Baixar Tabular Editor 3 3.22.0 per a Windows

New in 3.22.0
Check out our release blog to get a brief overview of the most important updates in this release.
(Enterprise Edition only) Tabular Editor 3 now fully supports Direct Lake on OneLake and Direct Lake on SQL. This means:
- The Table Import Wizard has new import options for Fabric Warehouses and Lakehouses. You no longer need to know the SQL analytics endpoint of the Warehouse/Lakehouse you wish to connect to. Instead, we show you a list of all the Warehouses/Lakehouses in your workspace, allowing you to choose the one you wish to use.
- On the last page of the Table Import Wizard, you can choose whether to use Direct Lake in OneLake or SQL mode, or whether to configure the table for plain old "import" mode.
- When the model contains tables in Direct Lake on OneLake mode, you can now also add tables in Import mode, in order to create mixed mode/hybrid models.
- When previewing data in Direct Lake tables, we no longer restrict the number of rows returned (so you can freely scroll through all the data in the table, just like in Import mode). Be aware that all columns are swapped into the Semantic Model in the Fabric capacity when you perform a data preview. Consider using a DAX Query if you are concerned about memory consumption on your Fabric capacity.
Note
We no longer configure Direct Lake models to use a case sensitive collation. If you plan to use Direct Lake on SQL with a Fabric Warehouse that uses a case sensitive collation, you must manually fill out the Collation property of your semantic model before adding any tables or other objects.
For more information, view our Direct Lake Guidance article.
Improvements in 3.22.0
- When saving a model as a single file, the Save file dialog now suggests "model.bim" as the default file name.
- Updated various dependencies to their latest versions, including AMO/TOM to 19.98.0.3. This update fixes longstanding authentication issues with Work or School accounts and repeated sign in promts.
- Changed the default Compatibility Level for new Power BI / Fabric semantic models to 1609, which allows you to control how Selection Expressions on your Calculation Groups behave by default, through the Model SelectionExpressionBehavior property.
- DAX query editor support for the MPARAMETER keyword, see #1467.
- Our DAX editors now support word based auto complete search terms. In other words, if you type sales ytd in the editor, the auto complete will now suggest measures such as [Sales Margin YTD], [Sales Revenue YTD], etc.
- With the June 2025 update of Power BI Desktop, external tools can now perform any write operation on the semantic model. As such, we no longer restrict the modeling operations available in Tabular Editor, when connecting to a model in this or newer versions of Power BI Desktop.
Bugfixes in 3.22.0
- Added exception handling when encountering IOException's during save-to-folder operations (where we sometimes need to delete folders in the folder structure, which can be blocked by file system locks placed by version control systems).
- Fixed a bug where certain keyboard shortcuts (both default and customized) didn't work when the "What's new" page was focused.
- When assigning a custom keyboard shortcut that is already assigned to another command, we now show a warning message and remove the shortcut from the other command, to avoid ambiguity.
- Fixed a bug where the Semantic Analyzer showed a false error message when using unqualified column references in a DAX window function. See #1460.
- Fixed a bug that could sometimes cause the Data Refresh view to freeze, requiring an app restart to resolve. See #1461.
- Fixed several smaller DAX editor issues related to auto complete, auto formatting, and syntax highlighting of keywords.
- Query Group names (i.e. "display folders" for Shared M expressions and M partitions) can now be modified. Also, if pasting an object with a query group into a model that doesn't contain said query group, it will be created automatically, so as to avoid the error message on subsequent Save / Deployment of the model.
- If pasting an object containing a Shared Expression reference to a model that does not contain the Shared Expression, we remove the reference to avoid "object does not exist" errors when saving/deploying the model to Analysis Services / Power BI.
- Fixed a bug where it was sometimes not possible to Undo previous changes, after saving the model metadata back to the server.