KB19516, and the other KBs linked at the bottom of it, are interesting and relevant, and may provide a level of detail which coordinates with what MCL and CO have informatively contributed to this discussion. (Since apparently not mentioned before, I'll just note that "fragmentation" is related to the designed arrangement of how file blocks are laid out on the surfaces of the disk platters. Ie, a file's blocks may well not be strictly contiguous due to eg the relation of read or write timing to disk rotation speed. So fragmentation may be said to be when the way a file's blocks are recorded on the disk is not the optimum.) On Fri, 20 Dec 2002, Mark C. Langston wrote: . . . > What I'd really love to find is a document laying out the > data structures of HFS+ and the atomic operations, without requiring > that I go download and pore over the Darwin source. . . .