carlos / platform Public

Astra brief: tenant-directory fix 4 confirming

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Paul Campbell pushed by claude 1 week ago 333f13f0823adccc07d931a95f8016ff88dfec48 parent 400a330

1 file changed, +23 −0

docs/superpowers/plans/CODEX-REVIEW-REGION-HA-TENANT-DIR-BRIEF-3.md +23 −0

CI cancelled — superseded by a newer push — run details

CI log

=== gate ===
cd launchctx && go vet ./... && go test ./... -count=1
ok  	amadan.net/carlos/platform.git/launchctx	0.012s
?   	amadan.net/carlos/platform.git/launchctx/controlname	[no test files]
?   	amadan.net/carlos/platform.git/launchctx/wireid	[no test files]
go vet ./...
gofmt -l .: clean
go test ./... -count=1 -timeout 30m

amadan: job cancelled — a newer push superseded it

diff --git a/docs/superpowers/plans/CODEX-REVIEW-REGION-HA-TENANT-DIR-BRIEF-3.md b/docs/superpowers/plans/CODEX-REVIEW-REGION-HA-TENANT-DIR-BRIEF-3.md

new file mode 100644

index 000000000..1f69cb3a4

--- /dev/null

+++ b/docs/superpowers/plans/CODEX-REVIEW-REGION-HA-TENANT-DIR-BRIEF-3.md

@@ -0,0 +1,23 @@

+# Astra: confirming review of tenant-directory fix 4

+Your round 2 (`CODEX-REVIEW-REGION-HA-TENANT-DIR-2.md`) found three things. P2 is answered in 400a3302; the two P1s are filed.

+**P2: the writer/reader cap mismatch.**

+- The 256-component cap is dropped.

+- The single 4095-byte path cap (`withinTreeCap`) is now enforced in every publishing path (`Stage`, `Commit`, `publishOne` and the watcher) and in every reader.

+- The watcher skips an over-cap path for good and logs it once per run.

+- A 257-component path now round-trips from publish to restore.

+**The two P1s are filed as #279, an activation gate.** They are pre-existing in main's `hostfiles`:

+- `RemoveAll` has no bound;

+- the walk's width is unbounded.

+The implementer added a third to #279: `Restore` is O(D³) in depth. These are not fixed in stage B, because relocation is dormant until they are.

+**Judge:**

+1. Is P2 closed? Do the writer and the readers now agree on every path?

+2. Can the watcher's skip lose a legitimate file silently, or wedge publishing?

+3. Is the #279 deferral acceptable, given that relocation stays dormant until #279 is fixed?

+4. Anything new?

+Give findings as P1/P2/P3 with file:line, then a verdict.