From: vdjole@EUnet.yu   
      
   Hi;   
      
   When displaying the contents of a directory, or information   
   about a selection of files to be e.g. copied (in a desktop program   
   or in a file copy program), it is usual to display the total count   
   of "bytes used" in the selection. This number is (almost always?)   
   obtained by summing the "useful" bytes of all relevant files, which   
   can results in numbers of a magnitude of several billions that appear   
   to be accurate down to the last byte.   
      
   It might be argued that this is not always the best approach,   
   because each file actually occupies an integer number of clusters   
   on the disk, and a cluster can have a size of several kilobytes.   
   The count of used clusters, multiplied by cluster size in KB, may   
   actually give a more meaningful and easier-readable information about   
   the really occupied space on the disk. It would be a more accurate   
   measure of the effort required to perform a copy (or other similar)   
   opeation and enable easier estimate of wheter free space on the target   
   location is sufficient. Beside that, it would in a simple way avoid   
   the problem of overflowing the 2 GB or 4 GB limit for total byte count   
   that appears when using simple 32-bit integer arithmetic.   
      
   On the down side, this approach would be contrary to the "tradition"   
   originating from MS-DOS (or CP/M ?) o displaying the sum of "useful"   
   bytes.   
      
   It may be noted that, on large disk volumes with large cluster sizes   
   the difference between summation of "useful" bytes and calculation   
   based on summation of used clusters can be significant.   
      
   The proposition therefore is that, when a desktop program displays   
   information about  "bytes used" in a directory or "bytes to be copied",   
   the used clusters are to be counted and summary information is expressed   
   in kilobytes (except in very rare cases, cluster sizes appear always to   
   be at least 1KB). Individual file sizes would still be displayed as   
   "useful byte" counts.   
      
   Any opinions?   
      
   cu;   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)