Thanks for those kind words. It is getting a bit long in the tooth as shown by the references only to PPC processors, but the principles are still valid.
The disappearance of the "green", Free Memory, simply means that all of the physical RAM has been Mapped into applications as either Wired, Active, or Inactive. At the point when there is insufficient free RAM for an active task, data in Inactive RAM will be written to the swapfile(s), ie "paged" out, and the RAM thus freed up plus any code or instructions in Inactive memory will be overwritten by code or data demanded by the active application(s). If there is insufficient Inactive RAM then an active application or task will have to be made inactive and paged out to memory. The algorithm used to determine which application is to be paged is generally based on least recently used and/or least often used but I do not know the specific algorithm used in OS X. In any case the next time the paged task(s) become active this process will have to be reversed and something else will be paged out so the reactivated task can be paged in.
The terms paged in and paged out refer to the fact both RAM and disk memory are not allocated or mapped on a byte by byte basis rather on a page by page basis where a page is 8 KB.
