Release Notes 6.12.0
This technical version delivers significant improvements to performance and consistency, helping teams manage data more efficiently at scale. This release focuses on backend refinements that improve maintainability and data accuracy—particularly around unit management and aggregation logic.
Project-Specific Units: Units are now managed within individual projects to improve performance and reduce clutter.
Improved Aggregation Behavior with
SHIFT
and Time Dimensions: Consistent application of the SUM aggregation to ensure more predictable and accurate results.Stable Ordering for
FIRST
andLAST
Aggregations: Aggregation logic now follows a deterministic sort order for consistent results.
Project-Specific Units
Units have been restructured to belong to individual projects, significantly reducing clutter and improving performance in instances with long histories or many unused units. During the upgrade, all units currently in use are safely retained and reassigned to the appropriate projects. Only unused, non-default units are cleaned up automatically to keep your environment lean and efficient. Default units, such as Euro, USD, Currency, Percent, and Quantity will automatically be included in every new project. For a complete list of the default units, please check out the Configure Units page in our documentation.
The units overview page in the admin UI has also been optimized. Project admins can now manage units directly if they hold a modelling license. This change improves data governance and boosts performance across modeling environments.
Improved Chart Rendering Performance
Rendering performance has been significantly improved—especially for large grid charts, data input charts, and line-item tables. This results in noticeably faster load times and smoother tab switching across the workspace. In one example, switching tabs with large datasets (30k+ rows) now takes less than a second, compared to up to 30 seconds previously. All chart types benefit from this optimization, making navigation and interaction much more responsive.
Improved Aggregation Behavior with SHIFT and Time Dimensions
When using the SHIFT
function with time dimensions that result in implicit aggregation (e.g., shifting from a 31-day to a 30-day month), the system now consistently applies the SUM aggregation. This ensures more predictable and accurate results.
Stable Ordering for FIRST and LAST Aggregations
The behavior of FIRST
and LAST
aggregations has been made deterministic by introducing a defined column sort order. Columns are now sorted by alphabetical dimension names and internal level order, eliminating inconsistencies across environments. This ensures that aggregation results are stable and predictable, even when working across projects or node structures.
As part of version 6.12.0, improvements have been made to enhance performance, accuracy, and long-term maintainability. In rare cases, these changes may affect numerical results—particularly in complex aggregations. If you notice any unexpected differences, please reach out to our support team. We're here to help and ensure your models run as expected.
Other Improvements
Calculation Engine Performance Improvements: Enhanced memory management and internal optimizations improve the efficiency and stability of calculations, especially when working with large datasets or models with many levels.
Snapshot and Presentation Support: Adjusted to align with new unit scoping.
Project Export/Import Logic Updated: Includes scoped units for clean, self-contained project transfers.
Security Updates: Updated Handsontable and Bootstrap to their latest versions to address known vulnerabilities and improve compatibility.
Various bug fixes and minor enhancements to improve overall platform stability and user experience.
This release brings important improvements under the hood. While mostly technical, these changes lay the foundation for a faster, more predictable, and more scalable experience.
For detailed information on these features and more, visit our documentation.
Happy Simulating!