Roadmap

OmniSciDB is being intensively developed and is evolving quickly. In order to help the developer community understand our near term priorities, here is a roadmap of work targeted for the next 3-6 months. Many of these items are the result of community requests from developers, like you.

Database

Completed

• Top k sort on GPU. Ability to efficiently order by a projected column (Completed 3.2.2)
• JOIN improvements. Additional join types and performance improvements (Completed 3.2)
• JOIN improvements. Table orderings and optimizations for larger numbers of tables (Ongoing improvements)
• TRUNCATE. Efficiently remove contents of table or part of table (Completed v3.2)
• Batched UPDATE and DELETE (Completed 4.0)
• Interval type, and TIMESTAMP_ADD, TIMESTAMP_DIFF (Completed v3.2)
• Table Sharding. Table sharding types in addition to the existing round robin (Completed v3.2)
• Table-level security, GRANT, REVOKE. Ability to grant and revoke table access to users/roles (Testing in 3.2.3, released in 4.0)

Upcoming

• Use of shared memory to accelerate certain group by operations
• More efficient memory layout for certain group by operations
• More efficient range joins
• Better usage of both GPU and CPU, depending on memory requirements of queries
• Create table as select (CTAS) improvements (performance, and handling variable length data)
• Update improvements, including handling variable length data, update via subquery, and more efficient columnar output
• Delete improvements, including immediate/delayed vacuum capability
• More granular timestamp support (to nanoseconds)
• Row-level User Defined Functions (UDFs)

GIS

Completed

• Geo Types. Point, Line, Polygon (Completed 4.0)
• Basic Geo functions (ST_Contains, ST_Distance) (Completed 4.0)

Upcoming

• Additional OGC geospatial types: Multi(Point|Line|Polygon)
• Additional OGC geospatial operators (ST_Intersects, ST_Within, ST_Area, ST_Perimiter, etc)
• Support for geometric constructors (ST_MakePoint, etc)
• Accelerated geospatial joins (with dynamic spatial hashing, not relying purely on brute force loop joins)
• Fixed length arrays for POINT datatype to conserve memory
• Geodatabase/geopackage import
• WMS support

Data Science/GPU Data Frame (GDF)/Apache Arrow

Completed

• Data frame as input to OmniSciDB. Ability to insert into a table from a data frame (Completed 3.2.4)
• PyMapD DB-API Python client
• Ibis backend for OmniSciDB

Upcoming

• Increase efficiency of Arrow serialization

Contributing

Developers are encouraged to contribute to this Open Source project to expand and enhance OmniSciDB capabilities. Check out our Contributing page.

Making your first contribution?

Check out contribution wishlist with highlighted issues that would make great first contributions to this project. We also checking out the Issues to see the latest updates or proposed content.

Contributor License Agreement (CLA)

In order to clarify the intellectual property license granted with Contributions from any person or entity, OmniSci must have a Contributor License Agreement ("CLA") on file that has been signed by each Contributor, indicating agreement to the Contributor License Agreement. After making a pull request, a bot will notify you if a signed CLA is required and provide instructions for how to sign it. Please read the agreement carefully before signing and retain a copy for your records.

Need Help?

If you have questions and would like to connect with the maintainers of this open source project, please visit the official online forum and community.