This document outlines the future development plans for TuGraph, including features currently under development, those not in the development pipeline, and completed functionalities not included in the open-source version.
TuGraph aims to be an open-source, high-performance graph database. It adopts a centralized storage approach for graph data and, in the short term, does not consider data sharding. Instead, it employs a master-slave replication mode to address high-concurrency read scenarios, while utilizing cloud-based storage solutions to address storage capacity challenges.
TuGraph was open-sourced on September 1, 2022, and has received regular bug fixes and enhancements based on community feedback.
| Version | Functionality | Date |
|---|---|---|
| 3.3.0 | Initial open-source release | 2022.9.1 |
| 3.3.1 | Refactored graph analysis engine with multi-mode support | 2022.10.14 |
| 3.3.2 | Added OGM support and improved unit test coverage | 2022.11.21 |
| 3.3.3 | Iterative improvements to link authentication mechanism and addition of English documentation | 2022.12.23 |
| 3.3.4 | Cloud deployment support and streamlined LDBC SNB Audit process | 2023.1.28 |
| 3.4.0 | Added support for OLAP Python API and upgraded offline data import | 2023.3.11 |
| 3.5.0 | Introduced POG (Procedures On Graph query language), frontend upgrades | 2023.6.5 |
| 3.5.1 | Added learning engine, Procedure Rust API, and storage-property separation | 2023.7.14 |
| 3.6.0 | High availability support, log system upgrade | 2023.8.11 |
| 4.0.0 | ISO GQL support, 11 new open-source graph algorithms, m1 Docker support | 2023.9.6 |
| 4.0.1 | Support for temporal edge sorting, 5 new open-source graph algorithms | 2023.9.28 |
| 4.1.0 | Bolt protocol support, fast online full import, support for geospatial data types | 2023.12.25 |
In addition, TuGraph has established a comprehensive quality system, including automated unit testing, integration testing, and performance testing.
For more detailed information, refer to the "[root]/release/CHANGELOG.md" file in the source code directory.
In addition to the aforementioned core functionalities, the following components are planned for development in 2024.
| Version | Functionality | Planned Date |
|---|---|---|
| 4.2.x | HA support for Witness role and management tools | 2024.3 |
| 4.2.x | Bolt support for stream processing and parameterized queries | 2024.3 |
| x.x.x | GeaX support for Cypher | 2024.6 |
| x.x.x | Support for composite indexes | 2024.6 |
| x.x.x | Optimization of data import functionality | 2024.6 |
| x.x.x | [Community Feature] Support for geospatial data types | 2024.6 |
| x.x.x | Enhanced Cypher capabilities | 2024.9 |
| x.x.x | Support for rapid schema changes | 2024.9 |
| x.x.x | Vectorization support | 2024.12 |
| x.x.x | RPQ support | 2024.12 |
| x.x.x | [Optional] Query engine upgrade | 2024.12 |
| x.x.x | [Community Feature] Support for GraphAr | 2024.12 |
Currently, the development team's resources are limited, and we cannot implement all the desired features for TuGraph. However, during the feature planning process, we have identified a range of ideas worth exploring. The team has conducted some initial exploration, and we welcome community collaboration in developing the following features:
| Version | Functionality | Planned Date |
|---|---|---|
| x.x.x | Rich graph algorithm library | 2024.x |
| x.x.x | Support for default attribute values | 2024.x |
| x.x.x | Best practice of embedded TuGraph-DB usage | 2024.x |
| x.x.x | Support for transaction in Bolt clients | 2024.x |
| x.x.x | Support for composite data types like List, Map, and Decimal | 2024.x |
| x.x.x | Support of multiple storage engine | 2024.x |
For simpler features, we will label them as "good first issue" on GitHub issues, and we welcome discussions from technology enthusiasts interested in graph databases.