Previous assignment edition. Resume or export work already started here. New work uses the current edition; do not resubmit completed work unless instructed. Current edition

AFTER-CLASS · CHAPTER 10

Dependable Systems: a decision you can defend

Practise first, make your own decision, then respond to a new constraint.

30–40 minutes4 points totalDeadline: course LMSIndividual response
  1. Warm-up
  2. FIT → BOUND → ACT + EVIDENCE
  3. Commit Part A
  4. STRESS → REFIT
  5. Review and submit in LMS
Checking storage…Not yet saved

1 · WARM-UP · 5 MINUTES · NOT GRADED

Start with one clear distinction.

An online timetable opens whenever requested, but sometimes displays yesterday’s timetable. Is availability alone enough to call it dependable?

Show / hide warm-up answer
No. Availability concerns whether the service is accessible when needed; reliability concerns correct service over time under stated conditions. A page can be accessible while returning incorrect information. State what users need and what evidence would establish correct behaviour.

No written warm-up response is required. Now transfer the reasoning to a different service in Part A.

Lecture guide and sentence stems

Open the lecture’s Units menu: Rule 03 — Observable outcomes; X01 — dependability properties; Rule 07 — Decision card; Rule 12 — Owner, evidence, sign-off; Rule 13 — Breaking variable; Rule 15 — AI assist + human sign-off; Rule 17 — Constraint mutation. Use C01 — Source ledger to locate source-slide references.

FIT: The governing property is … because this service must …

BOUND: This decision remains acceptable only while …; reopen if …

ACT + EVIDENCE: My decision record says …; it is supported by …, checked by …

READ BEFORE ANSWERING

How your work is assessed · 4 points

CriterionScoring
FIT · 1 point1: Relevant dependability property, a correctly linked mechanism, and a scenario-specific reason.
0.5: Relevant concept, but a missing mechanism or incomplete link to this context.
0: Absent, irrelevant or materially incorrect.
BOUND · 1 point1: Observable condition and clear trigger to reopen, consistent with the plan.
0.5: Condition or trigger only partly specified.
0: Absent, uncheckable or contradictory.
ACT + EVIDENCE · 1 point1: Feasible immediate action with responsible role, plus inspectable evidence tied to its boundary; unknown or proposed evidence is labelled.
0.5: Action or evidence is concrete, but the other is incomplete or poorly connected.
0: Neither supports a defensible plan, or results are fabricated.
REFIT · 1 point1: Explains the effect on the original boundary, justifies RETAIN, REVISE or REPLACE, and supplies a coherent final plan.
0.5: Plausible final choice, but a weak link to the original boundary or an incomplete final plan.
0: No defensible response, or material contrary evidence is ignored.

Any REFIT choice can earn full credit when justified; changing your answer earns no automatic credit. The instructor judges the reasoning, not language polish, length or box completion. The lecture’s L1–L4 maturity descriptions are a separate practice guide.

Lecture outcomes practised: explain dependability properties; build a bounded verdict with inspectable evidence; defend the verdict when a constraint changes. Comparing alternatives is practised when relevant; a full sociotechnical layer trace is not assessed in this first task.

How this assignment connects to the lecture

FIT: X01 / Rule 05 — properties and prediction. BOUND + EVIDENCE: Rules 07, 09 and 10 — decision card, failure conditions and risk. ACT: Rules 06 and 12 — engineering flow and accountable roles. REFIT: Rules 13 and 17 — breaking variables and changed constraints.

FIT → BOUND → ACT + EVIDENCE → STRESS → REFIT is the after-class record of the same decision-and-evidence approach used in the lecture.

Before you start

Use Save draft while working and Download draft / backup before leaving a shared device. Part A becomes read-only only after you confirm commitment. You can continue editing Part B afterward.

For this staged exercise, do not read another student’s STRESS before committing. The new evidence opens automatically after you commit Part A. No unlock code is required. Your initial reasoning stays saved and read-only. The page sequences the task; it is not a secure examination system.

Identity

Loaded from the entry page. Drafts are separated by Student ID in this browser.

Submission details

Check the current deadline and submission instructions in the LMS.

2 · PART A · 15 MINUTES

Assessed scenario · campus course registration

At 11:30, a fictional university's course-registration portal is experiencing intermittent errors during peak registration. Registration closes at 12:00. Two application servers support the portal, and automatic failover is enabled. The available failover test passed during a quiet period; no peak-load failover result is available. Current checks show that students can sign in and that confirmed registrations can be retrieved. No lost or duplicate registrations have been confirmed, but this has not yet been comprehensively checked. You are the duty software engineer. The registrar owns the registration deadline and may authorize an alternative submission route; you cannot change either policy yourself.

Your decision: Recommend one immediate operating plan: continue with stated controls, limit operation, or pause the affected function. Start with one governing dependability property, one observable boundary, one action with a responsible role, and one inspectable evidence item. Explain any mechanism you rely on. State what is known and what you would still need to verify. Your plan must address students' ability to complete registration and the integrity of confirmed registrations. There is no single required choice.

Teaching simulation. Produce a decision record using the stated facts. Do not invent test results or certify a real service. Mark missing evidence as unmeasured and specify how you would obtain it.

Aim for 2–4 focused sentences per reasoning field, about 250–400 focused words overall as guidance. ACT can be a compact record with labelled lines. Include at least one relevant lecture unit/title or assigned source-page citation supporting your mechanism. No code, external experiments, real student records or numerical reliability calculation is required.

A1 · FIT · 1 POINT

Select the leading dependability property, explain the consequence for users, and justify one relevant mechanism. Distinguish the property from the mechanism used to support it.

A2 · BOUND · 1 POINT

State an observable applicability condition and a trigger for reopening the decision. Name what remains unknown. Label any threshold you propose as proposed and requiring owner approval. Do not guess the later STRESS.

A3 · ACT · SCORED WITH EVIDENCE

State the action or verdict, its reason and operating condition, and the role responsible for it. This is your actual short artifact, not a promise to write one later.

A4 · EVIDENCE · SCORED WITH ACT

Name a record, test or measurement; say why it supports the decision, which version or conditions it must cover, and who checks it. Distinguish supplied scenario facts, assumptions, and evidence you still need. Label proposed tests or logs as requested evidence, not completed results. BOUND is the condition; EVIDENCE is how you check it.

Commit is not LMS submission. Confirmation freezes your original reasoning before the new constraint appears. Your final submission comes after REFIT and human review.

Privacy: responses are stored locally in this browser under Chapter + Student ID. They are not sent to the instructor by this page. Download a backup you can open, save the final PDF, and submit it through the LMS.