docs(tools): X-82 re-measured on the fixed build, and a lead that is stale by dates - #266
Merged
Conversation
…stale by dates One lane-B combat run on 502d45a -- the first build carrying the X-90 capsule lift. Zero probe=MISS across the whole run, no wire-volume kill, no [spawn] warning, and all three placements logged the lift. The only [fall] event was two ticks of a body that LANDED: probe=HIT dist 1.664 then 1.239 against 'Terrain 0,0', the gap closing and the report stopping. That is a drop off a ledge, not this row's signature, which is a body with no surface under it falling to y = -5344. THE ROW STAYS OPEN, and the reason is arithmetic rather than caution. The row itself puts the event at ~1 in 30 runs across four map locations. One clean run against a 1-in-30 base rate neither reproduces nor refutes; recording it as evidence of a fix would be closing a row on a coincidence, which is the thing this row's own remark warned about when it refused to ship a speculative repair beside its detector. THE TEMPTING READING IS NAMED AND NOT TAKEN. X-90 buried half the capsule, a CharacterController de-penetrates a buried capsule, and a lateral component of that push is a plausible mechanism for the 1.6 m an idle body travelled before falling. Plausible is not measured. The X-90 work's own end-to-end evidence cannot be re-checked from this checkout either: /artifacts/ is gitignored and predict-01 / lift-01 do not exist on this machine, so "zero [fall] lines in all four logs" is a claim about another machine's tree, recorded as such. THE LEAD THIS ROW NAMES IS STALE, established by dates rather than by argument. x64-after-02/server.log -- the measurement the row leans on -- is dated 2026-09-03. X-71, "the server walks a claimed player body", was CLOSED 2026-08-31. The walking defect was already fixed when the fall was measured, so "see X-71" cannot be the owed question. What is owed is unchanged in substance and smaller in shape: how a body arrives somewhere with no surface under it, on a build where nothing walks it and nothing buries it. Roll-up unchanged and verified: recount_debt_ledger.py --check agrees with the file.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
X-82 re-measured on
502d45a, the first build carrying the X-90 capsule lift. Therow stays open, and the reason is arithmetic rather than caution.
The run
probe=MISS— this row's signature[fall]eventprobe=HIT dist=1.664→1.239,collider='Terrain 0,0', gap closing, report stops[spawn]warnings+ 0.90 m capsule liftA drop off a ledge, not a body falling to y = −5344 with nothing beneath it.
Why one clean run closes nothing
The row itself puts the event at ~1 in 30 runs across four map locations. One
clean run against that base rate neither reproduces nor refutes. Recording it as
evidence of a fix would close a row on a coincidence — which is precisely what this
row's own remark warned about when it refused to ship a speculative repair beside
its detector.
The tempting reading, named and not taken
X-90 buried half the capsule; a
CharacterControllerde-penetrates a buriedcapsule; a lateral component of that push is a plausible mechanism for the 1.6 m an
idle body travelled before falling. Plausible is not measured. The run that
would separate it has not been done.
The X-90 work's own end-to-end evidence also cannot be re-checked from this
checkout:
/artifacts/is gitignored, andpredict-01/lift-01do not exist onthis machine. Its "zero
[fall]lines in all four logs" is a claim about anothermachine's tree, and the row now records it as such.
A lead that is stale by dates, not by argument
x64-after-02/server.log— the measurement this row leans on — is dated2026-09-03. X-71, "the server walks a claimed player body", was CLOSED
2026-08-31. The walking defect was already fixed when the fall was measured, so
see X-71cannot be the owed question any more.What is owed is unchanged in substance and smaller in shape: how a body arrives
somewhere with no surface under it, on a build where nothing walks it and nothing
buries it.
Roll-up unchanged;
recount_debt_ledger.py --checkagrees with the file.