| New Features |
|---|
| 1. | The StiOptions.Web.AllowGCCollect option is now supported by the report viewer and the report designer in ASP.NET MVC and ASP.NET Core. It starts the garbage collector after the server responses, which reduces the memory used by the application at the cost of slower responses under a large number of simultaneous requests. |
| 2. | The RowNumber() function for dashboards. It returns the row number (1..N) of a row in the final sorted result. |
| 3. | The MS SQL data adapter based on the Microsoft.Data.SqlClient provider, which supports strict SQL Server encryption (Encrypt=Strict). It works only on .NET Framework. To use it, create a standard "MS SQL" connection, then select "Microsoft.Data.SqlClient" for the "MS SQL Data Adapter" option on the "Data" page of the report designer options, or set the StiOptions.Dictionary.UseMicrosoftDataSqlClient property to true. The report designer downloads the provider from NuGet. |
| 4. | Sparkline in Cross-Tab and Pivot Table. |
| 5. | Added the ability to set the color of chart points from a data column in dashboards. It works for the "Pie", "Pie 3D", "Doughnut", "Funnel", "Treemap", "Sunburst" and "Pareto" chart types, as well as for clustered column and bar charts. |
| 6. | MonthYear, QuarterYear, WeekYear and DayMonthYear functions in reports and dashboards. They group data by the month, the quarter, the week or the day within the year, keeping the chronological order. |
| 7. | The PrimaryKey property has been added to the data columns in the data dictionary. The key columns are taken from the data source schema or from the Key attribute of business objects, shown with a key icon in the dictionary tree, and preselected when creating a relation in the report designer. |
| 8. | The relations of the data dictionary are now created from the foreign keys of the database when retrieving columns from a SQL connection of any type: MS SQL, MySQL and MariaDB, PostgreSQL, Oracle, Firebird, SQLite, SQL CE, Access, Teradata, Snowflake, DuckDB, VistaDB, as well as the connections through ODBC and OLE DB. |
| Enhancements |
|---|
| 1. | Support for VistaDB 6 databases: the VistaDB connection now works through the VistaDB 6 provider, databases of VistaDB 5 are still opened, and the primary keys of the tables are retrieved. |
| 2. | When the StiOptions.Web.AllowGCCollect option is enabled, the garbage collector now starts after the response is sent, so the web viewer and the web designer no longer delay the responses while it runs. |
| 3. | The error of a broken expression of a dashboard element is now displayed on the element itself in the report designer and in the rendering messages of a report. Before, such an error was skipped silently and the element just showed no data. |
| 4. | Retrieving the database schema of an Informix connection in the report designer, including tables, views, columns, stored procedures with parameters and relations. |
| 5. | Retrieving the database schema through an ODBC driver has been reworked: the primary keys and the relations are retrieved, and the system objects are no longer listed among the tables. |
| Fixed Bugs |
|---|
| 1. | No rows in the "View Data" window and in the report for the tables of a Teradata or VistaDB connection created from the database schema in the report designer. |
| 2. | Synchronizing the columns and the primary keys of the data sources of an OData V1-V3 connection with the "Synchronize" command of the data dictionary. These data sources were skipped: their columns were not updated, and a primary key unchecked in a column did not come back. |
| 3. | Retrieving the types of the columns which an ODBC driver names in its own way when reading the database schema: int4, float8 and timestamptz of PostgreSQL, DOUBLE PRECISION of Firebird, datetime2 and the identity columns of MS SQL. Such columns came as strings, the tables with a space in the name came without columns, and the relations of these tables and key columns were lost in the data dictionary. |
| 4. | The data sources added through an ODBC driver failed to build in a report when the name of a table, a view or a stored procedure contained a space or mixed-case letters, or when the table belonged to a schema: the names in the query were quoted the wrong way for the database, and the schema was added twice. |
| 5. | The server memory used by the web designer and the web viewer was not released after closing or reloading the page, and was kept while the application was idle. |
| 6. | The undo history in the web designer could be lost when the session of another user expired. |
| 7. | Displaying the views of an Oracle database in the "Select Data" form when connecting through the Devart provider. |
| 8. | Wrapping a one-character word that did not fit at the end of a line in a text component using the SixLabors library: the component did not grow to accommodate the extra line, and the last character was clipped. |
| 9. | The "Interaction" editor of the dashboard elements reports a broken expression right under the field which holds it - the tooltip, the hyperlink or the value of the parameter. |
| 10. | The data type of the FLOAT and SMALLFLOAT columns of an Informix data source, FLOAT is now double precision and SMALLFLOAT is single precision. |
| 11. | Missing string columns in the schema of an OData V1-V3 connection in the report designer when the properties are marked as nullable in the service metadata. |
| 12. | Connecting to an IBM DB2 database with the installed IBM Data Server Driver in the report designer; the "DB2" connection is available again in the list of data sources. |
| 13. | Positioning of underline and strikeout lines for text drawn with synthesized fonts when exporting a report to PDF. |
| 14. | Losing the version of an OData connection when saving a report to an *.mrt file with JSON markup or previewing it in the report designer. A connection configured for OData V1-V3 was treated as V4, and the data source remained empty. |
| 15. | Saving the connection parameters of a Google Sheets data source when saving a report to an *.mrt file with JSON markup. |
| 16. | Saving the code page of a dBase data source when saving a report to an *.mrt file with JSON markup. |
| 17. | Retrieving the database schema through the Oracle ODBC driver no longer crashes the report designer. The columns of the whole database were requested from the driver at once, and it brought the process down; now they are read schema by schema. |
