Skip to content

eras.md

Latest commit

 

History

History
170 lines (111 loc) · 3.39 KB

eras.md

File metadata and controls

170 lines (111 loc) · 3.39 KB

Eras

This document defines Eras in Continuum.

Eras are the mechanism by which execution policy changes over simulated time. They control how the simulation runs, not what the simulation is.

1. What an Era Is

An Era defines a named execution regime.

An era specifies:

  • the base timestep (dt)
  • which strata are active or gated
  • cadence modifiers for active strata
  • the conditions under which the era transitions

An era does not define simulation logic. It defines how frequently logic executes.

2. Why Eras Exist

Many simulations require different execution regimes over their lifetime.

Common reasons include:

  • coarse timesteps during early evolution
  • finer timesteps as systems become sensitive
  • delayed activation of subsystems
  • selective suspension of expensive processes

Eras allow the simulation to change resolution and scope without changing its causal structure.

3. Eras and Time

Eras operate on top of the global simulation timeline.

  • Time advances via ticks and dt
  • An era defines the dt used while it is active
  • Era changes do not reset or fork time

Eras change temporal resolution, not meaning.

If changing eras alters qualitative behavior, the model is incorrect.

4. Era Membership and Activation

At any tick, exactly one era is active.

  • The active era determines execution policy
  • All other eras are inactive
  • Era identity is explicit and stable

There is no implicit blending or interpolation between eras.

5. Strata Control Within Eras

Eras control strata execution.

Within an era:

  • strata may be active
  • strata may be gated (paused)
  • strata may execute at reduced cadence

Strata identity does not change across eras. Only their execution eligibility does.

6. Era Transitions

Era transitions are explicit and signal-driven.

A transition specifies:

  • a source era
  • a target era
  • a condition expressed over resolved signals

Transitions:

  • are evaluated at tick boundaries
  • must be deterministic
  • must not depend on fields or observers

An era transition changes execution policy, not simulation state.

7. Initial and Terminal Eras

A World defines:

  • exactly one initial era
  • zero or more terminal eras

A terminal era:

  • has no outgoing transitions
  • may run indefinitely
  • may represent steady-state or real-time execution

Reaching a terminal era is not a failure.

8. Eras and Determinism

Era behavior must be deterministic.

Given:

  • the same World
  • the same Scenario
  • the same seed
  • the same causal history

era transitions must occur at the same ticks and select the same target eras.

Non-deterministic era switching is forbidden.

9. Eras and Scenarios

Scenarios must not alter era structure.

A Scenario may:

  • select the initial era (if allowed by the World)
  • configure parameters used by era transition conditions

A Scenario must not:

  • add eras
  • remove eras
  • alter transition logic

Era structure belongs to the World.

10. What Eras Are Not

Eras are not:

  • phases
  • domains
  • timelines
  • world variants
  • observer constructs

They are execution policy regimes.

Summary

  • Eras define execution regimes
  • They control dt, strata gating, and cadence
  • Eras do not define logic or state
  • Transitions are signal-driven and deterministic
  • Eras change resolution, not meaning

If an era change alters causality, the abstraction has been violated.