An empty answer is not a measurement - #4465
Merged
Merged
Conversation
From the review log: `#4448 unmet=1 passed=4 failed=1 oracle=pass`. The bee had done the work, the generated Zig compiled and ran, four criteria of five were met, and the fifth read: - 4. `t27c spec-status specs/fpga/router.t27` prints `` (today: ) `spec-status` printed nothing at feed time - a stale symlink in the bee-bin directory is enough - and claims_hold compared the measured value with the claimed value, both the empty string, and let it through. The check that exists so no number is ever typed rather than measured could not see the one case where there is no number at all. Eight issues went out that way and every one came back sendBack at four of five, against an oracle that passed. Both feeders refuse an empty answer now, and the eight open issues were repaired in place by re-measuring: all eight now read `prints IMPLEMENTED`. Third place this week where an instrument could not tell "nothing" from "zero": the pusher (github_readable) this morning, the feeder (skips_are_fresh) at noon, this at three. Closes #4464 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
gHashTag
enabled auto-merge (squash)
September 20, 2026 15:43
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #4464
From the review log on the running deployment:
The bee did the work, the generated Zig compiled and ran, four criteria of
five were met — and the fifth was:
spec-statusprinted nothing at feed time.claims_holdcompared the measuredvalue with the claimed value — both the empty string — and let it through. The
check that exists so that no number is ever typed rather than measured cannot
see the case where there is no number at all.
Eight issues went out that way; every one came back
sendBackat 4/5 against apassing oracle. Both feeders refuse an empty answer now, and the eight open
issues were repaired in place by re-measuring — all eight read
prints IMPLEMENTED.Third place this week where an instrument could not tell nothing from zero:
the pusher (
github_readable) this morning, the feeder (skips_are_fresh) atnoon, this at three.
🤖 Generated with Claude Code