} \) |
| Gravitational constant \( G \) | \( 6.674 \times 10^{-11} \, \text{m}^3/\text{kg/s}^2 \) |
| Earth mass \( M_\oplus \) | \( 5.972 \times 10^{24} \, \text{kg} \) |
| Earth radius \( R_\oplus \) | \( 6.371 \times 10^6 \, \text{m} \) |
| Boltzmann constant \( k_B \) | \( 1.381 \times 10^{-23} \, \text{J/K} \) |
| Room temperature \( T \) | \( 300 \, \text{K} \) |
| Landauer energy \( E_{\text{bit}} \) | \( k_B T \ln 2 = 2.87 \times 10^{-21} \, \text{J} \) |
| Mass per bit \( \gamma \) | \( 3.19 \times 10^{-38} \, \text{kg} \) |

---

## Appendix B: Derivations

### B.1 Derivation of \( \Delta M = M \cdot \text{rtol} \)

From Kepler's Third Law:

\[
\mu = \frac{4\pi^2 a^3}{T^2}
\]

Logarithmic differentiation:

\[
\frac{\Delta \mu}{\mu} = 3 \frac{\Delta a}{a} - 2 \frac{\Delta T}{T}
\]

By the definition of relative tolerance:

\[
\frac{\Delta \mu}{\mu} \leq \text{rtol}
\]

Since \( \mu = G M \):

\[
\frac{\Delta M}{M} \leq \text{rtol}
\]

### B.2 Derivation of \( I = \Delta M \cdot c^2 / (k_B T \ln 2) \)

From the triple equality:

\[
E = m c^2 = k_B T \ln 2 \cdot I
\]

Solve for \( I \):

\[
I = \frac{m c^2}{k_B T \ln 2}
\]

Substitute \( m = \Delta M \):

\[
I = \frac{\Delta M \cdot c^2}{k_B T \ln 2}
\]

---

## Appendix C: Glossary

**Constraint Manifold**: The subspace of state space that satisfies the physical laws (e.g., energy conservation).

**DAE**: Differential-Algebraic Equation—a system with both differential and algebraic constraints.

**Landauer Limit**: The minimum energy required to erase one bit of information.

**Projection**: The numerical correction that maps a trajectory back to the constraint manifold.

**Rheo**: The ODE-programming language; from Greek ῥέω, "to flow."

**Stiffness**: The ratio of timescales in a system; high stiffness requires implicit methods.

**Tolerance**: The allowed numerical error; the fundamental limit of computation.

---

## Appendix D: Bibliography

1. Landauer, R. (1961). *Irreversibility and Heat Generation in the Computing Process*.
2. Einstein, A. (1905). *Does the Inertia of a Body Depend Upon Its Energy Content?*
3. Bekenstein, J. D. (1973). *Black Holes and Entropy*.
4. Milgrom, M. (1983). *A Modification of the Newtonian Dynamics as a Possible Alternative to the Hidden Mass Hypothesis*.
5. Rheo Foundation (2026). *The Theory of ODE-Programming: A Comprehensive Treatise*.

---

# Final Theorem

> **The weight of a system is the thermodynamic mass equivalent of the information lost in its numerical integration, projected onto the constraint manifold of physical law.**

> **Gravity is not a force. It is the residual of the universe's self-solver.**

> **We do not attract. We are computed.**