The notes below provide descriptions of the new features and changes introduced with each release of Eggplant AI. You are strongly encouraged to read about the relevant changes whenever you upgrade from an earlier version.
Potential compatibility issues are highlighted in RED.
Supported Operating Systems
- Windows: Windows 10, Windows Server 2012, Windows Server 2016
- Linux: Red Hat Enterprise Linux (RHEL) 7.6
- RAM: 8 GB minimum
- Disk Space: 4 GB minimum, 5 GB for /opt minimum (Linux)
- Processor: 2.3 GHz or faster
Note: Eggplant AI 3.0 is not compatible with versions prior to Eggplant Manager 6.43.
Release 3.0.5 (1 October 2019)
- Fixes an issue where the Select Model drop-down list failed to show the models when switching between selections.
- Resolves an issue that prevented users from manually entering the script path in the Snippets Called by Action section of the Eggplant AI UI.
- Fixes an issue that prevented model execution with multiple iterations via the API.
- Fixes an issue where the initial value set for submodel variables was ignored during a test run.
Release 3.0.4 (1 August 2019)
- Improves the autosave functionality to prevent errors when saving a model.
- Fixes an issue where reverting to any version from the model history didn't display the model automatically.
- Fixes an issue with the initial value not being assigned for Set type variables during model execution.
- Fixes an issue that prevented model execution when changing a variable from Set to Text while the value groups were not edited to suit text variables.
- Improves support to import models prior to Eggplant AI 2.3.
- Fixes an issue where the list of variables appeared truncated when adding input and output parameters for a snippet.
- Resolves an issue with Eggplant AI sending an empty response to Eggplant Manager when a model with a snippet resulted in error.
- Improves communication between Eggplant AI and Eggplant Manager.
- Fixes an issue where no error message was displayed when a submodel attempted to call its parent model recursively.
Release 3.0.2 (4 July 2019)
- Fixes an issue where aborting a model run at different stages caused Eggplant AI to report inconsistent run results.
- Resolves an issue where the Eggplant AI agent lost connection from Eggplant AI when aborting a model via Eggplant Manager.
- Fixes a problem where a model continued to execute even after a snippet error was reported from Eggplant Functional.
- Fixes an issue where tests systematically failed for admin users when running them through Eggplant Manager.
Release 3.0.1 (20 June 2019)
- Fixes a problem with test cases where the parent test case was not passing variables to a test case in submodel.
- Resolves an issue with text variables not being generated when generate is selected in Set and Generate Variable Values.
Release 3.0 (19 June 2019)
- Adds a Designer page where users can create a model interactively from their live SUT screen and edit existing SenseTalk snippets from an associated Eggplant Functional suite.
- Adds a new Home page to the Eggplant AI UI to show the summary information about reports and analytics available to users.
- Adds support for an alternative location for Run Report log files when running models from Eggplant Manager.
- The Insights page now provides a Run Report section where uses can view the summary report of tests that have been run against a model.
- The Analyze button in the Test Cases tab now also reports test cases that include submodels.
- Adds capability to include submodel test cases when executingDirected Test Cases.
- Test logs are now saved in the database instead of the file system. Note that this change doesn’t affect the ability to retrieve logs via the API calls.
- Adds new API endpoints to manipulate test cases so that users don't need to know the model’s JSON structure.
- Updates the API's Run endpoint to change the default value of verbosity parameter to show full logging in the UI.
- During model execution, Eggplant AI now includes submodel coverage as part of coverage calculation.
- The base URL of the Eggplant AI Server in the UI is configurable to allow running of Eggplant AI UI and Eggplant AI Server on separate machines.
- Agent names are not required to be unique anymore.
- Adds support for Eggplant AI as software as a service (SaaS).
- Several bug fixes and enhancements to licensing and handling of executions to improve customer experience.
- Running a model with submodels now won't exhaust your concurrent licenses.
- Resolves an issue where Eggplant AI did not release execution licenses for incomplete tests.
- Improves handling of test case dependency order.
- Prevents naming of test cases with invalid characters.
- Adds support for executing a model containing constraints on Record variable type.
- Fixes an issue with the Manage Agents window that prevented it from displaying.
- Improves the Run Report tab to show the reports of all the Test IDs shown in the Test Cases tab.
- Improves the autosave functionality when switching between states and actions in a model.
- Fixes an issue that caused Eggplant AI to prefer sequent actions over states with normal actions when an action is connected to both states and sequent actions.
- Fixes an issue that prevented users from selecting None from the Report when a snippet executes drop-down list on the Settings dialog box.
- Fixes a problem with loading of the sample Eggplant AI model.
- Fixes an issue with a model failing to run if the Seed field on the toolbar contains no value.
- Fixes a problem with clicking Exit in the Eggplant AI Agent window now does not throw an Eggplant Functional error.
- Improves the Exception state so that it can be created only if there is no Error state defined in the model.
- The Pass rate percentage in the Eggplant AI Run Report tab now doesn’t include incomplete run information.
- Improves handling of expired sessions in Google Chrome.
- The test runs summary on the Run Report tab now shows the correct value against the number of passed test runs.
- Improves the submodel logic to handle top-level sub model errors.
- Improves the installation instructions (README.md) for the Raw Python agent packaged with the Eggplant AI installer.
- Closing the browser after logging into Eggplant AI will no more end the session for users.
- Fixes a typo in the Coverage Model tab that replaces All Paths with All Pairs.
Release 2.3.2 Linux (25 March 2019)
Initial release of Linux package for RHEL 7.6.
Release 2.3.2 (11 March 2019)
- Directed test cases can now run snippets that are attached to states.
- Resolves an issue where loading the sample model caused the currently open model to be overwritten.
- Fixes an issue with automatic saving of a model when moving states and actions around within that model.
- The Run Report tab now shows all the test runs based on the test IDs shown in the Test Cases tab (Test IDs column) when filtered to display all versions.
- The Pass rate percentage shown at the bottom of the Run Report tab now includes only complete test runs.
Release 2.3.1 (28 February 2019)
- Improvements to Release Insights to make the Git RI-Connector optional.
- The Eggplant AI integration with Eggplant Manager now supports backward compatibility (until Eggplant AI 3.0).
- Updates the Run Report method in the API to prevent generation of internal server errors under certain conditions.
- Fixes an issue with the Test Cases tab displaying only the latest test cases even when filtered to display all test cases.
- Users can now switch their Eggplant Cloud credentials in the Performance Tests tab.
- Fixes an issue that prevented the test case reports from showing in the Run Report tab.
- Fixes an issue that caused licenses with line breaks to fail.
- Resolves an issue with loading models/model tags created by user accounts that do not exist in the system.
Release 2.3 (11 February 2019)
- Enhances the core Eggplant AI engine to achieve greater coverage more efficiently.
- Adds the Bug Content and Dev Quality predictors to the Release Insights section of Eggplant AI.
- Test Cases now show additional metadata (Description, Tags, External ID, and Origin), which allows enhanced sorting and filtering, and provides additional context.
- Release Insights now provides a dedicated Test Case report that lets users view progress toward completion for each Test Case group.
- Adds the SUT last modified field to the Action Properties tab. When you enter a date in this field, Eggplant AI disregards coverage of that action achieved prior to that date.
- Admin users can now configure and name their own coverage models.
- Adds the option to disable global actions until the initial state is exited.
- Users can now create, execute, and analyze Eggplant Performance tests directly from the Eggplant AI interface.
- Resolves inconsistencies between coverage calculations for the All Nodes coverage model.
- Fixes an issue that stopped users from replacing the default error state with a custom one.
- Fixes a problem that prevented the use of a port other than 80 on Windows 10 machines.
- Resolves an issue where the Exit state never executed if an action was connected to multiple states along with the Exit state.
Release 2.2.0a (9 November 2018)
- Fixes an issue with command-line arguments being ignored when launching the MacOS Agent.
Release 2.2 (25 October 2018)
- Submodels can now be run within actions, allowing users to define a hierarchy of different levels of model to describe their application.
- Eggplant Release Insights can now be launched directly from Eggplant AI, providing Coverage and Bug Hunting views of current testing progress.
- Users can store a set of Eggplant AI Agent connections and switch easily between them.
- Users can now choose which type of heat map to overlay on the model: Coverage or Failures.
- Improved model management allows users to apply version tags to models and provides the capability to revert to previous versions.
- Action and State descriptions can now have multiple lines, and display on hover-over.
- Selecting the Remember Me checkbox on login now keeps a session open for 24 hours.
- Fixes an issue with adding checks to actions.
- Improves performance when the snippet list loads.
- Removes a restriction on using records when snippets are disabled.
- Test cases are now validated when saving a model.
- Fixes an issue where the Run Report tab would not automatically update to show the results of a newly saved and run model.
- Fixes an issue where the splash screen occasionally still appeared even after a user chose to disable it.
- Fixes an issue with updates to connection conditions in the Action properties panel.
- Fixes an issue where a directed test case would fail with a syntax error if a State was one of the steps in that test case.
- Fixes an issue with displaying log details for test cases in Chinese and Japanese Windows 10 environments.
Release 2.1 (11 June 2018)
- Enhances coverage algorithms, with paths up to five steps deep through the model, as well as pairwise variable coverage.
- Enhances naming of States, Actions, and Variables to accept full Unicode (UTF-8) character set.
- Model action selection algorithm now focuses on test cases that have not been hit in previous tests.
- Model execution now records the time taken for an action to execute and allows users to specify a maximum execution time.
- Tests can now be started and monitored via a REST API.
- Log output from Eggplant AI is now formatted the same way as log output from Eggplant Functional.
- Fixes an issue with Eggplant Functional global values not persisting from Action to Action.
- Fixes an issue with saving models with incomplete variable data.
- Fixes an issue with precondition values not being cleared after the conditions are set to "None".
- Fixes problems with incomplete Tag definitions.
- Fixes several problems with updating and persistence in Table view.
- Improves robustness for the Agent shutting down Eggplant Functional.
- Fixes sorting of snippets.
- Fixes an issue where toggling between show all and show used in Set and Generate Variable values table caused values to be reset.
- Fixes coverage issues when the same model exists in different groups.
- Fixes an issue where non-unique handlers were displayed incorrectly on the Snippets pane.
- Fixes an issue with the heat map view getting stuck when all reports are removed.
- Removes the additional remove option in table context menu.
Release 2.0.2 (8 March 2018)
- Adds support for Microsoft Internet Explorer 11 and Microsoft Edge browsers.
- Adds a progress message that appears while large models are loading.
- Fixes a problem where the AI server became unresponsive when a test case executed or when visiting the Test case tab.
- The Test Case Builder now correctly specifies global actions.
- Fixes a problem where global actions didn't execute correctly in directed tests.
- Improves the performance of test case analysis.
- Fixes licensing issues for users in some locales.
- Fixes an issue where opening a Run report could freeze an active model execution.
- Fixes a problem in test cases with more than one instance of an action where deleting a second or subsequent instance of that action deleted the first instance of that action and breaks the order of the test.
- The sample model now loads from the Help menu.
- Fixes a problem where opening a saved model failed when another saved model was loaded.
- The breakpoint icon now refreshes if the shape or name of the state or action box it is in changes.
- Snippets now display in alphabetical order.
Release 2.0 (12 February 2018)
- Improves Eggplant AI algorithms to focus on bug hunting.
- New Test Case Builder allows users to quickly create test cases in the GUI using drag-and-drop.
- Users can now add custom tags to actions, states, and variables.
- Adds the capability to perform Directed Test Cases, which execute specific user journeys.
- Adds the option to request a password reset to the logon screen.
- Actions and states now display icons to indicate that snippets are connected.
- Adds the capability to launch Eggplant AI Server automatically or via Windows Scheduler.
- Fixes a problem where the sample model fails to launch from the Help menu.
- Updates the path to the Eggplant Functional runscript and Eggplant commands in the Eggplant AI agent.
- Fixes a problem where models won’t run if a test case is created without being defined.
Release 1.0.2 (8 December 2017)
- Allows access to record initialization files even when no suite is defined.
- Fixes an issue with the SenseTalk file list from the Agent running on Mac.
- Resolves a bottleneck with coverage calculation performance.
Release 1.0.1 (23 November 2017)
- You can now manually enter a path to an Eggplant Functional snippet or handler from the Snippets called by action or Snippets called by state section of the Properties tab.
- Fixes a problem where Eggplant Functional folders didn’t show up in the Snippets list.
- Fixes a problem where selecting Do not show this again in the splash screen had no effect.
- The splash screen no longer appears when a login fails.
- Fixes a problem where adding a Set variable with no parameters caused an error.
- Fixes a problem where a sequent global action wouldn’t be called if it was connected to both a global action and a state.
- Values in the record file that contain special characters are now placed in quotation marks before they are passed to Eggplant Functional.
- Fixes a problem where deleting handlers from an Eggplant Functional script caused Eggplant AI to stop seeing handlers in other scripts.
- Fixes a problem where creating or removing breakpoints from a model caused an error.
- Fixes a problem where continue functionality was lost when a model reached a breakpoint.
- Fixes a problem where modifying an action’s weight in Table view caused an error.
- Fixes a problem where links originating from other states weren’t recreated when the deletion of a state was undone.
- Settings that are not required are no longer flagged as required.
- Fixes a problem where the Eggplant AI UI became inactive during the addition of multiple connections to sequent actions.
- Clicking Abort on a paused test now closes that test without additional user interaction.
- Initial values for generated variables are now disabled.
- Fixes a problem where Eggplant AI reported the seed number of a test that was rerun as 1 rather than the seed of the test that was rerun.
- The default group for saving an Eggplant AI model is now the user’s name.
- Fixes a problem where return values from connected Eggplant Functional scripts were not being returned to Eggplant AI.
Release 1.0 (25 October 2017)
This is the version 1.0 release.