Sequential reasoning
Follow a defined sequence, open a subcycle when needed, and return to a saved position.
Two designs / a shared approach
M1 & M3
Study two ways to organize reasoning with explicit inputs, results, memory, and recovery. The book includes their detailed rules and procedures.
Dependency-based reasoning
A transition becomes ready when all required inputs of the right versions have been accepted.
How the architectures differ
| Question | M1 | M3 |
|---|---|---|
| Organization | Sequential work | Dependencies between results |
| What is preserved? | Position, accepted results, and return paths | Required inputs, their versions, and result dependencies |
| What starts the next step? | The next admissible transition in sequence | Acceptance of the required inputs |
| What needs its own test? | Execution on the chosen tools and conditions | Execution and any actual parallel capacity |
Read first. Test in a defined setting.
Reading a design and running it are separate activities. Running requires the exact package, compatible tools, permissions, and an observable startup result.
