Release version v1.1.15

Bug Fixes

This release focuses on stability improvements, enhanced tooling, and better development workflows.

• Capped references
• Fixes issues with edge:references
• Enhanced subscription functionality
• Improved CLI tooling
• Enhanced logger functionality with performance tool for benchmarking purposes
• Improved type definitions and error handling
• Code cleanup and optimization across multiple files

Complete Monthly Changelog: v1.1.4...v1.1.15

Release version v1.1.4

v1.1.4

(Nov 7th 2025)

High performance DB subscriptions

• Integration in modify core
• Id based subscriptions + include
• Filters based subscriptions
• Sort (integrated within the separate sort command)
• Time based subscriptions - evaluate "now" syntax
• Greedy type subscriptions

Partials

• Improved reference(s) implementation (step 2 of partial)
• Space requirement improvements
• Partial loading of large databases (Experimental) better memory and concurrency management

Edges

• Support all property types in edges
• Added edge aggregation
• Fixed an issue where edge props were not always saved properly

Schema

• User- sortable references as an option on the schema
• Full validation for all string formats (email, creditcard, etc)
• Dedicated validate function to validate payloads
• Fixed issues with one- sided edges
• Removed role and display properties from schema

Misc

• Added capped option for circular ids for nodes per type
• Modify individual errors
• Fixed bug with alias + filter queries
• Handle body data correctly for PUT and PATCH http methods
• Fixed bug with expire syntax + dependent
• Fixed bug with edges and migrations
• Refresh .expire()
• Improved load/save speed for references
• Removed weak references support
• Server/client integration improved performance by allowing sending and receiving of chunks on the client, avoiding memory copies

Breaking changes

• Removed display, role and readOnly option in @based/schema

Release version v0.1.3

(Oct 1st 2025)

• Optimized Server client integration
  • Complete protocol package for each format of based server / client e.g. response formats
• Improved reliability of client-server communication by replacing the current get and set processes with atomic upserting
  • Improved reference(s) implementation (step 1 of partial)
  • edge relations as true node types
• update reference upsert behavior to support defaults & hooks
• New result type for aggregations
• Fix: Bug with expire
• Aggregation with multiple distinct function types
• Fixed first letter bug in text search
• Fixed bug with edges on both sides
• Optimized and reimplemented DB read protocol for browser (40% smaller and fast)
• Implemented REST fallback protocol (excluding channels) (based/server)
• Added onClose option on context.session in the server
• Added support for all http methods in (based/server)
• Option to bind data from the session context in basedQueries (based/server)
• "Insert" option for upsert (only create if alias does not exist)
• Cardinality property migration fix
• Cardinality resource consumption optimization
• Improved Cardinality API - Precision and sparse/dense mode selection
• Order main buffer for better alignment
• Complete schema infer type
• Improved db hook system: Property hooks
• Implemented individual response protocol for modify operations (returning a response per modify)
• Meta information for last-ids (in zig, stored in common instead of schema)
• Fix regression: migrate should skip incompatible values
• Fix: load from dump with node using schema.dependent
• Fix: Incorrectly including drain time in query exec time measurement in .inspect()
• Unified @based/sdk acting as a single point for all other based packages. Consumers only need to install and import a single package @based/sdk.
• New centralized versioning system
  • Automatic cohesive updates across all suite of packages
  • Start of CI/CD

Based DB v0.1.1

🚀 v0.1.0 (Sep 2025)

Include

• 20% faster performance from internal optimizations
• New API feature: Meta information for adding contextual information like crc32, size and checksum.
• New API feature: "End option" allowing getting the start of a property, by default counts characters for strings and text including multibyte characters.
• Read is now fully supported in the @based/client, removing the need to call .toObject() when returning data from based functions
• Reading and deserializing database results is now 30% faster.

Query

• Id on a based query response now returns the id of a node if the target is single node
• Added queryId that is the

Backups

• Backup improvement. IoWorker threads for non blocking saving and loading (huge improvement)
• Start of partial loading of backups (works for types without references)
• Improved backup checksums

Modify

• Refactor of all client code, much cleaner and easier to understand codebase
• Offset system in place allows offloading validation & modify buffer preparation to db-client
• Growing preallocated modify buffer (saves memory if a low amount is used)
• Better support for async reference
• Better migration using native lastIds where applicable
• Better validation for most types - works better with customValidators as well
• New api feature: Standardizes modify result: Promise

Aggregations

• Date/time & intervals in Aggregations: Aggregate functions now supports date, time, and interval data types.
• Grouping by time windows, named intervals (day, week, ISO week, ISO year, etc) with timezone support
• Group by reference node IDs
• BaseType options for 'vector' (8/16/32-bit integers and 32/64-bit floating-points)
• New harmonic_mean aggregation function
• Improved [stddev and variance option to compute on sample or population mode

Interoperability

• New package: Export to CSV
  • We're introducing a new dedicated package for generating compliant CSV files from dumped datasets

Schema

• Infer typescript utility to infer types from the schema
• Introduced Schema Migration format:
  • Allows you to define custom transformations between schema versions making it easy to migrate
• New API feature: Schema hooks
  • Allows overwriting of each method for specific nodes
    e.g. url docs

General Fixes & Improvements

• Fixed stability issues with .expire()
• Fixed condition length of var string filters
• Fixed out of bounds read for an edge case with search
• Fixed incorrect timestamp validation, not allowing dates before 1970
• Fixed incorrect reading of references data for certain edge cases
• Fixed incorrect length encoding of filters on cardinality (from 2 to 4 bytes)
• Fixed defaults for enums are now handled correctly
• Fixed arbitrary unloading of WorkerCtx by napi external
  • solves filtering on compressed data (avoids segfault)
  • solves db process getting stuck when the db variable goes out of scope in js
• Fixed all handling of modify cursor mismatches on error

Documentation & Examples

• Start of documentation can be found here
• Northwind sample dataset - comparing SQL to BasedDb

Based Platform

• Alpha version of new environment hub & cli
  • Builds and watches large projects
  • Reconnects
  • Uses based-dbn on new env-hub this will go further in the future
• Launcher for Linux testing on MacOS
• Start of @based/protocol will centralize all protocols in 1 place
• @saulx/utils is replaced with @based/utils use this in the future
• @saulx/hash is replaced with @based/hash use this in the future

🚨 Breaking changes

• *Changed the $count result in aggegrations to count
• Filter operation has has been renamed to ìncludes. As shown in the docs
• Removed transform opton from schema properties in favor of Schema Hooks
• Removed upsert option from references modify api in favor of newly created db.upsert command. See the docs
• BasedQueryResponse now uses id for node ids vs the queryId
• Don't allow setting a string to text props without a locale

Based DB v0.0.71

🚨 Breaking Changes

• Schema version bump to 5.0.0-alpha.27 (ed077bd5f)

🚀 Features

• New CLI Tool:
  • Initial CLI implementation with login/logout (4f419bc7a, 74e40e4b0, 5ca200d0c)
  • Deploy command scaffolding (81025f932, 1266c051c)
  • Watch/build system integration (55104428b, 1672d3f02)
• Aggregation Enhancements:
  • cardinality aggregation fixes and tests (527d2f66e, 4e03b4307, b7a203f3a)
  • hll_union implementation (30df23317, 41dfeebf4)
• Schema & Data:
  • Support path: null in schemas (179d698c9)
  • New set function for data manipulation (271a0d4ae)

🐞 Fixes

• Serialization:
  • Fixed integer serialization bug (5d7460acb)
  • Hash (big number) serialization test (999b35e7d)
• Aggregations:
  • Fixed result placement for cardinality (5f881a060)
  • Validation error handling (527d2f66e)
• Edge Cases:
  • Default edges handling (b65703a1a)
  • Transform test fixes (f4a83aba6)

🛠 Refactoring & Maintenance

• Build system improvements:
  • Dynamic imports (13da7264d)
  • Concurrent watch scripts (7da428c85)
• Code cleanup:
  • Removed debug logs (b70551bf2, 4d3ef51f9)
  • TypeScript and formatting (4a0ec71c7, eceadc405)

⚡ Performance

• Fixed cardinality aggregation performance in group-by (b7a203f3a)

---

Note: This release introduces a new CLI tool and significant schema changes (v5 alpha). Developers should test thoroughly with the new schema version.

Based DB v0.0.70

Here’s the release notes for DB v0.0.70 schema 5.0.0-alpha.26:

---

🚀 Features

• Timestamp Control:
  • Allow overwrite of create/update timestamps (2291e111c)
• Schema Migrations:
  • Update schema transform during migrations (c457da958)

📚 Documentation

• Added Text/String field documentation (896515cc1)

📦 Dependencies

• Added fs-extra package (267574caf)

---

Note: This is a minor release focusing on timestamp flexibility and schema migration improvements.

v0.0.69

Based Db v0.0.69:

Warning: This release doesn't support node v21.*.

🚀 Features

• Transforms & Filters:
  • Added text/filter transforms (ac8a6bf80, 812b4c000, 0aca984c9)
• Defaults & Schema:
  • Support for default values in schema (786ce4e28, 68d4467ad, 9083074d1)
  • Async schema save on changes (467829b86, 9c00d3555)
• Performance & Optimization:
  • Slight serialization optimizations (9eb4ccd97, 5140e8843)
  • Better block verification (04f09ea28, 6ff75bfee)

🐞 Fixes

• Schema & Saving:
  • Fix race condition in save() (0e8e61434)
  • Only save if path exists (29304e68b)
  • Handle schema edge cases (467829b86, a13382683)
• Worker & Threads:
  • Graceful shutdown of DbWorker/IoWorker (5bff37007, f682af8c2, fc27b4cb5)
  • Fixed async termination (33cf41cb1, 1f5ad5f1a)
• Misc:
  • Revert to var pointers for stability (4582295e1, df30e3517)
  • Debugging fixes (8ceac38d7, 0615d9465)

🧹 Refactor & Cleanup

• Codebase:
  • Removed deprecated selva_fields_set() (b50f179f2)
  • Cleaned unused dependencies (4594d7329, 4903267ed, 4a105b65b)
• Documentation:
  • Added aggregation API docs (9d57897b4)
  • Improved doc comments (8d7a1a7e0, 72f05f828)

📦 Infrastructure

• Build & Dependencies:
  • Node.js v22+ requirement (029d61eb5)
  • Removed Turbo/analytics packages (4903267ed, 4a105b65b)
  • Fixed build scripts (9083074d1)

⚡ Performance

• Increased aggregation test threshold (92261d6f9)
• Optimized block handling (e4dbeb9f4, 6c4204d50)

---

Full Changelog: See commits.

---

Icons: 🚀 Feature, 🐞 Fix, 🧹 Refactor, 📦 Infrastructure, ⚡ Performance

DB Release v0.0.68

🚀 Release v0.0.68

Since packages are combined in the main Based repo this release still refers only to DB package changes since v0.0.66 (0.0.67 was deprecated due to Linux incompatibility).

🚀 Features

• GroupBy Enhancements:
  • Faster GROUP BY operations (1c8dc7400)
  • Support for ENUM types in references (12fbad6c9, b801bf2fa)
  • Cardinality blueprint and stddev aggregation (4306bd80e, e33cdf16c, 76cf72812)
• Numeric Types & Aggregations:
  • Improved handling of numeric types (564f24395, 5d917f757)
  • Aggregation fixes for min, avg, and f64 reading (5654f2885, 865c71612, fd0a8b0dc)
  • Descriptive stats functions (56e142e10)
• Performance & Scalability:
  • Partial block loading (c9d24f7c7, 331e84420)
  • Configurable BLOCK_CAPACITY in schema (2802ef8b8)
  • Optimized colvec support (d9413f864, 08cd4e890, f63e85707)
• Worker Improvements:
  • Generic Worker class for extensibility (2d2b5a2f5, 1f44402eb)
  • Better IoWorker job handling (d279cb782, 37f35b007)

🐞 Fixes

• Memory/type fixes: uint8array in deepCopy (e1dac9670), memcpy/memset implicit declarations (e06064b45)
• Schema/loading: Fix schema.bin usage (91e4a515e), block unloading (8c0b7495f)
• Aggregation/reference bugs: Result sizing (64e40e7c6), misplaced current_value (625cde407)
• Fallbacks: Text search (251daab29), filter handling (68b372128)

🧹 Refactor & Cleanup

• Replaced csmt with simpler struct (bae2353ca), renamed merkleTree to verifTree (63f33e5d7)
• Deduplicated finalization logic (7afecb7db), cleaned worker initialization (9b400cd73)
• Removed unused code (a339ed4df, e33cdf16c)

📚 Docs & Tests

• Added doc comments (178960680, 3c70bf10d)
• New tests: Enum types (15a2d70a6), numeric types (7afecb7db), aggregation references (4b179a770)

⚡ Performance

• Faster refs handling (946330c2a), optimized dirty range checks (0cde0c048)

Db v0.0.67

🚀 Release v0.0.67

Since packages are combined in the main Based repo this release still refers only to DB package changes since v0.0.66.

🚀 Features

• GroupBy Enhancements:
  • Faster GROUP BY operations (1c8dc7400)
  • Support for ENUM types in references (12fbad6c9, b801bf2fa)
  • Cardinality blueprint and stddev aggregation (4306bd80e, e33cdf16c, 76cf72812)
• Numeric Types & Aggregations:
  • Improved handling of numeric types (564f24395, 5d917f757)
  • Aggregation fixes for min, avg, and f64 reading (5654f2885, 865c71612, fd0a8b0dc)
  • Descriptive stats functions (56e142e10)
• Performance & Scalability:
  • Partial block loading (c9d24f7c7, 331e84420)
  • Configurable BLOCK_CAPACITY in schema (2802ef8b8)
  • Optimized colvec support (d9413f864, 08cd4e890, f63e85707)
• Worker Improvements:
  • Generic Worker class for extensibility (2d2b5a2f5, 1f44402eb)
  • Better IoWorker job handling (d279cb782, 37f35b007)

🐞 Fixes

• Memory/type fixes: uint8array in deepCopy (e1dac9670), memcpy/memset implicit declarations (e06064b45)
• Schema/loading: Fix schema.bin usage (91e4a515e), block unloading (8c0b7495f)
• Aggregation/reference bugs: Result sizing (64e40e7c6), misplaced current_value (625cde407)
• Fallbacks: Text search (251daab29), filter handling (68b372128)

🧹 Refactor & Cleanup

• Replaced csmt with simpler struct (bae2353ca), renamed merkleTree to verifTree (63f33e5d7)
• Deduplicated finalization logic (7afecb7db), cleaned worker initialization (9b400cd73)
• Removed unused code (a339ed4df, e33cdf16c)

📚 Docs & Tests

• Added doc comments (178960680, 3c70bf10d)
• New tests: Enum types (15a2d70a6), numeric types (7afecb7db), aggregation references (4b179a770)

⚡ Performance

• Faster refs handling (946330c2a), optimized dirty range checks (0cde0c048)

4.0.4

• Support cluster: 'local' in connect options
• Handle invalid payload in error handler