Advanced Engineering Extension
Use this sequence after the relevant core prerequisites. These are newly authored lesson studios awaiting independent review and a learner pilot. Workload estimates are additional to the core and include implementation, reading, review, and revision. The program overview distinguishes this extension from existing semester and specialization readiness.
| Order | Studio | Entry diagnostic | Evidence |
|---|---|---|---|
| 1 | Computational limits and reductions | prove a loop invariant and distinguish polynomial from exponential growth | verified reduction and bounded exact solver |
| 2 | Types, compilers, and semantic preservation | trace a lexical closure and recursive AST evaluation | typed language, IR transformation, differential tests |
| 3 | Numerical computing and optimization | solve a 2-by-2 system and explain a dot product | gradient derivation, convergence and sensitivity report |
| 4 | Finite models and protocol correctness | trace a race and state an invariant | bounded state exploration and counterexample |
| 5 | Performance experiments and queueing | compute weighted means and explain a confidence limitation | workload contract and repeated experiment |
| 6 | Storage and distributed failure semantics | distinguish commit, acknowledgment, and persistence | crash matrix and operation histories |
| 7 | Accessible clients and offline state | implement a form, HTTP request, and explicit error state | state machine, keyboard task, conflict tests |
| 8 | Property testing, fuzzing, and oracles | write a parser and explain a precondition | minimized defect and regression evidence |
| 9 | Paper reading and result reproduction | explain a result from lesson 5 or 6 with assumptions | reproduction plus one-variable extension |
| 10 | Maintenance and engineering residency | deliver the core capstone with a clean setup path | maintained releases, migration, recovery, handover |
How to work through a studio
Attempt the diagnostic without the lesson. If it fails, return to its named core prerequisite and solve one changed example before proceeding. Study the worked reasoning, complete the guided assignment, then close the page for the transfer task. Use the assessment rubric for every submission.
The studios intentionally cross course boundaries. For example, the language from lesson 2 becomes a fuzzing target in lesson 8, and the storage simulator from lesson 6 becomes a reproduction subject in lesson 9. Reusing a component is encouraged; each gate still requires new evidence.
You can use local simulators and small datasets. No paid cloud service, cluster, specialized hardware, or external contribution acceptance is required. If a lab depends on OS-specific tooling, record the environment or use the specified deterministic model.
Completion packet
Keep a manifest containing lesson ID, commit or file version, commands, expected results, actual results, assumptions, assistance used, reviewer, and remediation. Supply a text alternative for diagrams. Label self-review honestly. The final packet should let a reviewer reproduce one success and one failure per studio.