Member rambrain::cyclicManagedMemory::swapIn (managedMemoryChunk &chunk): in rare cases, these two lines are necessary as the readEl is also filtered. We should find out if we can leave this out somehow more elegant.
Member rambrain::cyclicManagedMemory::swapOut (global_bytesize min_size): investigate if subtracting swapped out preemptive bytes is affecting performance ( too much preemptive action possible ). Naively testing, this is not the case.