Digital Enterprise Suite

Bug fixes:

• Clicking on the plus icon to open a sub-process or a collapsed stage would not work under certain circumstances.
• Fixed syntax highlighting in FEEL expressions after a function with parameters is used inside a returned context.
• Paste by value/by reference doesn’t paste where the contextual menu was opened.

New and improved features:

• APIs Guru (an open marketplace for APIs) is now integrated in the Operation Library to facilitate the access to thousands of integrations.
• Interfaces in the Operation Library can now have their icons customized.
• When testing a workflow process, it is now possible to move/create/delete tokens.
• Process and cases reused through the Digital Enterprise Graph now offer their end/milestones variables for output mapping.
• It is now possible to update the data type of a decision logic in addition to creating a new one.
• When dropping or synghronizing a data type with the Digital Enterprise Graph, it is now possible to select a Replace All option.
• An option to automatically refresh the page on changes is now available in the service instances and my tasks views of Cloud Execution.
• Executing java classes in Business Knowledge Models (DMN) or Script Tasks (BPMN) now allows the usage of the java.lang.Math library in all deployment scenarios.
• A new Public API allows to fetch deployed models inside of services in the Digital Automation Suite.
• A new Process Service API allows to reassign user tasks to other participants.
• Allow users to decide if they want to update existing data type when importing from JSON Schema.

Bug fixes:

• When right-clicking on a shape immediatly after dropping it on the canvas, the browser contextual menu could be shown.
• Digital Enterprise Graph synchronization issues are not displayed anymore in the Kommunicator.
• When adding input requirements to a decision model, existing test cases would report missing inputs.
• Deploying a CMMN model through the Public API deployment resource could result in an error.
• When a decision was defined as outputting a Text but something else was actually returned, the test case could not be saved.
• Some workflow process services that invoke case tasks would not properly deploy.
• Various improvements to the FEEL parser inside filters.
• In an activity data mapping, a context typed as Any can’t always be filled properly.
• Some tasks in the My Tasks view could have been generated with an incorrect link. This will require an administrator re-index.
• Digital Distributed Containers were not properly establishing user sessions in the system of engagement which caused an error to be logged in the container.
• Adding an issue to a model in the example places should not be allowed.
• Can’t assign performers to signal/message events and sub-process.
• The issues overlay does not always refresh properly.
• Removing an interface from the operation logic of an imported BKM triggers an error.
• The breakpoint tab in the testing functionality of the workflow modeler would not display all breakpoints when a model was calling another one.
• The contextual menu would not be shown when clicking directly on a gridline.
• Some data types were wrongly identified as being used when selecting the delete all option.

Non-Backward Compatible Changes:

• We have deprecated and removed the XPDL import and export, the BPEL export and the Project import from the Workflow Modeler.

Bug fixes:

• The notification bell on the main page of the Digital Modeling Suite could not be clicked.
• A workflow service containing a data output named end and a case service containing a case file item named milestone would not be able to generate a valid OpenAPI definition of their service.
• Certain FEEL expressions were impacted by the right operand and and or expressions being executed without being required.

New and improved features:

• It is now possible to migrate long running service instance (process and cases) between different versions of a service.
• A new admin API is available on long running services (process and cases) to move, create or delete tokens.
• The place filtering was replaced by a place search that allows to find files in a local place or through all places.
• It is now possible to promote a place into another place. Promoting a place replace the content of the target place with the content of the source place without affecting the history of the target place and optionally assigning a version and state.
• Constraints can now be added to collection data types.
• It is now possible to add more than one flag on an element.
• When renaming a data type, its name will also be updated in the FEEL logic using the instance of function.
• A loading feedback was added for large model requiring some time to load.
• Included decision model elements now have a link to their included model in the detail panel.
• New FEEL warning messages are now reported when trying to do boolean operators (equal, not equal, greater than, and, or, …) using two expression of incompatible types.

Bug fixes:

• The instance of FEEL operator would not properly differentiate years and months duration from days and time duration.
• When reaching a dead end in a workflow, the process test functionality could incorrectly keep tokens that did not exist anymore on the canvas.
• It was not possible to hide some links in the Knowledge Entity Modeler and the Workflow Modeler.
• The browser default contextual menu was not displayed anymore when right-clicking on a rich-text input.
• Administrator using the test functionality could sometimes get incorrect results because they would also receive other users results in parallel.
• Importing an OpenAPI that does not define its output and/or error make them unusable for data mapping.
• Markdown tables not imported in the OpenAPI import.
• There are missing functions in the context of the FEEL expressions in the data type in the Knowledge Entity Modeler.
• Included decisions that are not in the current context were incorrectly added to the whole model decision service.
• When an included model could not be found, the model might not be able to load its pages.
• Custom FEEL functions were not updated in item definition constraints when the prefix was changed.
• Decisions from included models did not depict ellipsis on missing requirements.
• When clearing the search field in the data type dialog, some letters may stay highlighted.
• The save button of the products section of the user provider tab in the administrator instance would not properly save for certain clients.
• The dependency panel and where used panel would not always load.
• Improvements to the FEEL context parser with an untyped global variable, a global variable inside a loop and on a collection inside an inline context.

Bug fixes:

• Test cases from some models with data types defined recursively could not be loaded.
• CMMN XML models that had sentry graphically depicted under the element to which they were attached could not be imported.

Bug fixes:

• Optimized the delay when saving a model with a large number of data types to a different place on a high latency internet connection.
• Some models published before version 11.10.0 and using the healthcare custom FEEL library were not loading properly.

Bug fixes:

• The execution of iterators on range on dates was not working as expected for some models.
• Test cases where a data type was changed from a Text to a Collection of Text could not be edited in the form.
• The des-cli command line utility now supports self signed HTTPS certificates.

Bug fixes:

• Models that were included using a prefix that collided with the prefix of a data type would not be included anymore.
• Models with a large number of data types would take longer to display the Test results.
• Publishing and testing an included decision model locally using included data types could report non-existing errors.

New and improved features:

• Implemented most of the changes from the new version of the DMN 1.5 standard.
  • The Decision Modeler now imports and export in DMN 1.5
  • The Digital Automation Suite now supports all of the new features of DMN 1.5.
  • It is now possible to include decision models without a prefix into the default namespace by specifying an empty prefix.
  • It is now possible to include custom FEEL function libraries without a prefix into the default namespace by specifying an empty prefix.
  • Support for the alternate graphical notation of Input Data in DMN that aligns with Data Objects of BPMN.
  • It is now possible to assign a specific type of range as a type.
  • The Test and Form functionalities now have proper support for ranges of different types.
  • Improve the support for subtyping of lists, contexts, ranges and functions in the instance of operator in FEEL.
  • The FEEL language now has an updated range() function that takes a string as a parameter (ex: range("[18..21)")).
  • Support the scientific notation to define numbers (ex: 2.3e-5).
  • Negative durations are now supported in the FEEL language (ex: -duration("PT1H")).
  • Suppor the @ notation for temporal expressions in the modeling environment (ex: @"PT5H").
• When loading a test case, modifying it and saving it, its original name will be offered by default.
• It is now possible to add external links in rich texts inputs in the modelers.

Bug fixes:

• Under some conditions, the selection of the data type contextual menu could be partially hidden under the data type selection dialog.
• Special character such as ‘<‘ were not visible when a FEEL expression was displayed as FEEL unparsed.
• When pasting a service task link to an operation with errors, errors are not properly linked.
• Replacing an existing term in a model could cause the modeler to hang.
• The translation model scroll bars are not always aligned with their container scroll bars.
• In the test cases, an empty string could be incorrectly serialized as a null in the test cases.
• Loading a test case with an empty list incorrectly trigger a detection of a type change.
• The file data type was not always properly saved in test cases.
• When entering a non FEEL value in the scripting language picker, reloading the modal could cause an incorrect value to be displayed as the scripting language.
• Creating a data type from the output of a decision service would always create it as Any instead of properly proposing a data type based on the Decision Service structure.
• Importing data types from a JSON schema now support importing nullable objects that define themselves as either a type or null.
• After completing a graph synchronization, it was possible that the current page would automatically switch without user interaction.
• The missing status of an element was not updated when that element was replaced by a newly dropped one from the graph.
• The severity level of messages at deployment times were aligned with the validation severities.
• Improved the delay when importing a JSON Schema in a model already containing multipe data types.
• When overing on the elipsis marker on a shape, under certain conditions it was possible that the wrong or no shape feedback was displayed.
• Term definitions would not follow a copy and paste in a different model.
• The email configuration channel save and test button were not properly saving and testing the configuration.
• The offline Trisotech Accounts option was not offered for the proper license type.
• The date, date and time duration and years and month duration FEEL functions were syntaxed highlighted in red incorrectly if a variable was used as a parameter.

Bug fixes:

• When synchronizing changes on large models, a performance issue could significantly delay the synching of an element.
• CDS Hooks would refuse calls without prefetched data.

Bug fixes:

• Fixed an issue where a ghost layer was prenventing the proper scrolling on a decision logic page.

Bug fixes:

• When exporting to DMN 1.4, some literal expressions were not converted from FEEL 1.3 to FEEL 1.4.
• When loading test cases, it was possible that the type matching was too eager resulting in an incorrect dialog of type incompatibility.
• It was possible that the initial values for user tasks to not properly filled when testing a process.
• When deploying a case to automation, some errors were not properly propagated to the user.
• Under some specific circumstances, the horizontal and vertical spacer were not properly following the mouse.
• The label placement algorithm for embedded stages was wrapping the text too eagerly.

Bug fixes:

• Some services (Workflow, Case, Decisions) generated with data types containing spaces were not always generating the system of engagement forms properly.
• Saving a test case with a list of list could not be loaded properly.
• Saving a test case with a text value of empty string would save it as a null value instead.
• Excel import of complex type in test cases could result in a list of letters for a collection of data types that were derived from text.

New and improved features:

• User tasks can now be visualized for a given execution environment.
• The REST API of the Digital Enterprise Suite now contains an additional resource that allows to query user tasks across services and environments.
• The Help now contains a documentation page with the possible operations on the Digital Enterprise Suite REST API.

Bug fixes:

• Downloading a Digital Distributed Container that did not exist returned an error message that was not properly displayed by the Docker client.
• Improved our support of advanced OpenAPI 3.0 constructs in the Operation Library import.
• When renaming an included model, inputs data in the Scope Panel were not renamed and we could not drag them.
• Comparing a KEM model that has a change on some properties fails.
• Editing a data type that has hundreds of components takes a long time.
• When importing a data type from the graph, sometimes the takes a long time to render the result.
• When sorting data types, reused data types were not sorted.
• The data type association is not kept when pasting a term in the Knowledge Entity Modeler.
• Renaming a data object without a name could fail under certain rare conditions.
• A process with only a data store in a pool/expanded call activity does not generate the proper BPMN XML export.

Additional notes:

Bug fixes:

• Subscribers of the Digital Enterprise Graph that did not deploy accelerators could loose access to the Digital Enterprise Graph.
• Fixed a documentation link that was broken for the technical documentation of the forms renderer in the Trisotech Help.

New and improved features:

• Integration with Microsoft Teams:
  • New notification channel for user tasks in workflow and cases.
  • Send teams messages as part of a workflow or case.
  • Receive messages from teams to resume or start a workflow or case.
• It is now possible to develop custom JavaScript widgets for the system of engagement forms (inputs and outputs). More information on building your own widget and tooling is available in our public GitLab repository.
• SVG export now generates SVG files without foreign objects to increase the compatibility with SVG editors.
• The image export label clarity was greatly improved.
• Publishing a service now generates an SVG image instead of a PNG.
• Publishing as an HTML report now uses an SVG image instead of a PNG for better quality.
• The PDF images generated now has clearer and selectable text.
• The performance of the full-text search in the Digital Enterprise Graph was greatly improved. It also now require a minimum of three character to start searching.
• When a WebHook emitter is disabled, the notification email now includes the last encountered error.
• The settings page to configure emitters now allows to download the audit logs directly from the user interface when defining an emitter of type Audit.
• The pager of the Cloud Execution Running instances was optimized for a large number of pages.
• The display of the name and tags of instances in the Cloud Execution was improved to completely display the information.

Bug fixes:

• APIKey identity encoding used in the query parameter were not properly encoded when it contained special characters.
• Fixed a problem that could cause an error when the automation was under very heavy load.
• The JSON schema data type import was not supporting reusable schema portions defined under $def.
• Types with the same name as FEEL types are not recognized when importing JSON Schema.
• Use the file name as a default for the type created from a JSON schema import when there’s no title.
• External link markers replace lock markers if a synchronization occurs while testing or animating a model.
• Custom FEEL functions associated data types did not properly autocomplete.
• Data type validation would not always be checked for collection on a constrained type.
• In FEEL, a for literal expression with multiple iterators depending on each other will now generate the right value.
• The validation would output an incorrect message when trying to access an attribute of a variable defined in a for literal expression.
• The number FEEL function can now be used with a variable number of parameters.
• Script tasks without data inputs were not able to properly access their inner data inputs variables at runtime.
• Allow null was removed from enumerations when importing a foreign DMN XML.
• Reusing a data type with a missing type could fail.
• Terms could be duplicated when reusing a term with a recursive reference.
• The transfrom and delete action buttons are not updated when changing the type of a decision logic.
• The dialog informing that a testing session was closed due to innactivity could not be dismissed.
• The content of a Knowledge Entity Model or Capability Model imported from Excel did not properly display in the Digital Enterprise Graph.
• Terms references were not always properly cleaned up when a local term was deleted.
• The instance of FEEL autocompletion now offers the option for list, context and functions.
• The instance of FEEL autocompletion was using display name instead of the actual data type name.