sub-title=History

Version 2.9.1:

0. Bug Fix:  Command line arguments that took parameters, such
   as /s{filename} previously could dereference a null (or un-
   initialized) pointer if the second parameter was missing.

0. Bug Fix:  Fixed possibility of a memory overrun error during
   the reading 16-bit FATs which could have halted COMMAND.COM.

0. Bug Fix:  Inserting a file with the 'I' command improperly
   handled the file buffer becoming full and if full, could have
   inserted too much data.

0. Bug Fix:  The reading of the volume label was inconsistant.

1. Added the display of disk Serial ID and File System information.


Version 2.9:

0. Bug Fix:  When changing a sector, the '=' command tested the
   argument for greater than the buffer size rather than greater
   or equal to the buffer size which could let the buffer be changed
   past the end.

0. Bug Fix:  The physical to logical calculation was in error
   when calculating to sector numbers larger that 65535.  This
   showed up only when using the 'next/back-track' and the 's'
   commands.

0. Bug Fix:  My calculation for setting to head 0 (when using the
   'h' or 's' commands) was wrong and would result in a "sector out
   of bounds" error at either the first track or the last track on
   hard drives.

   This was due to number of "hidden" sectors.  On drives where the
   number of hidden sectors was not a multiple of the sectors per
   track the error would happen on track 0, and on others on the
   last track.  (Hey, why do drives/DOS do such stupid things!)

   Actually, it makes (some) sense when the physical size of the
   drive is not a multiple of the cluster size, i.e. there is some
   "extra" space (or something like that).

   Anyways, I left the phyisical set commands in and just put some
   bounds checking tests around the calculation to "fix" it.  The
   drawback is that, 'h' near the end of the drive may go to the
   last sector only.

0. Bug Fix:  With a drive command line argument, DISKED did not
   restore the default drive upon exiting.  This is the second
   time I fixed this!

0. Bug Fix:  The 'P' command did not re-read the current sector
   when finished and displayed the wrong data.  If the starting
   sector of the 'P' command was 0, the go back back command, '^Q'
   tried to go to an out of bounds sector.  Both are fixed.


Version 2.8:

   I forgot what I did with this version.


Version 2.7a:

1. Fixed CRASH bug when <more> displayed for user to press any key
   to continue.  The macro clreol() cleared 0x4F00 columns rather
   than 0x4F.


Version 2.7:

1. Removed extra CR/LF on full sector display.

2. Disk search finds over sector boundries.


Version 2.6:

Changes (most significant):

0. Bug Fix:  Get sectors display numbers (minor fix).

0. Bug Fix:  Edit directory bug -- =0 selected out of bounds.

1. Set to default directory upon startup.