PacificDB Community · v1.1.1

A database connection.
Your first document.

Create a database, add a collection, and connect directly from Node.js, Python, Java, the CLI, or Workbench. Version 1.1.1 removes project selection from everyday workflows while preserving existing metadata and legacy APIs.

Start the quickstartUpgrade guide

What’s new in v1.1.1

Added

  • Direct PacificDB.connect(url) entry points for Node.js, Python and Java, with authenticated URLs, verified TLS and explicit connection lifecycles.
  • Python and Java document CRUD, bulk, aggregation, explain and named operation families.
  • Python and Java checksummed, bounded file uploads/downloads and full backup export, with explicit resume IDs.
  • Workbench database URLs and matching CLI/SDK connection examples.

Removed

  • Mandatory project selection from database and collection workflows.
  • Friendly project commands in native/npm shells and project screens in Workbench.
  • Stored project IDs and tokens from updated CLI context. Legacy engine and SDK project APIs remain available.

Improved

  • Database creation selects after success. Database switching checks the authorized catalog.
  • Java/Python clients pool connections, preserve request scope, expose typed errors and close pending work.
  • Fixed Node.js pool scheduling after sequential warmup: completed slots become available before caller promises settle, letting a concurrent burst use the configured connections. A peer regression verifies four concurrent requests with a four-connection pool.
  • Media transfer validates chunks and whole files; verified downloads replace destinations only after successful checks.
  • Workbench limits Overview count work and stops scheduling it when another view is selected.

Try the database-first flow

Install PacificDB 1.1.1, then run pacificdb. Use a fresh database name:

create database app
create collection users
insert users {"id":"1","name":"Ada"}
find users {"id":"1"}

For an existing database use use app or pacificdb --url 'pacificdb://127.0.0.1:9000/app' --no-start. A database URL speaks the engine protocol; browser Workbench uses its separate printed HTTP address.

Workbench v1.1.1

Workbench now lists databases directly, creates or selects a database without a project, and nests collections under that database. Existing project-mapped databases remain accessible without moving stored data.

The connection dialog requires a selected database and generates its database URL, native CLI command, and Node.js, Python, and Java examples. Authenticated examples read PACIFICDB_URL privately; credentials are omitted from copied examples.

Overview loads missing collection counts through one loader with at most four jobs in flight, caches results per database, invalidates affected summaries after mutations, and displays a dash when totals are incomplete. Documents, Query, and Media views do not initiate full collection count scans.

The Debian installer registers the launcher and configures the Chromium sandbox helper. Windows, macOS, Debian, and portable Linux artifacts are available from the Workbench v1.1.1 release.

Compatibility and limits

No project/data migration is required. Legacy createProject/useProject and raw project/catalog actions remain available for existing consumers. Creation is explicit and not idempotent. Use matching 1.1.1 SDK packages for new convenience methods.

Interrupted writes can have unknown outcomes and are not retried automatically. File transfers use explicit resume IDs; practical file limits remain disk, network, and request constraints. Workbench uploads have a 64 MiB UI limit. Community indexing remains non-unique, single-field B-tree indexing.

Use Node.js, Python, Java, CLI, and Workbench guides for install, authentication, CRUD, vectors, media, closing, and troubleshooting.