At Timescale, they value both innovation and stability. To resolve this tension, they introduced a new experimental schema for their database, where they can develop and ship features at an even faster pace than before. This schema allows them to validate new functionality with real users and get feedback, while also setting expectations that these features are still experimental and may break. The first feature released in the experimental schema is time_bucket_ng, which adds support for bucketing by years and months, as well as timezone functionality for use in queries and continuous aggregates. This enables developers to analyze data over longer periods of time and provides more flexibility in their analysis. Additionally, TimescaleDB 2.4 introduces improved elasticity and high-availability features for multi-node deployments, allowing users to move chunks between data nodes and copy data across the cluster. The experimental schema also includes a function to clean up failed copy or move operations. To provide feedback and drive the project, community members are encouraged to try the experimental schema in a test-safe environment and share their experience.