db> The FAT file system is an abomination - in theory, you can have any
 db> number of positive integer FATs from one upwards .. but most software
 db> is hard-coded for two, and expects them one after the other.  Of
 db> course, by the time one is corrupted the data is copied straight over
 db> the second, corrupting it as well.

 db> The fact that Paul was able to recover so much is testimony to just
 db> how reliable HPFS really is when it comes to the crunch.  With FAT, he
 db> would have had no chance.

i recovered a system once where it looked as if they had been hit by a
power failure halfway through rewriting the fat. the data was all there,
but it was just as if the second half had been bit shifted, probably due to
a phase difference between the old and the new data. aha i thought just
copy the second fat over the first! well it let me get all the data off the
disk onto a tape, but i had to reformat the disk before i could write to
it, just kept on coming up with "disk full!". i never found out
why as the owner of the shop whose pos system it was was rather anxious to
havce hid system back. it does show, though that you can get quite a lot
back from any system if you have the motivation. 

i do have to admit however that microsoft did quite a good job when they
designed hpfs (:

                     keith
@EOT:

--- WinPoint 0.2.07 Alpha