Data and databases

The data/ directory contains .sql database migrations and a cache.json file with currently known asset files metadata.

SQLite Databases

For simplicity and portability, the project uses SQLite. The data/sql/ directory is embedded into the server binary at compile time, and migrations are run automatically at startup.

core.db

Schema: data/sql/core_db/

The main SQLite database. Contains the classified list of available assets and random names used in the sign-up scene.

Managed by goose using migrations in data/sql/core_db/updates/. Shortcuts for creating, applying, and resetting migrations are available in the Makefile.

An initial squashed migration (data/sql/core_db/base.sql) includes both schema and seed data.

Use data/sql/core_db/squash.sh to squash update migrations into base.sql.

Files under data/sql/queries/ are named example SQL queries available to the admin dashboard`s SQL Explorer.

blob.db

Schema: data/sql/blob_db/

A separate SQLite database storing the raw binary content of game asset files.

Downloaded on startup if blob.download = true and the database is missing.

When a client requests /cdn/{cdnid}, the server fetches the blob column for that cdnid.

The database can be populated using the blob-tool or the admin dashboard.

Cache Metadata

cache.json

A summary of known cache files, including name (CDN ID), size, type, hash, GSF OID and basic bundle info (version, object counts, scene roots).

Generated with the cache-tool Python script:

python tools/cache-tool/cache.py /path/to/cache/folder \
  --summary-file data/cache.json

This file is already included in the core.db base migration. For future updates (if new unique cache assets are discovered), add a new migration under updates/ with the additional data.