Digital Enterprise Suite

Bug fixes:

• The detection of the absolute URLs was not working in some configurations causing absolute URLs to be displayed. This prevented the container location strings from being valid.
• A term containing healthcare coding missing a system definition could cause a deployment issue.
• Fixed the screen flickering when loading the service library.
• Custom attributes that were renamed did not properly show up in the diff view.
• Method and Style validation of decision table was incorrectly reporting issues when the output of the column was a variable with the same data type as the column.
• When testing a decision logic that returned a number, in some cases the info marker would not display the execution result.
• When publishing for automation a workflow model linking to a decision model that uses a namespace from a different vendor, that model would not be properly packaged resulting in an incomplete deployment.
• The FEEL unparsed language was not always kept after a reload in some rare conditions in the workflow and case modelers.
• The ellipsis marker and menu could be out of sync when undoing the morphing of a data association.
• In the word/html report, the data mapping color could be incorrect.
• Clicking on a type inside the data mapping of a call activity would not properly open the data type dialog.
• There was a rare issue that would duplicate the content of the ellipsis menu for included decision.
• Updating the name of an element in the original model would not always update in the more referencing it for copy of links.
• State information could be duplicated when updating it from a different language.
• Undoing a decision logic clear could leave the decision logic page blank.

Bug fixes:

• Decision models logic containg boxed expression rendered blank in the Kommunicator.
• The translation dialog was not properly displaying for all users.
• Method and Style validation of adhoc subprocess does not return errors anymore when containing tasks without connections.

New and improved features:

• Kafka messages can now be generated from the workflow automation.
• It is now possible to subscribe to Kafka topics to trigger start or intermediate events while automating workflows.
• A new Kafka emitter type is available to publish events (modeling and automation) to Kafka brokers.
• Data Mapping FEEL expression issues are now reported when deploying a workflow to the automation.
• The data object documentation is now displayed when available for user tasks in the Try It form.
• The PKCE authorization flow is now supported for obtaining bearer tokens from the Digital Enterprise Suite.
• The repository content Public API now offers models in DMN 1.3.
• Triggers can now be edited from the Operation Library.
• Iterator, Conditional and Filter decision logic are now usable in workflow process data mapping.
• Data Mapping for case tasks is now available in the case modeler.
• Operation library outputs now allow to define any Content-Type for the body of a response.

Bug fixes:

• An error could occur when executing test cases in decision models that included other models.
• When not using a gateway to branch after a script task in a workflow model, an unexpected error could prevent the proper execution of the model.
• Under rare conditions, the process name was not properly updated after a manual task.
• With a container deployment, setting only the SMTP_HOST environment variable without the SMTP_PORT caused an error instead of using the default value.
• Publishing decision models linking terms with sources information would fail.
• Renaming a decision with the same name as an input data could generate an error when undoing the operation.
• Creating a data type from the output of a decision table inside a context could fail.
• Clicking on the data type of a context entry could edit the global data type of the logic instead of the specific one clicked on in rare conditions.
• Adding a new row to a decision table while editing a table cell could cause the operation to silently fail.
• HTML tags were not properly escaped in the information, warning and error markers displayed on the diagram (validation errors, export problems, …) .
• Exporting to an image with a shape selected causes the shape to appear black in the rendering.
• Exporting to an image with a shape containing term links and user entered new lines could generate a blank label.

Bug fixes:

• Certain internal services results were not properly unwrapped when they were used in Business Knowledge Model (BMK) in decisions.
• Attended tasks with multiple inputs could under certain conditions display one of the inputs as a GUID.
• A decision table column with an enumeration containing an empty string could fail when editing or entering a value.
• A decision model referencing a collection of X data types only inside an iterator, conditional or filter boxed expression would fail to properly validate.
• Assigning twice the same name in the input mapping clears the context when adding a new line.
• Missing line borders for Decision tables inside filter, conditional and iterator boxed expression.
• Certain codes were not properly expanded when retrieving them by ID from terminology servers.

Bug fixes:

• The dependency panel was not always properly refreshed when a workflow task had data associations.
• When searching for Value Sets with a space in the name, some FHIR server could return an error that prevented the proper matching.
• Value Sets returning codes without coding systems were not indexed in the graph.
• Using the underscore character to link to a term in Discovery Accelerator does not properly link the term.

Bug fixes:

• Making a REST call that returns information in an HTTP header would fail if the Header was not present.
• Copy/pasting a boxed context result row would generate an invalid model.
• Data types are not updating their missing status in the dependency panel.
• When searching for a term in the graph, the automatic matching should not occur if more than one match is found.
• When renaming a recursive data type, the recursive names are not updated properly.
• Removing an included model does not properly sync with the dependency panel.
• Decision services with missing nodes cannot be unlocked.
• Hitting CMD on MacOS while a shape is selected could clear the label.
• Changing a collapsed sub-process to/from an event sub-process does not update the icon next to the page tab.
• Internal DES services can now be called from Decision Models.
• Discovery accelerator could fail to load models whith a dependency to an accelerator.

New and improved features:

• The Knowledge Entity Modeler tab Coding was renamed to Healthcare and now offers autocompletion of codes using ValueSets of CodingSystems from a FHIR Terminology server.
• New healthcare specific FEEL functions (healthcare codes, healthcare term codes, healthcare code in term) are now available to access terms codes at automation.
• The FEEL functions ribbon was moved to the LEARN ribbon and added to the workflow and case modelers.
• A new admin interface allows to configure the FHIR Terminology server settings.
• We now support decision and workflow models that defines CDS Hooks Cards as outputs to explicitly control the content of the returned cards. The default implicit mapping still applies for existing models.
• A new wizard is accessible to define emitters. This wizard is accessible under the /management/emitters URL. From this wizard, both user and system emitters can be created.
• It is now possible to document item components of data types.
• The identity management page (/management/identities) now allows to share identities with groups. Shared identities are usable by member of the group but can be deleted only by their owner.
• The identity management page (/management/identities) now offers an improved wizard to create HTTP identities.
• The progress when searching for a term in the graph is now visible in the rich text editor.

Bug fixes:

• Assigning a constant value in a user task mapping could generate an error for numeric data types.
• A workflow model containing an operation without input or output would not be allowed.
• Posting invalid or incomplete CDS Hooks prefetch data could generate an error without further information on how to fix it.
• The term selection dialog style was aligned with the other modeling products.
• The view ribbon was not fully displayed on smaller screens.
• It was not always possible to change the column data type of decision tables embedded in the return statement of iterator boxed expression.
• Under rare conditions, it was not possible to switch back from Natural Language to FEEL for an expression.
• Under certain circumstances, it was not possible to publish a workflow model including a decision model that includes another decision model.
• Data type subcomponents were truncated after 1 level when exporting the documentation.
• OpenAPI import does not import properly files with undefined object or arrays.
• Creating recursive data types was not allowed in the user interface.
• Links could be created for basic FEEL data types in the report.
• The OData create operation output was not properly poupulated with the OData-EntityId. That value was always null.
• A call activity that call a non executable process did not provide an adequate error message.
• Including a PMML model that contains spaces in its name would fail at publishing.

Bug fixes:

• Under some circumstances, the operation path containing spaces could be incorrectly encoded.
• To improve the support of legacy OData servers, query parameters containing spaces are URL encoded using %20 instead of +.
• When expanding a knowledge entity model that contains business rules from the digital enterprise suite, it could create an error and prevent the expansion.

Bug fixes:

• Some Workflow and Decision models invoking models could fail to publish to the automation.
• In certain conditions, decision services with missing data types would generate an error when synchronizing with their reference.
• The dependency panel could fail to load if a graph was unknown.
• When an element that is reused from the digital enterprise graph is reconnected, the lock stay red.
• The name of elements in the dependency panel is not always properly refreshed from the source model in Discovery Accelerator.

New and improved features:

• The automatic model comparison is now available in Discovery Accelerator (subscription required).
• The dependency panel is now available in Discovery Accelerator.
• A new security page is available to admins to control the web session length, the cross origin and iframe usage settings.
• For services that have a single input, the REST endpoint now supports submitting the input name or omitting it in the API call. By default the Open API representation of the input will omit the single input name but this change is backward compatible and messages containing it will still behave as they did before.
• The expression context in the Case Modeler now offer FEEL autocompletion.
• Roles cannot be assigned to Case File Items.
• The severity level of messages when opening a model where there are missing references to the Digital Enteprise Graph were adjusted to warnings.
• A new shortcut option was added to compare with the last saved version of the current model.
• The reuse buttons in the ribbon bar of the modeling products now display the product names instead of the standard they implement (BPMN to Workflow Modeler).
• Now supports the header x-forwarded-port when deployed behind a proxy.
• A new function is available for workflow automation to extract the text of an HTML string.
• When deployed in audit and debug mode, invoking BKM and Decision Services will now emit ServiceNodeStarted and ServiceNodeFinished messages.

Bug fixes:

• Re-importing OData definition disconnected the invocations to the operation.
• Generating a Word report containing test cases could produce an invalid Word document.
• It was possible to create an invalid BPMN XML file when using attended tasks.
• Sometimes Data Types would show up in the model comparaison without having been modified.
• Attachments URL and obsolete alternative names are now properly compared between models.
• Some performers of tasks could be impossible to rename.
• Data Types imported from a different Digital Enteprise Graph were not always properly mapped.
• Automatic translation could fail when translating a text with large images.
• Events are interrupted from being delivered to websocket emitters after a timer execution.
• A start message without data mapping introduced a context with the message name to the target data object.
• When resuming a workflow from a timer event, the base url setting of the automation is not considered when generating absolute URLs.
• The OpenAPI definition file could incorrectly represent messages that were input for messages when they were collections.
• The same message can now properly be used for both starting and during a workflow automation.
• The scope input field size was not aligned with the description and notes.
• Exporting to BPMN, the Data Objects are not properly connected to their activities.
• It should not be possible to lock missing copy of relationships.
• Reuse links can now be changed to copy links (unlocking) even when the link is broken.

Bug fixes:

• The typing context for iterator and filter boxed context could be lost preventing the proper parsing of ambiguous FEEL expressions.
• When making a REST call, certain characters in Query parameters are not properly encoded.
• Importing an OData metadata file containing entities without primary keys would fail.

Bug fixes:

• Attachments on elements can’t be deleted.
• Sometimes the HTML export fails when there are items with copy or reuse links in the model.
• With some configurations, the accepted Access Requests invites do not disappear.
• The FEEL now() function could cause an incompatibility when used in a decision table to compare with another date.
• The signal events in email sending incorrectly required an identity to be triggered.

Bug fixes:

• When exporting a Discovery Accelerator model to the Workflow Modeler (BPMN), the data objects links will now all be connected.
• Events changes are now properly reported by the model comparison.
• The term addition is not always rendered correctly in Firefox.
• Comments changes are not displayed when comparing models.
• Custom attributes changes on models are not displayed when comparing models.
• Data isCollection attribute changes are not displayed when comparing models.
• Removing a code sometimes doesn’t show properly when comparing models.
• Certain newly added elements would not show up when comparing models.
• Conditional expression on intermediate events cannot be modified.
• Only updated elements are now refreshed in the dependency panel.
• Improve how we present the model compare when a decision logic type was changed.
• Hit Policy and aggregation are not properly saved in Process Instance Name and Process Instance Tags decision tables.

Bug fixes:

• The DMN validation would report errors on validation for models that did not contain errors.

New and improved features:

• It is now possible to compare the current model to a previous version of that model. This option is available in the Teamwork ribbon (separate subscription required).
• When publishing a service (workflow or decision), 3 profiles are now available:
  • Production: Will generate a minimal amount of service events at automation for performance.
  • Audit: Will generate additional service events for each model node traversed at automation.
  • Debug: Will generate additional debug information for traversed nodes used for debugging purpose only.
• A new Dependencies side panel lists all reuse of the current model to external models.
• Terms can now be dragged from the Digital Enterprise Graph as entities in the modelers.
• Test cases, process instance names/tags and data mapping are now available in the documentation (html/word).

Bug fixes:

• Event emitter update API function could fail for certain emitter types.
• A workflow service with a dot in its process name would generate files with the wrong extension when trying to download the automation results.
• Reading XML context data now ignore non-alphanumerical character from attribute names when matching to the data types.
• The OpenAPI definition of a service containing a collection of collections could generate invalid files.
• Defining the condition of a complex gateway could fail.
• Parameters and outcomes are not properly synchronized when reused on decision services.
• Sometimes autocompletion is not offered inside boxed context entry in a deep context.
• OData import does not properly support multiple namespaces in the metadata definition.
• Fixed an issue where creating a boxed context inside another boxed context could create a rendering issue.
• Fixed an issue where creating a conditional expression inside a boxed context could create a rendering issue.
• Fixed an issue where a FEEL unparsed marker could not be cleared properly.
• Translation on hidden items could prevent the model from being opened.
• The second and third parameter of a BKM invocation could experience some autocompletion issues in specific situations.
• Workflow modeler could freeze when drawing a data association between two lanes (not in a pool).
• Invocation of reused decision service with missing type gives a blank page.
• Input/Output on case are not exported.
• FEEL autocompletion not suggesting sub-components after a filter in a for loop.
• Invalid test cases could prevent a decision model to be exported to DMN XML.
• Data input/data output cannot be synchronized with their related graph element.
• Resume point form description could in some rare cases be incorrectly rendered.
• The product FEEL function edge case product([]) = null was corrected.

Deprecation:

• In version 10.10.0 (June), authenticating using the AuthToken session cookie to the APIs will not be supported anymore. Please use bearer tokens generated from Client Apps.

Bug fixes:

• The rich text editor of the detail dialog was not properly displayed preventing the edition of rich text.
• The excel export dialog is not always visible.
• The REST based emitters defined by users could run into a race condition preventing them from being correctly re-instantiated from configuration.
• The REST API user interface would appear frozen when an error happened when expanding an operation.