Repository navigation
Releases
Brainwire v0.23.0-beta.2
Brainwire v0.23.0-beta.1
Brainwire v0.22.0-beta.4
Brainwire v0.22.0-beta.3
Brainwire v0.22.0-beta.2
Brainwire v0.21.0-beta.2
Brainwire v0.21.0-beta.1
Brainwire v0.20.0-beta.2
Experimental beta: evaluate with rebuildable data and a tested backup.
What's new
-
Accounted query growth now uses a sampled cgroup margin alongside RSS. Under pressure the server can refuse new query reservations; this is not a hard RSS ceiling. A local Linux A/B test verified rollback and retry in one short synthetic episode.
-
Label-qualified primary-key lookups avoid intermediate allocations. Timeline examples reject timestamp precision or timezone offsets that could be silently truncated.
-
DuckLake checkpoints now bind the source generation, mapping, graph, writer identity and endpoint policy. Existing checkpoints without the new contract are refused rather than silently reused.
-
Decimal values that cannot be represented exactly as floats are rejected; map text for exact values or explicitly choose DOUBLE for approximation.
-
A reproducible synthetic benchmark covers 1 million nodes and 2 million relationships on three engines, with raw receipts and restart checks. Brainwire results are medians of three runs of a locally built candidate; the published image is a separate build. Durability and client settings differ.
-
The README, agent quickstart, Colima guidance and documentation site are simpler. CI now exercises real DuckLake checkpoint races and incremental synchronization.
-
An authenticated Bolt regression checks that a stale ACL stamp in the same transaction conflicts with a concurrent revocation and that the revocation survives restart. Carrying a stale ACL to a new transaction can still restore access; reread it and handle conflicts.
Upgrade DuckLake synchronization
--source-identity (or source_identity=) and the ImportCheckpoint.contract_fingerprint ontology property are required. Stop and reconcile previous workers and uncertain commits before migrating. A new graph with a verified full reload is the safest starting point. Failed synchronization can leave earlier batches committed and its lease held. See parallel ingestion.
Evidence and limits
Local tests covered DuckLake concurrency, ACLs, process and quorum recovery, and short 512 MiB/2 GiB/3 GiB cgroup pressure scenarios. Separate 30-minute synthetic read and mixed read/write runs at 2 GiB stayed well below the limit and are not saturation tests. These do not establish a process-wide RSS ceiling or qualify distributed failover across machines. Cypher and search remain bounded, with no native ANN/BM25; native Bolt TLS is not implemented. See the benchmark report and current limits.
Brainwire v0.19.0-beta.1
Brainwire v0.19.0-beta.1
An experimental step toward a graph-backed context layer for applications and agents.
- Scoped context: narrow configured providers' ranked candidates to up to 100 explicit IDs before authorized reads and reciprocal-rank fusion. Providers are supplied by your application. The scope is not an ACL or a shared snapshot, and provider top-k can limit recall.
- Company context example: build a disposable synthetic graph of a team, project, decisions and sources; retrieve a project-scoped result and trace a source link. The bundled fixture is not text or vector search.
- A refreshed identity: cobalt network mark, light documentation site and a terminal palette update. The terminal workspace remains read-only.
- Operational checks: the candidate was exercised with access-control and local quorum tests and retained-volume recovery. See the repository journal and guides for scope and evidence.
Start with the quickstart and use a rebuildable graph and tested backup. Brainwire does not yet provide native ANN or BM25 indexes, historical graph revisions, graph branch/merge, a guaranteed global RSS ceiling or qualified multi-machine failover. Shared deployments require authentication and a verified encrypted boundary.
Brainwire v0.18.0-beta.6
Brainwire v0.18.0-beta.6
This experimental beta makes Python batch ingestion more predictable and expands recovery checks.
- Reject obviously oversized node, relationship and deletion items before making a complete JSON copy.
BatchTooLargeErrorremains compatible withValueErrorand reports the item type and configured limit. - Measure generated code-graph JSON in small groups rather than serializing a second complete graph just to check its size.
- Test crash recovery of two independent graphs on the same persistent volume, including rollback of uncommitted writes.
- Document explicit certificate-authority trust for Python runtimes that do not honor
SSL_CERT_FILEand exercise the supported configuration. - Add short, reproducible 512 MiB, 2 GiB and 3 GiB synthetic pressure-test receipts.
Run the image with the installation guide and read the current limits before using real data. The image is published only after the release workflow verifies both Linux architectures.
The client checks do not bound source, driver, server or process memory. Pressure results do not establish a global RSS ceiling or sustained production readiness. Multi-machine failover remains experimental and unqualified. Use a rebuildable graph and tested backups.