We hope you had a wonderful summer! Our beautiful database sapling continues to grow along nicely. In the last month, we worked on pgvector compatible vector support and PostgreSQL replication for our upcoming alpha release.
But that is not all that is new:
What’s new at CedarDB
We have launched a community Slack:
To make it easier for us to interact with you, our users, and answer your questions, we have created a community Slack channel. Feel free to join and reach out to us there!
We kept busy with our Blog:
- We hit it off with an in-depth discussion of Colibri, our storage engine, optimized for both, fast scans and fast writes. If you’re interested in the nitty-gritty details, you can also take a look at the paper this blog post is based on, presented at this year’s prestigious VLDB conference.
- We then wrote an overview highlighting how CedarDB Reclaims SQL’s Declarative Power. If you want to know what makes our query optimizer so special and how it can help you write easier to understand SQL queries, this is the blog post for you!
- Finally, we seem to have stirred up a hornet’s nest with our opinion piece on Why We Prefer Exceptions to Error Values, which has sparked up many interesting discussions in the broader programming languages community.
Database Changes
New Features & more
- CedarDB now supports vectors and is compatible to pgvector! Read our new tutorial on vectors and learn why the concept most dissimilar to CedarDB is Counterinsurgency.
- We have substantially improved cardinality estimation for very selective predicates. This leads to over 10x speedups for some queries.
- We have implemented a heuristic that inlines Common Table Expressions that materialize a large state but are cheap to recalculate. This improves TPC-DS Q95’s performance by about 10x, for example.
- To further improve performance on time series data, we have taught our optimizer to use range joins in many scenarios where it would previously have used a slower variant.
- Finally, we introduced a lot of PostgreSQL compatibility features to make it even easier to use CedarDB with your existing toolchain. Features include to_regtype(), pg_typeof(), or a Postgres specific syntax for the trim() function.
Past Events
US Visit
Moritz and Chris visited Silicon Valley, Stanford University and the German Accelerator in Palo Alto. In addition to learning about the specifics of U.S. law and business practices and networking with local founders, they had the opportunity to meet face-to-face with investors, angels and fans of CedarDB.
Thank you!
That’s all for today. We’re looking forward to sharing more awesome progress in our next newsletter! Spoiler: I’m currently reviewing a pull request for as-of joins.
Until next time!
Your CedarDB Team