BD> 19 Feb 98 20:01, Tony Dunlap wrote to Bob Davidson:
 TD> That doesn't worry me as much as the apparent speed loss of FAT32 vs
 TD> FAT16. I beleive this may be because more "read operations" are
 BD>  One of the things I watched closely when I converted 
 BD> to FAT32/4K was for a difference in speed. If there is 
 BD> a difference, it is so minute as to be unrecognizable. 
 BD> I'm going to leave the hard drive I toss mail on for 30 
 BD> days before running defrag. This may give me a vague 
 BD> ideas as to what/if there are any differences worth concern.
I can't see any discernable difference in mail tossing speed with SQUISH 1.11 
DOS tossing mail in Windows 95 on my home system. It still tosses 15-35 
msg/sec typically doing around 20-25 with the small echomail bundles (Small 
bundles-I never toss more than 100-300 messages) I send to the home system. 
I've not noticed Exel (large spreadsheets), My DOS 9 year data base, games, 
anything else running any slower or faster than with FAT 16. My son says that 
it doesn't seem to make any difference in the speed his games run at in DOS 
only mode either, other than he has found a couple of games out of more than 
50 that won't work with FAT32 (No biggy, they were old games that he never 
used much anyway).
It seems to me that there really is very little difference between FAT32 and 
FAT16 in fragmentation or performance except cluster size that I've ever been 
able to notice. Seems that that is a consensus from people in this conference 
that have switched from FAT16 to FAT32 with Windows95 and spent a little time 
with it. Not scientific, just seat of the pants observations.
But both NTFS and HPFS are -far- superior file systems to FAT32 in both 
performance and storage efficiency. Such is the price one pays for "backwards 
compatiblity" with obsolete 15-25 year old software technology .
--- Maximus/NT 3.01b1
---------------