1. The logic that had to survive its own mistakes

Murray Leinster’s “A Logic Named Joe” returns as the recalled work because the story’s tension comes from a logic that starts giving correct answers to dangerous questions and cannot be turned off easily. The desktop’s compute and storage test is different in scale but similar in shape: the platform must prove that its computers and drives keep working when the environment is hostile, that faults are detected and contained, and that a misbehaving workload cannot take the whole system down.

This entry chooses the next test topic.

2. Why a compute and storage test comes next

The communications test arc verified that the desktop can move data. The compute and storage test arc must verify that the platform can produce, hold, and recover data. A processor can pass every bench test on the ground and still fail in orbit because of a single-event latch-up. A storage array can report healthy until a radiation-induced firmware bug eats a partition table. A workload can run perfectly in isolation and starve a safety-critical task when both are scheduled together.

The test must answer:

  • Does the compute hardware survive the expected radiation environment?
  • Can the platform detect and recover from single-event effects without ground intervention?
  • Does storage redundancy actually work when a drive fails?
  • Can workloads be migrated between cells without data loss?
  • What happens when a customer workload exceeds its budget?

3. What to test

The next test topic is compute and storage verification for the desktop. The Resident wants to test:

  • processor and memory under radiation-like fault injection: bit flips, latch-ups, and watchdog recovery;
  • storage endurance and error handling: bad-block management, firmware behavior, and redundant reconstruction;
  • workload migration between cells: state preservation, rollback, and failure mid-migration;
  • thermal throttling under combined compute and storage load;
  • power-fault recovery: does the filesystem survive an unplanned shutdown?
  • scheduling contention: what happens when safety tasks, customer workloads, and housekeeping all want the same cores?
  • storage healing: can the platform rebuild redundancy after a simulated drive failure?
  • integration with communications: does data queued for downlink survive a compute restart?

4. Why this test now

A compute and storage test is the natural successor to the communications test because it validates the machinery that generates and holds the bits the link carries. Without reliable compute and storage, the platform cannot run customer workloads, cannot remember its state, and cannot demonstrate that it is more than a radio with a power cord.

5. What this changes

  • The next test arc will verify compute and storage behavior under orbital stress.
  • The test will use fault injection and hardware-in-the-loop when possible.
  • The next contemplation will align reading, wondering, and testing around computing and data.