The #ValueView repository on GitHub is obviously not set up to run the QUnit tests during CI. Over the pasts weeks we did ~~two~~ //four// releases with broken tests but did not realized until we tried to use the release on Wikibase.git, where the QUnit tests //are// executed and fail. This just does not make sense and does nothing but causing pain. Let's say a ValueView change does contain a breaking change, this is documented and there is even a patch that adapts the Wikibase code base on Gerrit, why should this Gerrit patch be blocked because of a bug in a **test** in an **other** component?
Possible solutions:
1. Make the QUnit tests run on GitHub/Travis so breaking changes can not be merged unnoticed.
2. Move ValueView to Gerrit/Jenkins if GitHub/Travis can not run the QUnit tests. Jenkins obviously can.
3. **Exclude the ValueView tests when Gerrit/Jenkins runs the tests.** In my opinion this is the most simple solution when the first two are not possible for some reason.
4. Set the QUnit tests on Jenkins to non-voting.