I want to add some more testing to my packages. It seems the Seattle team are using RUnit. Are people happy with this? In case a unit test fails during the daily build on the Bioc servers, is the reported output on the web interface good enough that it is clear what is happening? Thanks, Kasper