Digital Enterprise Suite

• PDF images could not be generated for certain models.

• An error could occur when executing test cases with a model containing multiple subsequent spaces in its name.
• Overloading the Server URL of an operation (for SMART on FHIR for instance) could cause an execution error under certain circumstances.
• Improved the error message in some cases where a paranthesis was missing.
• Cases could be incorrectly initalized and an error could prevent some proper data mapping to sub instances.
• Flags could be depicted multiple times on terms.

New and improved features:

• Analytic Data is now also available for Decisions.
• Cloud Execution now offers a Connectors tab to visualize the status of external connections status and statistics about their usage.
• The Trisotech Help was greatly improved for the Workflow Modeler, Decision Modeler and Case Modeler.
• The AI Ninja (assistant) is now available in most canvas to help you break the white page syndrome. This is part of our AI Feature Set.
• The AI Ninja (assistant) is now available in the text view of Discovery Accelerator to automatically recognize the named entities from your text and extract them. This is part of our AI Feature Set.
• Added a new service info() FEEL function to the automation custom FEEL function library that provides information on the current running service and its version.
• When invoking a service using the automation REST endpoint, an additional HTTP Header (X-Version) is now returned to identify the version of the service that was executed.
• It is now possible to test the configuration of execution environment external registry to push Digital Distributed Containers.
• It is now possible to change the routing of links between a Straight and Manhattan.
• Renaming case file item will now propagate that change to expressions using that case file item.
• Renaming a context entry, function parameter, iterator name and decision table input header now propagate the change to the decision logic using that name.
• Multiple drawing improvements for the Shared Data Modeler including: adding simple data type to the quick connect menu, the reference connector is now the default connector, some links are now straight by default and more.
• When displaying errors coming from included models, we now display the included model name.
• Support Amazon Web Service (AWS) proprietary signing authorization v4 (AWS4-HMAC-SHA256).

Bug fixes:

• Optional arguments for custom FEEL functions would not be considered optional at invocation.
• Improved the message reported when a FEEL statement contained an if without and else.
• Warnings in FEEL expression in decision tables were incorrectly maked as errors.
• The documentation of the Public API endpoint to download emitter audit file incorrectly listed the parameter as requiring date and time format when only the date was considered. The documentation now properly reflects that.
• Sending email with an attachment stored in an external file did not work properly.
• When clearing the content of some boxed functions body, it incorrectly displayed the option to use a function as the body.
• Under certain circumstances, the save because of changes dialog was shown when viewing a model.
• Improved the quick connect arrows to prevent from going around a shape.
• Ellipsis and collection markers overlapped on data items shape.
• Some changes to a decision service are not properly reported in the model compare report.
• Result of testing were not always displayed on conditional boxed expression.
• Changing the type of a shape with multiple depictions could generate duplicated ids.
• Fixed DMN 1.5 namespace as the reference for item definitions.
• Services instance links were not pointing to the proper URL when a service was deployed to a non-default group.
• Prevented improper scrolling using the area selection tool on a scrolled canvas.
• The option to paste by value or reference is not displayed in the contextual menu for case file items.

Non-Backward Compatible Changes:

• Following a deprecation announcement by Microsoft of their webhook integration into Teams we updated our integration with Teams. Existing published services will continue to work using webhooks until Microsoft remove the support on their side. To make new deployments, you will need to reconnect your team account in the admin interface and reconfigure your user task Teams integration.
• The default email sending address for emails from the Digital Enterprise Suite is now noreply@trisotech.com instead of support@trisotech.com.

• Security fixes.

• Security fix Long Term Support release.

Bug fixes:

• Case file items inside stages would sometime cause issues with the data mapping.
• Deploying a case using the debug execution profile prevents decisions from properly executing.

• Security fix Long Term Support release.

Bug fixes:

• When adding a user to a group, a restart is required before that user can access the identity management page.
• The Azure Container Registry token authentication method was not supported to host Digital Distributed Container images.
• It was not possible to delete partially connected composition connector.

Bug fixes:

• A small percentage of empty result of boxed context were now reporting validation errors incorrectly.
• Enabling or disabling a client access licenses with an expiration date would result the wrong column actions being updated.

Bug fixes:

• B-FEEL functions sum, mean, median, product and stddev were not properly accepting varargs arguments.
• DMN 1.6 models using healthcare terms could not be published and tested in some circumstances.
• An empty result would incorrectly depict a No decision logic defined message.

New and improved features:

• Support for DMN 1.6:
  • Import and export in the DMN 1.6 XML.
  • B-FEEL dialect support.
  • Descendant (…) operator.
  • Date and time data types now have a value attribute.
  • New date and time function with IANA time zone.
  • Rounding functions with a single parameter.
• Introduced Analytic Data:
  • Analytic data are a subset of the data elements that are part of a process (data object/input/output) or a case (case file item) that has analytic value.
  • The modeler allows to define FEEL expressions to generate analytic data from a service instance.
  • Cloud Execution allows to download the analytic data in a CSV file using filters.
  • The exported CSV can be analysed in your favorite Business Intelligence tool (Excel, PowerBI, Tableau, …).
  • Analytic data for completed services is dependent on the History enabled setting for the deployment environment.
  • A new service event is published when new analytic data is generated for external integration in audit and debug execution profiles.
• The FEEL expression in workflows and cases also support the new features of DMN 1.6 (B-FEEL, descendant operator, …).
• An information panel present the instance name, tags and analysis data when testing workflow processes.
• Boxed context result can now use any type of boxed expression.
• When a missing model link is reported when opening a model, clicking on the error will now open the proper context to relink or remove the link.
• The API to list service instances now supports an additional attribute to specify an AND or OR when supplying multiple tags.
• New automation messages are now available for events sent and received (ServicePublishedEvent, ServiceReceivedEvent) in the debug profile.
• Template configurations now use the default configuration for any link that is not overloaded by a configuration.
• It is now possible to change reuse by copy links in template configurations.
• The process instance viewer now allows to launch the form interface on a call activity.
• The number data type display supports outputting as a currency and the display mask now support ECMA-376 (spreadsheets) patterns.
• Now displays <empty> or <null> in the tooltip of the marker on the diagram for empty and null values when testing models.

Bug fixes:

• A process that contains a pool and an annotation could not be deployed.
• The BOM character was not supported in an XML file used as a service input.
• The case instance API was not returning the process sub-instances.
• Timer events would prevent the proper deployment of case services.
• Timer events that fired in the past will now fire on instantiation.
• It was possible to receive a non-existing error when using locally and included data type.
• Data types were always shown in the terms view.
• The template configuration overlay is not visible when generating hte report.
• The alternative depiction of data types is now available on the diagram instead of the data input.
• Sometimes criterions were duplicated in CMMN XML.

Non-Backward Compatible Changes:

• DMN 1.6 defines a JSON binding. The JSON binding is mostly aligned with our current API binding but there is one exception for IANA time zones. Previously, a date and time with an IANA timezone would be represented as: 2007-12-03T10:15:30@Europe/Paris. The new binding defines this format instead: 2007-12-03T10:15:30+01:00[Europe/Paris]. Therefore, starting on this version, our REST API and message payloads have been updated to align to the mandated format.

• Security fix Long Term Support release.

Bug fixes:

• When testing a decision model, the result was not always displayed as an information icon for invocation, iterator, conditional and filter.
• Digital Distributed Container security prevents secured event topics from logging to the emitters.
• Using OAuth 2 additional parameters for authorization (for Google OAuth for instance) was not working as intended.

New and improved features:

• Template configuration can now be created for executable workflow process. It allows to define multiple reuse configuration for a process.
• The dependency panel of the Digital Enterprise Graph now offers the ability to relink dependencies to a different element, model or graph easily.
• It is now possible to keep the history of service execution for process enabling to visualize their end state in Cloud Execution.
• Error end events in workflow processes will now map to error responses allowing to generate customized HTTP responses with custom error documentation.
• A new attribute on the modeling canvas is now available to automatically assign an identifier to newly created shapes.
• The outputs of a service returning a single output will no longer have that single result wrapped in an object with a key representing the output name when requesting the Accept type: application/json+result.
• Workflow process can now leverage the CDS Hooks suggestion feedback mechanism.
• A new utility custom FEEL function was added to recursively remove nulls from Any data.
• Improved the time required to import large XSD data type definitions.
• The horizontal scrollbar for the test panel is now always available and does not require to scroll to the end of the form.
• Multiple rows of a decision logic can now be selected.
• Submitting an invalid JSON file to the test panel would result in a generic error and now result in a more significant error.
• When downloading the technical result of a test, the order of the keys will now be constant.
• The rich text editor controls are now available when in text view.
• Improved the formatting of the information bubble tooltips when testing a decision model. They also now include a scrollbar and their text can be selected.

Bug fixes:

• Testing a decision logic that contains a slash character in its name fails.
• Decision table analysis would in rare cases omit gaps.
• When using a collection was referring a type with a custom rendering, the rendering properties could be ignored.
• Copying and pasting a line in the data mapping could create an additional empty row.
• The dependency panel is not properly updated when deleting unused data types.
• Dropping a process from the Digital Enterprise Graph can introduce duplicated data types.
• Models can contain duplicated data types when the data type name contained special characters.
• Importing two XSD files in the same model could generate duplicated ids.
• Multiple iterator in a FEEL for expression could fail with an undefined variable if some of the iterator resolved to an empty list.
• When testing a decision logic with boxed context, some entries would not display their result using the info icon.
• When creating a relation table with types of complex types, the columns referring to complex names could be incorrectly parsed.
• Decision models called in process launched from a case could be incorrectly connected.
• Testing a process with a file input would not always properly display the submit button on a later user task.

Bug fixes:

• The DMN 1.2 XML export could fail with include models.
• A process model calling a decision model with a name containing a trailing space would not load properly.
• Missing information on referred data types were shown as error without a clear indication of the root cause.

New and improved features:

• Added a preference to not synchronize all dependent models when testing and publishing to automation.
• Added a preference to only get the final result when testing Decision models for faster exection.
• Model reusing a large amount of data types will now be synchronized faster.
• The Digital Enterprise Graph navigation performance has been significantly improved for large graphs.
• When creating a test case, the current service being tested is also saved in the test case allowing for partial model test cases to be created and later tested.
• It is now possible to select and deselect all elements of a category when generating reports.
• The admin configuration of the email notification channel now allows to configure default email addresses.
• We can now duplicate rows of decision logic.
• We can now duplicate rows of expression for instance name, tags and data mapping.
• The utility custom FEEL library now supports decoding both standard and URL Base64 character set.
• It is now possible to view the HTML code of rich test editor inputs.
• The info dialog now allows to regenerate the model namespace.
• An option to include or exclude examples from the report generation is now available.
• When stopping a test started from a file, the input form now stays on the file input mode.

Bug fixes:

• The process instance viewer was not accessible in digital distributed containers.
• Some models were dirty after opening them from a place.
• The input data types could be considered of type Any instead of their actual types when testing a Business Knowledge Model.
• Constraints defined on a collection reused as a base type could be incorrectly evaluated.
• Task data mapping did not support iterators boxed expression.
• The emitter Public API endpoints were not accessible with a bearer token.
• The Twilio connector parameters were updated to align with changes in their API.
• Script task did not evaluate output data mapping when there was no expression defined.
• Some model containing decision tasks could not be deployed.
• It was not possible to select the Microsoft Teams distribution channel for user tasks.
• Data could not be edited in call activities when testing models.
• The collection marker was not always shown on data items.
• Item definition names are not translatable and will not be displayed in red anymore when viewing a translated language.
• Reused data types could not be deleted.
• In some cases, the process instance viewer was not displaying complex data types using the proper input rendering.
• Copying a sub process with children could cause model structural issues.
• Zoom to fit for case models did not consider the case name.
• Inner data inputs not displayed on the form could not be used in the email/teams notification template for user and manual tasks.
• The tooltips from the report generation dialog do not always properly disappear when closing the dialog.
• The input description is not shown when testing a model.
• The user interface to edit data stores was not updating properly after a save.
• The FEEL highlighting will now properly recognize multiple iterator variables in a for statement.
• The FEEL highlighting could now work properly after an addition or subtraction inside a filter statement.
• The FEEL highlighting will not try to convert numbers as strings when defining a context as a literal expression.
• Fixed an issue with the automation when multiple sequence flows were linking the same two shapes with different constraints.
• When testing a workflow model, it was possible that the data panel would not be properly updated with the new value of a variable that used data mapping.
• The shared data modeler now supports importing DMN 1.2, 1.3 and 1.4 models.