Repository navigation
ci: balance offline module shards with normalized case clocks - #202
Merged
Merged
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Several newer test modules still used the default scheduling weight, leaving uneven offline groups. Refresh all 211 module hints using normalized case clocks from four accepted CI profiles. Preserve the deterministic four-group algorithm, every test and assertion, the existing concurrency limit and all native evidence gates.
Document the calibration provenance and its limits, and clarify that the historical synthetic wire corpus is separate from the later signed-wire custody candidate.
Validation: one fresh complete 4,322-test local regression passed with no skips or resource warnings, exact case counters, eight original streams, and artifact/privacy checks. Native input bytes are unchanged; no new standalone Go qualification is claimed. Fresh hosted PR and separate push/main acceptance remain required.
Recorded-clock models improve balance in all four calibration profiles and all four held-out same-source main profiles. A controlled wall-time speedup is not established. Application/core remain NO-GO and independent assessment remains unfilled.