Repository navigation
CLAUDE.md: the first checked exception to "the cache is always one of the paths" - #699
Merged
Merged
Conversation
… the paths" #691 and #696 were both checked the way that rule demands -- each repro run three times against a warming cache with .nbc files confirmed present, plus three runs in one NodusRuntime and a second runtime in the same process. Stable every time. Recorded because the rule is stated absolutely, so the next reader would otherwise re-derive this, and because the reason generalises: the cache broke #394 because that fix MARKED something, and a mark survives compilation but not serialization. #691 and #696 mark nothing -- they resolve at call time from objects that are live either way. A resolve-don't-mark fix has nothing to serialize, so it cannot have a cache-shaped sibling path. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EdYjdCRLuedehwRCdi2hXb
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A follow-up to #695 and #698, recording a checked negative.
CLAUDE.md states the cache rule absolutely — "the bytecode cache is always one of the paths" — on the strength of #521, #400 and #394. Both closure/chunk fixes were checked the way that rule demands and both came back clean:
.nbcfiles confirmed present in.nodus/cache;NodusRuntime, exercising the module memo (BUG-457: ModuleLoader.compile_only silently returns the first source's bytecode when module_name is reused #457's shape);NodusRuntimein the same process.Stable every time, in both directions.
Recorded because the reason generalises, not just because it passed. The cache broke #394 because that fix marked something, and a mark survives compilation but not serialization — so the bypass came back on run 2. #691 and #696 mark nothing; they resolve at call time from objects that are live either way (frames, and the module's own
functionstable). A resolve-don't-mark fix has nothing to serialize, so it cannot have a cache-shaped sibling path.Without this note the next reader re-derives it, since the rule as written admits no exceptions. The note keeps the instruction to spend the second run — knowing why it passed is the point.
Docs only;
nodus_gate --staticand--versionsgreen.🤖 Generated with Claude Code
https://claude.ai/code/session_01EdYjdCRLuedehwRCdi2hXb