Neodymium 5.3.0
Neodymium 5.3.0 is mostly a bug-fix release with which we try again to eliminate bugs from our code, but we also add some new features to, well, try new things.
5.3.0
- Feature: Retry Flaky Tests
- Flaky tests, caused by minor test or application issues, can be a major drain on productivity and undermine the general perception of the test reports. To stabilize such tests, we added the
@Retryannotation, allowing the test automation to retry a specific test on specific error messages.
- Flaky tests, caused by minor test or application issues, can be a major drain on productivity and undermine the general perception of the test reports. To stabilize such tests, we added the
- TODO capabilities and video issue TODO
- Data Set Ranges: some
- If you are handling a huge number of data sets but want to run your test just for a certain range, handling such test data can be quite difficult with only IDs or indices, so we introduced a range for these test cases. Just add
@DataSet((3,16))to your test to run the test with data sets 3 to 16.
- If you are handling a huge number of data sets but want to run your test just for a certain range, handling such test data can be quite difficult with only IDs or indices, so we introduced a range for these test cases. Just add
- fixed bug #335 - Fixed an issue with our URL filter on excluding certain paths.
- fixed bug #341 - When highlighting the viewport, we always blurred the non-viewport part, which could hide important error messages. This is now configurable.
- fixed bug #354 - Our Lighthouse audit assertions missed part of the list due to different JSON structures; this is now fixed.
- fixed bug #355 - Our Neodymium random seed was not reset after a test run, so it was only possible to rerun a full suite with the same seed instead of single tests. This is now possible.
- fixed bug #354 - The configuration value to take a screenshot on every step was not working in the last version but has been fixed.
- fixed bug #360 - We have different ways of retrieving test data, and not all of them ended up as an attachment in the report.
- fixed bug #363 - Running a test without a specific browser configuration used the Selenide WebDriver and broke a lot of features. Now,
@Browserenables the default browser with videos and everything. - fixed bug #364 - Our property substitution was not working correctly in the Allure environment section.
- fixed bug #369 - Our pop-up blocker did not work properly with SVG links. We have also done some optimization of the whole implementation.
- fixed bug #390 - Our test data view in the report showed issues with long values that were outside the JS number range. We added a workaround to handle this.
- fixed bug #393 - The
beforeEachandafterEachmethods in JUnit 5 caused some problems, which are now solved. - fixed bug #383 - In our default browser, the Chrome password warning was sometimes interfering with site handling; we suppressed this by default.
- fixed bug #396 - Special circumstances like open alerts, broke the video recording.
- fixed bug #397 - There is a system property that can enable/disable browsers from a CI/CD pipeline. This did not work previously, but does now.
- fixed bug #357 - We had to remove some links to our old repository name.