02.10.2014 21:21, Alexander E. Patrakov wrote: > I have looked at the patch, but came to the conclusion that I cannot > provide a good review now. Forgot to express one more point clearly: due to not doing any processing, module-virtual-sink is not a good model to test fixes on. This just leads to writing code that relies on assumptions that are invalid for other virtual sinks. OTOH, I am not sure if other sinks are good for testing this type of patches on, either, for other reasons such as non-audited latency-reporting code. -- Alexander E. Patrakov