Releases: aerokube/moon
v2.2.0: logLevel capability
Update Notes
This version introduced secretRef field in configuration object (kubectl get configs -n moon) custom resource definition. In order to use it - you need to delete and apply Moon custom resource definitions.
New features
Fixes
- #319 -
hostAliasdoes not work in Playwright \ Cypress
Improvements
- #322 - bump Golang to 1.18
v2.1.2: license manipulation fixes
v2.1.1: Selenium Devtools support for non-standard load balancers
New features
- #314 - Selenium Devtools support for non-standard load balancers
v2.1.0: Selenium 4 Devtools Support
This release adds Selenium 4 BiDi \ Devtools support which should work out of the box. An example test can be found here.
v2.0.0: first release of second Moon generation
Important
This release is completely different from 1.x series. The most important changes are described here. Trying to deploy Moon 2 on top of Moon 1 will not work because configuration formats of these two versions are incompatible. Moon 1 documentation is now located here.
New features
- Multiple Kubernetes namespaces support
- Improved configuration using custom resource definitions
- Improved browser performance and lower resources consumption
- Improved additional trusted TLS root certification authorities support
- Advanced Selenium features like copy-pasting images to clipboard
v1.9.6: Kubernetes 1.23 support
- #311 - Kubernetes 1.23 support
v1.9.5: OpenID connect status code fix
Fixes
- #306 -
moon-authdoes not work with some Java clients
v1.9.4: case insensitive user groups
Improvements
- #302 - case insensitive user groups in OpenID Connect
v1.9.3: using FQDN in DNS resolution
Improvements
- #295 - preferring FQDN in pod domain name resolution
v1.9.2: better Selenium timeout logic
Important
Starting from this release Selenium session idle timeout works a bit differently. Session is now considered idle between the end of previous command and the start of the next command (previously idle time included command execution duration too). The most visible difference can be with page loading in the browser. Previously too slow page loading (sometimes caused by missing firewall rules) led to session timeout and broken tests. Now the same will run forever and should lead to test case timeout. This should allow to easily make distinction between too long idle time between Selenium commands and issues with system under test.