I’m a bit late to this thread - but I slightly disagree with this statement. In my experience, Launchpad is usually slower than OSGi testing, for two reasons:
- In Launchpad you’re generally firing up the framework for each test, with OSGi testing you only fire it up once for the entire test run;
- With OSGi testing, you can take this a step further by using continuous testing - where the test framework stays running in the background across many test runs, and then automatically re-runs tests when you change the code. If your framework is well designed and the code you’re working on has few downstream dependencies, this results in a very tight dev-test-dev-test cycle.
I’ve come to prefer OSGi testing rather than Launchpad for most of the tests that I write.