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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The match search stopped when the newest hash entry collided, missing valid older matches. Continue through the bounded chain.
On the seven tracked corpora, output falls 0.67% in total at level 6 and 2.35% at level 3. No size regressions across 70 corpus/level cases. This is a compression-ratio tradeoff: short level-6 JVM measurements take roughly 7–72% longer on the larger fixtures; Native also shows slowdowns.
Linux validation: 178 JVM tests pass, all 70 outputs decode with both JDK and KFlate, and all seven Native/Wasm benchmark corpora pass setup checks. Only KFlate was timed; saved Kompress data was reused. Raw samples, before/after sizes, reproduction helpers and per-corpus tradeoffs are in
performance/2026-09-09-match/RESULTS.md.Timing calibration: unchanged baseline text varied from 60.37 to 85.43 ms across three fresh forks. Initial single-fork text ratios are provisional; compressed-size results are deterministic.