Skip to content

What's New in Version 0.11.0

October 3, 2025 3:36 PM UTC

Version 0.11.0 of the relationalai Python package is now available!

To upgrade, activate your virtual environment and run the following command:

Terminal window
pip install --upgrade relationalai
  • When preparing data streams that require a Change Data Capture (CDC) engine, the progress display now clearly shows when the CDC engine is being resumed or provisioned. Instead of a generic “Preparing your data” message, you’ll see a dedicated progress sub-task like “Provisioning CDC engine…” that updates as soon as the engine is ready and data streams start processing.

    This improvement makes it easier to understand what’s happening during potentially long waits, providing better transparency and reducing confusion.

  • The CLI and notebook environments now feature an enhanced progress indicator system that provides detailed, real-time tracking of main tasks and subtasks during multi-step operations like data streaming, engine provisioning, and relation population.

    Key improvements include:

    • Hierarchical subtask tracking with clear visual feedback:
      • Subtasks highlight in yellow when updated.
      • Completed subtasks turn green and automatically disappear after a short delay.
      • Errors are prominently marked with icons and color coding.
    • Adaptive display that adjusts output for terminal (CLI) and notebook (Snowflake, Jupyter) environments to ensure optimal readability.
    • Clear success and failure states for both main tasks and subtasks, improving transparency and troubleshooting.

    These enhancements make it easier to monitor complex workflows, quickly spot progress updates or issues, and enjoy a consistent experience whether working in a terminal or notebook interface.

  • Improved the error message shown when change tracking is not enabled on a Snowflake table. The message now clearly explains both how to enable change tracking manually with a SQL command and how to enable it automatically by setting ensure_change_tracking = true in your configuration or model. This makes it easier to resolve change tracking issues and choose the best approach for your workflow.

  • Improved the robustness of engine creation by handling cases where internal API responses are None or invalid. This prevents errors during simultaneous creation of CDC and user engines, ensuring smoother and more reliable engine provisioning.

  • Improved the robustness of the spinner polling logic by safely skipping over None entries in the polling data. This prevents unexpected crashes caused by incomplete or malformed JSON data returned by internal APIs, resolving errors like:

    AttributeError: 'NoneType' object has no attribute 'lower'

    With this fix, queries are more reliable and less likely to fail due to upstream data issues.

  • Improved the internal retry logic for network connections to catch all exceptions, including those from the requests library (such as connection aborted or reset errors). This enhancement ensures that transient network issues are properly retried, making network operations more reliable and reducing failures in automated workflows.

  • Updated the internal lqp Python package dependency to version 0.1.16. This change enhances internal validation clarity for relationalai package developers but does not affect any user-facing features, APIs, CLI commands, or Snowflake SQL Library functions.