As a maintainer of open source projects it's not always possible to support feature requests instantly. We were just getting started on supporting Vector databases and embeddings for March release. However, when one of the collaborators came with a pressing need to support Cloudflare D1 for a startup, we decided to take a look. Initially we were a bit hesitant, because D1 is meant to be used from Cloudflare Workers and Pages.
After initial analysis we figured out that it is possible to support this database and CRUD, in few days. The core framework is extremely pluggable now to support any database in just few days or even less. Really great work by the team. This was also an opportunity for us to refactor the core engine now to make it ready to support non-relational databases. I promise, that MongoDB support will arrive later this year.
What's Change
- Support for Cloudflare D1 - Query, Filter, Insert, Update, Delete.
- Refactoring the core to pave way for non-relational databases.
Whats's Next
Ok we are now back into simplifying AI/ML support around PostgreSQL with PostgresML and pgvector. Besides, this we will work on documentation, which is lagging. March 2024, is few hours away, knowing the team, we can deliver bit more as well....