[net.micro] MS-DOS technical brain teaser

lotto@talcott.UUCP (Jerry Lotto) (06/27/85)

	Due to the underwhelming response to my last inquiry, I am
rephrasing the question (Microsoft employees, are you listening?). MS-DOS
provides basic memory management capability via DOS function 48H et al. I
have found that COMMAND.COM when loading an .EXE program does NOT use this
capability to tell DOS to leave the stack (segment) alone as it does for
the code and data areas. Since the stack size AND location is available
to the command processor from the .EXE program header, unlike the
situation loading .COM files, I would expect that COMMAND.COM would deal
with this a little better. What goes on here? Is there a reason why this
cannot be done?

	Microsofts technical support line has cost me a literal fortune in
MUSAC, with no response (how is 40 minutes from MA to WA with nothing but
"You Light Up My Life", followed by a disconnect?). I can no longer afford
this "free" help. Almost everyone else in an "official" capacity from Boca
to Bellevue is hard pressed to understand why I want to know this.

	Another note: the new symbolic debugger from Microsoft is a great
program, but it does not appear to be compatible across the PC-DOS line.
Attempts to use it under PC-DOS 3.1 on an AT have met with many system
crashes and very little debugging. When this problem is resolved, I will
move over to this family of compilers and development tools very quickly.

	Thanks in advance for the response.

-- 
____________

Gerald Lotto - Harvard Chemistry Dept.

UUCP:  {genrad,cbosgd}!wjh12!h-sc4!harvard!lhasa!lotto
       {seismo,harpo,ihnp4,linus,allegra,ut-sally}!harvard!lhasa!lotto
ARPA:  lotto@harvard.ARPA
CSNET: lotto%harvard@csnet-relay