Release notes 2.2.2
Avola 2.2.2-20210419.1 was released on May 6th, 2021
Enhancements and bugfixes
As always there are a lot of small bug fixes in this version. Some of them are mentionable:
- Avola uses ‘Any’ as the hit policy for all decision tables. The DMN specification states that ‘the hit policy is summarized using a single character in a particular decision table cell’. That’s why we changed the ‘AC’ in the top-left cell of each decision table to an ‘A’.
- Ever tried to download all your business data from Avola and fell asleep while waiting for the Excel file to download? Good news: We significantly increased the download speed of this Business Data file.
- The name of a Business data can always be changed. However, when this business data was used in a formula, the name of the business data wouldn’t update there. This is now fixed. (Support ticket #618)
- Batch execution of test collections was very slow since the last release. We recovered the caching of the decision models for batch executions. With this release speed of execution is back to normal (up to 17 times faster!).
- We were noticed by a customer that the sort order of an Execution Search could be confusing. In order to prevent this, we changed the default sort order. Most recent executions will be shown first. Also we added the paging bar at the top of the result page to make it clearer that there may be more results than you see on the page. (Support ticket #2087)
- In the Execution Search you can reset the search criteria using the button ‘Clear’. Not all of the old search data was properly cleared by clicking it. From now, when you hit ‘Clear’, your search query is clean as a whistle.
- In your decisions you can calculate a TimeSpan using two input parameters of type DateTime. When one of those dates contained an empty value the calculation of the entire decision failed (‘Nullable object must have a value’). The TimeSpan calculation will from now on return a ‘No Conclusion’ in this situation.
- When you add an input parameter to an existing model, this new parameter will be added as an empty value in existing test scenarios. Regardless of the data type of this new parameter.
- Test collections can now be executed (again) without dependencies. You will not see the internal server error when executing the scenario(s). (Support ticket #2114)