It seems that Extras-testing is not currently working very efficiently. There are many applications in there that have been there for very long time. While proper testing of applications is necessary, it serves nobodies interests if takes too long. This especially problematic for updates of existing applications.

For example: Lets presume there is App X that has finally been promoted form Extras-testing to Extras. After a while, it is reported that it has an critical fault that affects not only the application but also the entire system. Sure enough, the developer notices that he had a typo in his code and quickly makes the fix. He then uploads it to Extras-devel. This time his fortunate, as it took less than an hour to appear there and he immediately promotes this to Extras-testing. This takes a bit longer, but in 5 hours, the update is finally in Extras-testing.

However, now the real problem starts. This critical update can reside in Extras-testing for days, weeks of even months! Of course, the developer could always recommend for the end-users to temporally activate Extras-testing to receive this critical update, but that would hardly be a good solution.

Brainstorm Link

This thread is relevent to this brainstorm.

This thread may also be.

I was vaguely wondering something about portable votes, but you'd need to know exactly what people are voting for: usefulness, UI etc. Those votes should be portable to an update, if you could identify them.