In order to make the testing/unstable system work, which has a big impact on our release process, people need to use unstable. If people don't use unstable, bugs are found in testing making unstable nothing more than a build phase and an unnecessary latency. The unstable/testing system only works if a reasonable amount of all bugs is found in unstable. Perhaps we need to talk more people into using unstable (and experimental while we're at it).

I did some primitive statistics and it currently looks like the number of bugs reported in unstable vs. the number reported against testing is almost 1:1. Whatever that means ...