We’re happy to share the April 2023 release of Tabular Editor 3, which contains a much requested new feature – the ability to connect to and import data directly from Databricks – see more below.
The headline this month is a native Databricks connector, covered first below.
You can now connect natively to Databricks and import/refresh table metadata, when working on a Power BI Dataset from inside TE3.
You will need to install the Simba Spark ODBC driver on your development machine.
To import/refresh table metadata, you must open or create a Power BI Dataset and then use the Import tables or Update table schema wizards with an Implicit data source.
To prepare, you’ll need the connection details of the SQL endpoint from your Databricks instance:
With the Server hostname and HTTP path in hand, it’s time to start up Tabular Editor 3 and connect to your Databricks instance (in this example Azure Databricks, but it supports all flavors):
We hope you like this new native connector, please let us know if you find it useful or find areas that needs improvement.
The Databricks connector is only available for Implicit data sources, which is why you’ll only see it on Power BI Datasets. If you need to connect to Databricks in an Analysis Services model or using Provider or Structured data sources (see the differences here), you will need to use ODBC as usual.
In the April 2023 update of Power BI Desktop the new functions RANK and ROWNUMBER were announced. Moreover, the ORDERBY function has a new syntax allowing the use of any scalar value for the expression parameters, and a new set of keywords for the order-by parameter.
Release 3.6.0 includes full support for the new functions and syntax in the DAX editor code completion.
You can see more info by visiting Jeffrey Wang‘s blog about the new functions.
There are several interesting improvements in this update, starting with first-class support for dynamic format strings.
When adding a dynamic format string expression to a measure, we now clear the static format string property (as it is not possible for measures to have both static and dynamic format strings assigned).
Default Compatibility Level for new Power BI models created with Tabular Editor is now 1601.
This compatibility level allows you to set the FormatStringDefinition on measures, allowing for dynamic format strings mentioned in last section.
Moreover, the “create new model” dialog now lets you specify any compatibility level.
Please see our release notes for full details.
We have made some bug fixes in 3.6.0; please see our release notes for full details.
If you have any feedback or experience any bugs please reach out in our community forum. We look forward to hearing from you
TMDL is a new format for serialization of a data model, that resembles YAML. You can see more about this new format in the official Microsoft documentation.
Since the big announcement of TMDL at SQLBits earlier this year by Mathias Thierbach & Gabi Münster (watch it here), there has understandably been a lot of hype about this great addition.
We added support for TMDL in Tabular Editor 2 (TE2) this week (get that release here) and perhaps people was expecting support to show up in TE3 in this release.
However, we have decided to wait until the release of TMDL preview-2, which should be out in the near future due to some high impact known issues.
This is not because of a lack of commitment to TMDL from our side (quite the opposite), but from a wish to provide the most stable environment possible to our customers.
As of 2026, Tabular Editor 3 now supports TMDL: you can save and load models in TMDL serialization mode and export any object as a TMDL script from the TOM Explorer (see the docs).
That's the April 2023 release: the headline is the native Databricks connector, with support for the new DAX functions and some format-string polish alongside it. TMDL held back for preview-2 at the time, though Tabular Editor 3 now supports it. Grab 3.6.0, and let us know how the Databricks connector works out for you.