A failure from earlier in the semester, read cold with the outcome removed — argue for or against having retried at all.
Due: 28 May 2027 · Weight: 40%
The brief#
Read the case with its outcome removed, and argue whether the retry it describes should have happened.
Eleven weeks have mostly been about retries going wrong, which makes the reflex assumption of this assessment “no” — and that reflex is exactly what week 12 asks you to distrust the same way week 1 asked you to distrust the reflex to say “yes”. The case will not tell you whether the retry helped. You have to argue it from what was known at the time, not from how it turned out.
What you submit#
A written argument, no more than the length stated in the lab brief, taking one side and defending it: what you would have retried or not, what that decision does to the dependency and to the caller still waiting on it, and what you would have measured to know you were right before the outcome was in.
Why there is no criterion table#
This one is marked holistically, on purpose. Splitting “took a position” from “engaged the opposing argument” would reward an answer that does both badly and scores on the boxes, because the two are the same act: the position is only worth anything once it has survived the strongest objection to it.
The spec
The brief poses the problem and leaves room for a response. The spec is the fixed part: what a marker checks. Some lines can be checked mechanically; the rest call for judgement.
- takes a clear position, for or against having retried, rather than presenting both sides unresolved
- names what that decision did to the dependency, to the caller's latency, and to the state a failed attempt left behind
- proposes a measurement that would have told the decision-maker whether they were right, before the outcome was known
- engages the strongest version of the opposing argument, not a weaker one it's easier to beat
How it is marked
Judged as a single argument rather than a checklist: whether the case correctly identifies what the retry actually did — attempts arriving at the dependency, latency on the path that was already working, the state a failed attempt left behind, who could see it happening — whether it commits to a position instead of hedging between both, and whether the measurement it proposes would actually have settled the question at the time, not only in hindsight.