I have been forcing my webservice data source into a workable solution for making GraphQL requests. Right now in order for us to use GraphQL, we have to build out the request manually with a string such as the following:

{"query: "query GetSomeResult($input: Int!) {result {id name}}",
"variables": "{"input": @parameter}"
}

While this is a normal way to hand craft a GraphQL request, the issue I have is that it shouldn't need to be that hard. Also I get no intellisense or suggestion in regards to the data returning. This means that when I try to bind data to a text box, I have to visually validate that the [=Fields....] is actually correct by looking at the GraphQL query and at the code base. I would love to see the ability to to just say have a set of standard api technologies we could implement from such as GraphQL queries, GRPc queries, or any other contract style request. I would expect that the data source logic would be smart enough to parse the query (at least for GQL) and be able to provide reccomendations.

Also a major issue we have had is in being able to pass that data source to subReports, or dealing with any nested objects such as a user.address.addressLine1 would not be able to be found if address is a nested values like the following

user: {
  address: {
    addressLine1
  }
}

The Telerik Reporting REST Service storage is automatically cleaned based on some configurable timeouts and hardcoded times. It requires an awaken service to be performed.

It would be very useful for the developer to be able to invoke manually the storage cleaning functionality either with code or through the service Web API.

All of my report parameters have their AutoRefresh setting set to False, and the parameters do not depend on each other, they are not cascading parameters. The parameters have available values.

If I change the value of one of the parameters, despite the above settings, all of the other parameters' data is re-retrieved(even if those parameters are not visible).

This causes issues because the "Preview" button is disabled until the parameters are refreshed, and makes the UX bad since if the connection is slow, the end-users will have to wait for the parameters area to refresh after changing any parameter value.  

I'm using VS 2022, the demo reporting (Telerik_Reporting_2026_Q1_20_0_26_211). When using the "Telerik Report 2026 Q1 Wizard", I select the following

• "Table Wizard"
• Current Data Sources
• Add New Data Sources
• Select "Object Data Sources
• I select my Assembly and data object.
• I select the Data source Member

See attach file when I finish the wizard.

Currently, when using External Stylesheets in the reports, the engine tries to first load the stylesheet as an embedded resource of the assembly. If it fails to load it from the assembly, a message is written to the trace along the lines of:

ESS: Attempt to load ExternalStyleSheet from EmbeddedResource from assembly Telerik.Reporting, Version=19.0.25.211 with resource name styles.xml is unsuccessful.

After this, the engine will try other ways to resolve the stylesheet such as from URI and if it succeeds, the report will render as expected.

This can create confusion because looking at the trace, one might think that there were errors with the rendering of the report because the stylesheet did not load but it was actually successfully loaded with one of the other approaches.

It would be better to write the errors to the trace only if the stylesheet does not load after all attempts.

Is it possible to get descriptive information exposed for the ReportsControllerBase swagger endpoints? do they exist?

Notice how the About and files endpoints have information next to them:

Like-wise how the files endpoint has descriptive parameters:

I run an ASP.NET Core application that has the functionality to export reports into PDFs in an Azure environment.

There is no way to install additional fonts on Azure so I provide the fonts used in my reports via the "privateFonts" element of the Reporting configuration.

This works for the most part, the font is embedded in the produced PDFs but if I have a textbox with a large text, the text may be cut off when using the Skia engine.

I tested running my application in Docker, I copied the needed fonts in the usr/share/fonts directory of the Docker Linux container, and tried exporting again.

When exporting from Docker, with the fonts installed in the container, the text is not cut off. It seems that there is an issue when the fonts are provided as private fonts.

Hi Telerik Team,

We are experiencing an issue related to tooltips in PDF exports from Telerik Reporting.
When we enable tooltips using the ViewerRenderToolTips setting both at the report level and through the device info in the report service, tooltips are appearing in the exported PDF—but they are being rendered as text annotations, not as normal tooltips.

Issue:
In the resulting PDF, tooltips are rendered as Text Annotations, which:

1.Appear as popups or sticky notes on mouse hover/click
2.Are not the standard tooltip experience expected by users
3.May affect accessibility compliance and user readability

Expected Behavior:
Tooltips should appear as standard PDF tooltips (like browser-style hover text) instead of annotations.

ADA Accessibility is a priority for our customers, so we are aiming to ensure logical tabbing order for keyboard users.

I've noticed that the RowHeadersPrintOnEveryPage attribute, while very useful for creating more readable PDF exports, seems to result in some strange screen reader behavior when set to True.

The first problem is with the strange tabbing order in the live environment. The second column header is the first to receive focus, followed by the remaining column headers, then the data fields under the first column receive focus in a vertical list, until finally the first column header receives focus. Without the default RowHeadersPrintOnEveryPage setting (False), the first column header is the first to receive focus, which is expected. 

In addition to the wrong tabbing order, the RowHeaderPrintOnEveryPage causes the screen reader to announce the incorrect column associated with a particular cell. I use the Windows Narrator. Other screen readers may result in slightly different behavior. 

When the current thread culture is set to Turkish (tr-TR), the lowercase and uppercase conversion rules differ from English — especially for the letter "i".

• In Turkish:

  • "i".ToUpper() → İ (with a dot)

  • "I".ToLower() → ı (dotless i)

Our assumption that the Telerik HtmlTextBox internally parses and maps HTML tags by converting them to uppercase. When the Turkish culture is active, this conversion causes tag names to break, which is not a valid tag, so ordered lists fail to render.

Best regards,
Denis

Currently, if the SkipBlankPages property is set to True and there is not any significant content in the report, the following message will be displayed in the designer and the viewer: The current data set presented in the report did not produce any significant content, so no pages were generated. If you need to see the whole report content, including blank pages, please contact the report author.

If you add transparency to a Graph line, for example, the Y-axis, with an alpha value: '150, 255, 255, 192', it is respected in the Standalone Report Designer, where the rendering format is Image.

In the Web Report Designer and in the HTML5 Report Viewer preview, the alpha value is ignored, and the color is rendered as '255, 255, 192', which removes the transparency.

I want to set the Table ColumnWidth to a particular value in the Properties of the Web Report Designer. There is no property ColumnWidth as in the Desktop designers, and I tried to set the Size.Width, which is available.

When I increased the Width from the Properties, in design time, there is no change in the column width, only dashed borders indicating the new size when selecting the corresponding TextBoxes. In runtime, though, the new size is respected.

The inconsistency repeats even when I open the saved report in another browser or through another application, hence this is not a caching issue.

The issue is caused by an issue in the SQL DataSource Wizard that doesn't provide the correct Data Provider name. In the wizard on the first page, the Data Provider must be MySql.Data.MySqlClient, but the option is limited only to MySql.Data. Since the engine cannot determine the type of the data provider by that name, it falls back to System.Data.SqlClient and claims that "Port" is not a supported keyword.

After upgrading Telerik.ReportViewer.BlazorNative to version 20.2.26.1007, rebuilding a project targeting .NET 9 or later fails with the following error:

Severity Code Description Project File Line Suppression State
Error (active) The asset 'C:\...\BlazorApp1\Debug\net10.0\compressed\14jp1i84hx-{0}-2x8nbomfaf-2x8nbomfaf.gz' can not be found at any of the searched locations 'C:\Users\<user>\.nuget\packages\telerik.reportviewer.blazornative\20.2.26.1007\staticwebassets\css\reporting-blazor-viewer.css.map' and 'C:\Users\<user>\.nuget\packages\telerik.reportviewer.blazornative\20.2.26.1007\staticwebassets\css\reporting-blazor-viewer.css.map'. BlazorApp1 C:\Program Files\dotnet\sdk\10.0.301\Sdks\Microsoft.NET.Sdk.StaticWebAssets\targets\Microsoft.NET.Sdk.StaticWebAssets.Compression.targets 276

-----------------------------------------------------------------------------------

The issue occurs because the NuGet package declares a CSS source map as a static web asset but does not include the corresponding file.

For the available workarounds, refer to the following Knowledge Base article: Missing CSS source map causes build failure in .NET 9 and later with Telerik.ReportViewer.BlazorNative 20.2.26.1007

The data source preview in the web report designer does not show the full shape of the query. It has no horizontal scroll bar to verify the preview data for the entire projection.

Please consider addressing this.

When a Table has a Group PageBreak that splits the Table into more than one page, the items below the table are rendered after the first part of the Table (after the first page break), rather than under the whole table.

The issue occurs in hard pagination: PrintPreview of the viewers, PDF, etc.