Release Notes
A patch release implementing a community-requested feature (#351): a reliable, cheap boot-time
check for whether a local DB is consistent with the WAL. Shipped as a patch alongside the
prerequisite global_seq re-derivation fix (#352) because the requesting deployment needs both
together.
Added
wal.applied_seq and wal.max_seq in knowledge_status, letting a client decide — from
one call and an integer comparison — whether its DB already reflects the WAL, needs an
incremental resume, or needs a full rebuild. applied_seq is a new piece of persisted DB
state (survives restart, not cached in memory) recording the highest WAL seq whose
mutations are committed in the current graph; max_seq is the highest seq actually present
across the WAL files, read fresh on every call so an externally-updated WAL is observed
immediately. applied_seq is null (unknown), 0 (nothing applied), or a positive integer
(a known position) — see docs/operations.md for the full consumer decision table and a
cross-language footgun around null in numeric comparisons. A pre-existing, populated DB that
predates this feature backfills a usable position on first open via
ADR-0026's retroactive episode-cursor mechanism,
rather than reporting null — except when that mechanism itself can't find the last episode's
uuid in the WAL, the one case where null remains the correct, documented report.
(#353, ADR-0353)
lcg-eval 0.12.2
Install lcg-eval 0.12.2
Install prebuilt binaries via shell script
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/verveguy/liminis-context-graph/releases/download/v0.12.2/lcg-eval-installer.sh | sh
Download lcg-eval 0.12.2
lcg-service 0.12.2
Install lcg-service 0.12.2
Install prebuilt binaries via shell script
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/verveguy/liminis-context-graph/releases/download/v0.12.2/lcg-service-installer.sh | sh
Download lcg-service 0.12.2
