Skip to main content

Workload and progression contract

Plan 440–640 focused hours for Semester 04. These are author estimates, not measured completion times or a claim that every learner needs the same duration. Record actual time by category and revise after learner pilots. Prerequisite remediation beyond the bounded bridges is additional. Reading, implementation, debugging and assessment all count; passive elapsed waits do not.

Allocation without double counting​

  • Module 1: 65–95 hours. Original C concepts, multi-file utility and practice 47–69; input-contract extension 18–26.
  • Module 2: 75–110 hours. Original representation, pointers, layout and memory investigation 55–80; ownership extension 20–30.
  • Module 3: 70–100 hours. Original ISA, disassembly, hierarchy and performance practice 52–74; measurement extension 18–26.
  • Module 4: 85–125 hours. Original processes, mappings, threads, sockets and debugging 67–99; partial-I/O extension 18–26.
  • Module 5: 85–125 hours. Original abstraction, higher-order procedures, full interpreter, mutation, streams and staging 65–95; lexical-contract extension 20–30.
  • Required safety/platform strand: 30–40 hours. Fuzzing and minimization 10–14; C/Rust comparison 8–10; mitigation analysis 4–6; SIMD evidence 8–10.
  • Semester integration: 30–45 hours. Assemble/adapt the existing project 12–18; cumulative review and checkpoint 6–9; three-hour exam plus two-hour defense 5; repair and delayed transfer 7–13.

Five modules total 380–555 hours; strand plus integration adds 60–85. Each extension replaces its former short guided lab rather than being added on top of it. Module budgets include quizzes, original practice and extension gates. Integration reuses their artifacts: do not rebuild a second line reader or interpreter just to assemble the packet. New integration code and the distinct strand experiments count here. One artifact can support several claims, but time is recorded once.

At 10 hours/week, plan roughly 44–64 weeks; at 15 hours/week, 30–43; at 20 hours/week, 22–32. Breaks and extended repairs add calendar time. The legacy ten-week arc is an ordering map, not a deadline. Optional compiler/GPU deep dives are outside this core budget; the required specialized-platform evidence is inside it.

Working sequence​

Finish Module 1 input contracts before building the fuzz adapter. Finish Module 2 ownership before the Rust comparison. Finish Module 3 measurement before the SIMD experiment. Complete Module 4 before integration of process and I/O behavior. Use Module 5 to distinguish language-level contracts from host implementation accidents. Keep each strand artifact alongside the module whose evidence it extends.

A learner can reuse a passing sanitizer trace from Module 2, but still needs a minimized fuzz case, Rust comparison, mitigation note and platform experiment. A generic row/column timing comparison alone is not specialized-platform evidence. The strand defines the missing experiment and its acceptance criteria.

Advancement rule​

All CI1–CI6, MP1–MP6, PE1–PE6, IO1–IO6 and LI1–LI6 gates must reach level 2 independently. All five strand categories must also reach level 2. Keep the existing checkpoint threshold of 80%, exam threshold of 70% overall and 60% per major section, project requirements and integrity declaration. These are conjunctive requirements: no average overrides a missing gate.

Maintain an evidence ledger with requirement, artifact path, command/trace, reviewer question, score, repair and fresh retest. A score of 2 means independently correct and supported; 3 means transfer to a new case with limitations. Missing or incorrect evidence is 0; prompted or partial reasoning is 1. Any required item below 2 means repair required — incomplete. Progression to Semester 5 follows successful repair, not a promise to repair later.

Fresh-case defense and bridge​

Spend two hours on a fresh input-capacity case, failed allocation, ambiguous performance observation, partial write and closure capture. Explain one assumption that could invalidate each conclusion. Then connect bytes and lifetime to process address spaces, descriptors to filesystem objects, partial I/O to stream protocols, and shared mutable state to scheduling. This defense complements the written exam and is inside integration time.

After at least 48 hours, revisit the weakest case without solutions. Preserve the original error, correction and new evidence. Waiting is not billed as study time. A learner-ready content label means the route is authored and locally checked; independent review and timing pilots remain pending.

Independent practice route and evidence reuse​

The following entry checks and changed-case practice make the existing module work inspectable. Use them within the current original-practice, lab and retention allowances; no full-book reading or new parallel project is added. An existing artifact can satisfy both an original exercise and this route when it demonstrates the same claim. Keep one evidence index linking each required gate to the exact artifact, actual run, assistance and reviewer status. Do not waive a topic simply because a different lab passes.

Keep authored material, author execution, learner attempts, independent review and delayed transfer as distinct statuses. Retain existing construction, security, HCI, AI-use, world-class and capstone strands and all per-gate minima. Real-system and human requirements remain pending until observed; planned observations are not completed evidence. Record actual study time before revising estimates.