Thanks to a trouble report from Tony Anderson, I've done some research on
major conflicts between AltLCD routines for the M100/102 and the ROM version
of Traveling Software's TS-DOS program.

   The first incompatibility is blatant: TS-DOS ROM uses a chunk of the AltLCD
area as an input-output buffer for communication with the TDD.  So, an
m/l routine stored in the AltLCD won't survive any operations that access the
drive.

   The second conflict is more subtle, and it can arise even when there is no
attempt at TDD access.
   It is standard practice to ORG AltLCD routines at 64704.  But, when you set
"DOS ON" with the TS-DOS ROM, at least one BASIC command (OPENing a .DO file
for any read/write operation) pokes new values into addresses 64706 and 64707.
   It also seems that "DOS ON" installs hooks in other BASIC keywords, and
routines at the new vectors make reference to the bytes at 64706 and
64707.
   So you've got a nasty situation if the "DOS ON" switch is set at the
same time you're running a BASIC program with two elements:

	a) any OPEN statement, and
	b) an AltLCD routine ORG'd at
	   64704

   If you poke or LOADM your m/l code =before= the OPEN, TS-DOS overwrites at
least two bytes of your program.  If you open the file first, and =then=
install your AltLCD code, you've screwed up the reference bytes that other DOS
hooks are trying to read.
   The sure-fire solution is to turn TS-DOS's switch "OFF" before running code
in the AltLCD.
   Beyond that, you could ORG your Altbuf code to an address above
64707.  I have not found any other Altbuf locations that TS-DOS routinely
changes (although it's always possible I've missed something).
   If anyone has related information that should be included in this file,
please pass it along.  Thanks.
	Wilson Van Alst [76576,2735]