Conformance
Run your work against the vectors to see whether it conforms, and at which level.
Every specification here is a draft (version below 1.0). Build to them and comment on them, knowing they may change. No licence is granted yet.
In short
Each specification publishes test vectors in one shared format. Run your work against them to see whether it conforms, and at which level. Only CTU metering has published vectors so far; the first vectors for the other five are proposed for their next versions.
The vector format
| Field | Meaning |
|---|---|
id | Unique within the specification, for example PF-3 |
spec | The specification slug |
level | The conformance level it belongs to |
case | One plain sentence |
input | The illustrative input (invented values) |
expect | The expected output, or the expected rejection |
since | The version that added it |
The vectors
| Specification | Id | Level | Case | Expected | Since |
|---|---|---|---|---|---|
| continuity-invariant | CI-1 | 1 | A change that keeps both halves of the Invariant | passes | proposed for 0.2 |
| continuity-invariant | CI-2 | 1 | A change that loses correspondence in any domain | fails | proposed for 0.2 |
| continuity-invariant | CI-3 | 1 | A change that drifts from purpose beyond tolerance | fails | proposed for 0.2 |
| proof-format | PF-1 | 1 | A pass with no losses | accepted | proposed for 0.2 |
| proof-format | PF-2 | 1 | A staged verdict without conditions | rejected | proposed for 0.2 |
| proof-format | PF-3 | 1 | A pass with a loss in one domain | rejected | proposed for 0.2 |
| proof-format | PF-4 | 1 | A corrected proof | refers to the old one | proposed for 0.2 |
| ledger-entry | LE-1 | 1 | An entry without a link to the previous one | rejected | proposed for 0.2 |
| ledger-entry | LE-2 | 1 | A correction | points back to the entry it corrects | proposed for 0.2 |
| ledger-entry | LE-3 | 1 | An entry with personal data | rejected | proposed for 0.2 |
| audit-method | AM-1 | 1 | A sample recorded after work started | rejected | proposed for 0.2 |
| audit-method | AM-2 | 1 | A re-run that does not match | produces a finding | proposed for 0.2 |
| staged-rollout | SR-1 | 1 | A move forward without a tested rollback | refused | proposed for 0.2 |
| staged-rollout | SR-2 | 1 | A failure at a stage | moves back one stage | proposed for 0.2 |
| staged-rollout | SR-3 | 1 | Live without an external audit | refused | proposed for 0.2 |
| ctu-metering | CT-V1 | 1 | A passing cycle on one device | One record; all fields present | 0.2.0 |
| ctu-metering | CT-V2 | 1 | A cycle rejected at the gate | One record; outcome reject; energy counted | 0.2.0 |
| ctu-metering | CT-V3 | 1 | A cycle that fails before the gate | One record; outcome failed; energy counted | 0.2.0 |
| ctu-metering | CT-V4 | 2 | Two cycles sharing a node | Two records; allocations sum to the node meter | 0.2.0 |
| ctu-metering | CT-V5 | 2 | Missing meter readings for part of a cycle | Record marked unverified; no silent estimate | 0.2.0 |
| ctu-metering | CT-V6 | 2 | A correction for a calibration error | A new record pointing back; the original unchanged | 0.2.0 |
| ctu-metering | CT-V7 | 1 | No CTU definition yet | Conversion: none; CTU empty; raw values present | 0.2.0 |
| ctu-metering | CT-V8 | 1 | A record with a personal name in it | Rejected | 0.2.0 |
All inputs are illustrative. No implementation has been tested against these vectors yet.
Implementation reports
You may publish which vectors your work passes, at which level and for which version. Such a report is your own statement. It becomes a statement by CRIS only after an audit under the audit method.
