Well, not having comments to understand what you are doing is pretty
tough when you don't have the time to analyze the code line by line. But
IIRC, the common way to check for the presence of an ANSI device driver
was to check via an INT 2Fh (multiplexer) call (don't recall the exact
call value (AX/AH)).
Will take a closer look at this later this evening when I am back, I am
heading out to this year's SCaLE conference... ;-)
Ralf
On 7/29/2022 2:14 AM, Jerome Shidel wrote:
Hi All,
While recently performing a re-install of everything (System Commander
Multiple DOS, Windows, Linux booting) on my Pentium Pro, I discovered
a minor compatibility issue regarding ANSI.SYS support. Without
digging into it more, I cannot say if the issue is NANSI.SYS or
elsewhere in FreeDOS.
Basically, one of the versions of my ancient (written back in the
early 90’s) Turbo Pascal directory listing programs used ANSI escape
sequences to provide color. Instead of blindly assuming ANSI support
is present, it first probes DOS to verify the functionality. It
performs the probe through standard DOS I/O function calls and works
on other DOS platforms (such as MS-DOS, PC-DOS and DR-DOS). However,
the test fails under FreeDOS.
function DetectAnsi : boolean; assembler;
asm
MOV DI, OFFSET @AnsiTestStr
@2:
MOV DL, CS:[DI]
INC DI
CMP DL, 0
JE @3
MOV AH, $02
INT $21
JMP @2
@3:
MOV AH, $06
MOV DL, $FF
INT $21
JZ @5
CMP AL, 27
JNE @5
MOV AH, $06
MOV DL, $FF
INT $21
JZ @5
CMP AL, '['
JNE @5
@4:
MOV AH, $06
MOV DL, $FF
INT $21
JZ @5
CMP AL,'R'
JNE @4
@5:
MOV AH, $06
MOV DL, $FF
INT $21
JZ @6
CMP AL,13
JNE @5
MOV AL, True
JMP @Done
@6: { Not Found }
MOV DI, OFFSET @AnsiTestStr
@7:
MOV DL, CS:[DI]
INC DI
CMP DL, 0
JE @8
MOV AH, $02
MOV DL, 8
INT $21
MOV AH, $02
MOV DL, 32
INT $21
MOV AH, $02
MOV DL, 8
INT $21
JMP @7
@8:
MOV AL, False
JMP @Done
@AnsiTestStr:
DB 27,'[6n',0
@Done:
MOV AnsiFound, AL
end;
/Sorry, I was never one for doing a lot of code comments. And yes, it
is far from efficient, elegant or pretty. But, the snippet of code is
roughly 30 years old./
/
/
:-)
As you can see, it sends an escape sequence to request information. If
it receives the data, the test passes. Otherwise, it cleans up the
displayed garbage and moves along without ANSI support.
I wouldn’t call it a serious compatibility issue. Even this ancient
directory lister provides support to bypass the probe and force ANSI
support OFF or ON.
dir l:
Interestingly, this same program reveals a couple other cross DOS
compatibility issues as well. For example, under MSDOS using MSCDEX +
OAKCDROM driver, the CD can be listed. It also likes NWCDEX + OAKCDROM
under DR-DOS. However, not under FreeDOS using our supplied drivers
(invalid drive reported) and crashes (divide by zero) when accessing
the CD under PC-DOS with IBMIDECD + MSCDEX.
Anyhow, the program source is available on GitHub under
https://github.com/shidel/DustyTP7/tree/master/D and a precompiled
binary at https://github.com/shidel/DustyTP7/blob/master/bin/D.EXE
:-)
Jerome
_______________________________________________
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel
_______________________________________________
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel
