FAILURE MAP
← Case archive

FA-18591 / Time representation / Open access

Rollback filter measures elapsed in wall domain or hides monotonic regression · case 01

The decoded time state disagrees with the explicit regression oracle for elapsed.

Verified by executionVariant 1 · 11 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Rollback filter measures elapsed in wall domain or hides monotonic regression.

VERIFIED REPAIR

Preserve the declared coordinate and state contract at elapsed: elapsed = r['mono'] - r['last_mono'].

Unsuccessful approach: The partial correction still substitutes abs(r['mono'] - r['last_mono']) at the same fault site.

Case contract

A presentation clock preserves monotonic output while exposing raw wall coordinate and rollback debt. It advances a floor by monotonic elapsed since previous sample; publish max(raw wall,floor) if continuity mode enabled, otherwise raw wall. On explicit reset, discard old floor. Output adjusted coordinate, added bias, rollback marker, debt, source mode and next checkpoint. Negative elapsed invalidates sample.

Why this case matters

Clock transfer and timestamp consumers require preserved coordinate, phase, validity and elapsed-time semantics.

1 / The failure

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json
from fractions import Fraction
N = 1
observations = []
def solve(r):
    elapsed = r['wall'] - r['last_output']
    valid = elapsed >= 0
    floor = r['wall'] if r['reset'] else r['last_output'] + max(0,elapsed)
    rollback = r['wall'] < floor
    output = max(r['wall'],floor) if r['continuous'] else r['wall']
    bias = output - r['wall']
    debt = max(0,floor - r['wall'])
    mode = 'floor' if r['continuous'] and rollback else 'wall'
    checkpoint_output = output if valid else r['last_output']
    checkpoint_mono = r['mono'] if valid else r['last_mono']
    return [valid,output,bias,debt,mode,[checkpoint_output,checkpoint_mono],rollback]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('fixture 1', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 5, 5, 'floor', [110, 20], True])
check('fixture 2', solve({'mono': 20, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 120, 0, 0, 'wall', [120, 20], False])
check('fixture 3', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 2}), [True, 105, 0, 5, 'wall', [105, 20], True])
check('fixture 4', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': True, 'continuous': True, 'last_bias': 2}), [True, 90, 0, 0, 'wall', [90, 20], False])
check('fixture 5', solve({'mono': 10, 'last_mono': 10, 'wall': 100, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 100, 0, 0, 'wall', [100, 10], False])
check('fixture 6', solve({'mono': 9, 'last_mono': 10, 'wall': 101, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [False, 101, 0, 0, 'wall', [100, 10], False])
check('fixture 7', solve({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 0, 0, 'wall', [110, 20], False])
check('fixture 8', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 20, 20, 'floor', [110, 20], True])
check('fixture 9', solve({'mono': 30, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 5}), [True, 120, 15, 15, 'floor', [120, 30], True])
check('fixture 10', solve({'mono': 10, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 0}), [True, 120, 0, 0, 'wall', [120, 10], False])
variant = [({'mono': 20, 'last_mono': 10, 'wall': 106, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 4, 4, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 107, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 3, 3, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 108, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 2, 2, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 109, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 1, 1, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 0, 0, 'wall', [110, 20], False])]
check("variant capture", solve(variant[N-1][0]), variant[N-1][1])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
fixture 1[True, 105, 0, 0, 'wall', [105, 20], False][True, 110, 5, 5, 'floor', [110, 20], True]Failed
fixture 2[True, 120, 0, 0, 'wall', [120, 20], False][True, 120, 0, 0, 'wall', [120, 20], False]Passed
fixture 3[True, 105, 0, 0, 'wall', [105, 20], False][True, 105, 0, 5, 'wall', [105, 20], True]Failed
fixture 4[False, 90, 0, 0, 'wall', [100, 10], False][True, 90, 0, 0, 'wall', [90, 20], False]Failed
fixture 5[True, 100, 0, 0, 'wall', [100, 10], False][True, 100, 0, 0, 'wall', [100, 10], False]Passed
fixture 6[True, 101, 0, 0, 'wall', [101, 9], False][False, 101, 0, 0, 'wall', [100, 10], False]Failed
fixture 7[True, 110, 0, 0, 'wall', [110, 20], False][True, 110, 0, 0, 'wall', [110, 20], False]Passed
fixture 8[False, 100, 10, 10, 'floor', [100, 10], True][True, 110, 20, 20, 'floor', [110, 20], True]Failed
fixture 9[True, 105, 0, 0, 'wall', [105, 30], False][True, 120, 15, 15, 'floor', [120, 30], True]Failed
fixture 10[True, 120, 0, 0, 'wall', [120, 10], False][True, 120, 0, 0, 'wall', [120, 10], False]Passed
variant capture[True, 106, 0, 0, 'wall', [106, 20], False][True, 110, 4, 4, 'floor', [110, 20], True]Failed

SHA-256 / 18b067f9d98ca23627255a69c9658ac598912a1a42db1f56aca28f6a61ed0628

2 / The unsuccessful fix

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json
from fractions import Fraction
N = 1
observations = []
def solve(r):
    elapsed = abs(r['mono'] - r['last_mono'])
    valid = elapsed >= 0
    floor = r['wall'] if r['reset'] else r['last_output'] + max(0,elapsed)
    rollback = r['wall'] < floor
    output = max(r['wall'],floor) if r['continuous'] else r['wall']
    bias = output - r['wall']
    debt = max(0,floor - r['wall'])
    mode = 'floor' if r['continuous'] and rollback else 'wall'
    checkpoint_output = output if valid else r['last_output']
    checkpoint_mono = r['mono'] if valid else r['last_mono']
    return [valid,output,bias,debt,mode,[checkpoint_output,checkpoint_mono],rollback]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('fixture 1', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 5, 5, 'floor', [110, 20], True])
check('fixture 2', solve({'mono': 20, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 120, 0, 0, 'wall', [120, 20], False])
check('fixture 3', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 2}), [True, 105, 0, 5, 'wall', [105, 20], True])
check('fixture 4', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': True, 'continuous': True, 'last_bias': 2}), [True, 90, 0, 0, 'wall', [90, 20], False])
check('fixture 5', solve({'mono': 10, 'last_mono': 10, 'wall': 100, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 100, 0, 0, 'wall', [100, 10], False])
check('fixture 6', solve({'mono': 9, 'last_mono': 10, 'wall': 101, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [False, 101, 0, 0, 'wall', [100, 10], False])
check('fixture 7', solve({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 0, 0, 'wall', [110, 20], False])
check('fixture 8', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 20, 20, 'floor', [110, 20], True])
check('fixture 9', solve({'mono': 30, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 5}), [True, 120, 15, 15, 'floor', [120, 30], True])
check('fixture 10', solve({'mono': 10, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 0}), [True, 120, 0, 0, 'wall', [120, 10], False])
variant = [({'mono': 20, 'last_mono': 10, 'wall': 106, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 4, 4, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 107, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 3, 3, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 108, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 2, 2, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 109, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 1, 1, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 0, 0, 'wall', [110, 20], False])]
check("variant capture", solve(variant[N-1][0]), variant[N-1][1])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
fixture 1[True, 110, 5, 5, 'floor', [110, 20], True][True, 110, 5, 5, 'floor', [110, 20], True]Passed
fixture 2[True, 120, 0, 0, 'wall', [120, 20], False][True, 120, 0, 0, 'wall', [120, 20], False]Passed
fixture 3[True, 105, 0, 5, 'wall', [105, 20], True][True, 105, 0, 5, 'wall', [105, 20], True]Passed
fixture 4[True, 90, 0, 0, 'wall', [90, 20], False][True, 90, 0, 0, 'wall', [90, 20], False]Passed
fixture 5[True, 100, 0, 0, 'wall', [100, 10], False][True, 100, 0, 0, 'wall', [100, 10], False]Passed
fixture 6[True, 101, 0, 0, 'wall', [101, 9], False][False, 101, 0, 0, 'wall', [100, 10], False]Failed
fixture 7[True, 110, 0, 0, 'wall', [110, 20], False][True, 110, 0, 0, 'wall', [110, 20], False]Passed
fixture 8[True, 110, 20, 20, 'floor', [110, 20], True][True, 110, 20, 20, 'floor', [110, 20], True]Passed
fixture 9[True, 120, 15, 15, 'floor', [120, 30], True][True, 120, 15, 15, 'floor', [120, 30], True]Passed
fixture 10[True, 120, 0, 0, 'wall', [120, 10], False][True, 120, 0, 0, 'wall', [120, 10], False]Passed
variant capture[True, 110, 4, 4, 'floor', [110, 20], True][True, 110, 4, 4, 'floor', [110, 20], True]Passed

SHA-256 / d1e5add67beba1444d56ebb3a91abca1a322f43bddfb0b96a6c8feefc7819394

3 / The verified repair

Exit 0
"""Failure Map reference implementation. Python standard library only."""
import json
from fractions import Fraction
N = 1
observations = []
def solve(r):
    elapsed = r['mono'] - r['last_mono']
    valid = elapsed >= 0
    floor = r['wall'] if r['reset'] else r['last_output'] + max(0,elapsed)
    rollback = r['wall'] < floor
    output = max(r['wall'],floor) if r['continuous'] else r['wall']
    bias = output - r['wall']
    debt = max(0,floor - r['wall'])
    mode = 'floor' if r['continuous'] and rollback else 'wall'
    checkpoint_output = output if valid else r['last_output']
    checkpoint_mono = r['mono'] if valid else r['last_mono']
    return [valid,output,bias,debt,mode,[checkpoint_output,checkpoint_mono],rollback]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('fixture 1', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 5, 5, 'floor', [110, 20], True])
check('fixture 2', solve({'mono': 20, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 120, 0, 0, 'wall', [120, 20], False])
check('fixture 3', solve({'mono': 20, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 2}), [True, 105, 0, 5, 'wall', [105, 20], True])
check('fixture 4', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': True, 'continuous': True, 'last_bias': 2}), [True, 90, 0, 0, 'wall', [90, 20], False])
check('fixture 5', solve({'mono': 10, 'last_mono': 10, 'wall': 100, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 100, 0, 0, 'wall', [100, 10], False])
check('fixture 6', solve({'mono': 9, 'last_mono': 10, 'wall': 101, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [False, 101, 0, 0, 'wall', [100, 10], False])
check('fixture 7', solve({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 0, 0, 'wall', [110, 20], False])
check('fixture 8', solve({'mono': 20, 'last_mono': 10, 'wall': 90, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}), [True, 110, 20, 20, 'floor', [110, 20], True])
check('fixture 9', solve({'mono': 30, 'last_mono': 10, 'wall': 105, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 5}), [True, 120, 15, 15, 'floor', [120, 30], True])
check('fixture 10', solve({'mono': 10, 'last_mono': 10, 'wall': 120, 'last_output': 100, 'reset': False, 'continuous': False, 'last_bias': 0}), [True, 120, 0, 0, 'wall', [120, 10], False])
variant = [({'mono': 20, 'last_mono': 10, 'wall': 106, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 4, 4, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 107, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 3, 3, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 108, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 2, 2, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 109, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 1, 1, 'floor', [110, 20], True]), ({'mono': 20, 'last_mono': 10, 'wall': 110, 'last_output': 100, 'reset': False, 'continuous': True, 'last_bias': 2}, [True, 110, 0, 0, 'wall', [110, 20], False])]
check("variant capture", solve(variant[N-1][0]), variant[N-1][1])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
fixture 1[True, 110, 5, 5, 'floor', [110, 20], True][True, 110, 5, 5, 'floor', [110, 20], True]Passed
fixture 2[True, 120, 0, 0, 'wall', [120, 20], False][True, 120, 0, 0, 'wall', [120, 20], False]Passed
fixture 3[True, 105, 0, 5, 'wall', [105, 20], True][True, 105, 0, 5, 'wall', [105, 20], True]Passed
fixture 4[True, 90, 0, 0, 'wall', [90, 20], False][True, 90, 0, 0, 'wall', [90, 20], False]Passed
fixture 5[True, 100, 0, 0, 'wall', [100, 10], False][True, 100, 0, 0, 'wall', [100, 10], False]Passed
fixture 6[False, 101, 0, 0, 'wall', [100, 10], False][False, 101, 0, 0, 'wall', [100, 10], False]Passed
fixture 7[True, 110, 0, 0, 'wall', [110, 20], False][True, 110, 0, 0, 'wall', [110, 20], False]Passed
fixture 8[True, 110, 20, 20, 'floor', [110, 20], True][True, 110, 20, 20, 'floor', [110, 20], True]Passed
fixture 9[True, 120, 15, 15, 'floor', [120, 30], True][True, 120, 15, 15, 'floor', [120, 30], True]Passed
fixture 10[True, 120, 0, 0, 'wall', [120, 10], False][True, 120, 0, 0, 'wall', [120, 10], False]Passed
variant capture[True, 110, 4, 4, 'floor', [110, 20], True][True, 110, 4, 4, 'floor', [110, 20], True]Passed

SHA-256 / 2c419a3c37a871fe92299030de8da56be0bad1440e620cff4008e5f3bdf9b4de

Verification & scope

Deterministic integer reference model with stipulated units and policies; not a complete clock, wire standard or platform implementation. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:39:59.044147+00:00.

Case digest / b9b7c50cabe035292a8b255b7e2346b8dda83559d7784b28dacc31c1a82db8ef