(09-26-2022, 04:46 AM)mnrvovrfc Wrote: I made a quick look at "libqb.cpp", toward the bottom, at the "big fat main loop". It doesn't look like "_EXIT" is being handled by ancient interrupts. What you're requesting might have to be programmed like that for "_EXIT" but check for a different keystroke combination. Why do you want to trap "Print Screen" key? Oh OK you must actually have a printer and it's annoying to press it. Or it's annoying that Windows interferes when it guesses the user must have wanted to print a text file via Notepad.

You have to remember that any trace of "CALL INTERRUPT" is emulated by QB64PE only with the first three or so mouse functions, because those calls were quite common in programs written for QBasic, especially those by "TheBob". Cannot emulate BIOS or MS-DOS like we could do in M$QB or QBasic many years ago. Therefore, any emulation of ancient interrupts have to be implemented. Oh no somebody is going to request the "BSOD" one or the disk sector one... just ignore them devs.

The same thing goes on with the ancient input and output ports. Because so many QB/QBasic programs played around with EGA/VGA bit planes, the ports were emulated at that end. One more thing to emulate might have had to do with the keyboard. Otherwise Windows Vista and later was going to make it a pain installing a DLL only to be able to use "INP()" and "OUT" and "WAIT" in a QB program away from 16-bit.

(09-26-2022, 07:13 PM)Spriggsy Wrote: I'm very confused as to why you are bringing up mouse interrupts and such.

(10-10-2022, 03:42 AM)eoredson Wrote: The assembly code I am using to trap x1b and x05 is this:

Code: (Select All)

;Traps and ignores Print-Screen and Control-Break. v1.1a PD 2020.

;Compiled with 'Tasm /t Keytrap;'

;Compile calling program using /fs with BC. (QBX forces /fs).

            .model  medium, basic          ;Stay compatible with basic.
    extrn  SETUEVENT: far
            .code
public      SETINT

SETINT      proc    uses ds
            mov    ax, 351BH              ;Get old control-break interrupt
            int    21h                    ;vector and save it.
            mov    word ptr cs:OldVector1,bx
            mov    word ptr cs:OldVector1+2,es
            push    cs                      ;Set the new
            pop    ds                      ;interrupt vector
            lea    dx, EventHandler        ;to the address
            mov    ax, 251BH              ;of our service
            int    21H                    ;routine.

            mov    ax, 3505H              ;Get old print-screen interrupt
            int    21h                    ;vector and save it.
            mov    word ptr cs:OldVector2,bx
            mov    word ptr cs:OldVector2+2,es
            push    cs                      ;Set the new
            pop    ds                      ;interrupt vector
            lea    dx, EventHandler        ;to the address
            mov    ax, 2505H              ;of our service
            int    21H                    ;routine.
            ret
SETINT     endp

public  EVENTHANDLER

EVENTHANDLER    proc
            iret                            ;eat keystroke. pop stack.

OldVector1  dd      0                      ;keep data in code segment.
OldVector2  dd      0

EVENTHANDLER endp

public      RESTINT

RESTINT    proc    uses ds                ;Restore the old
            lds    dx, cs:OldVector1      ;interrupt vectors
            mov    ax, 251BH              ;so things will
            int    21h                    ;keep working when
            lds    dx, cs:OldVector2      ;this basic program is
            mov    ax, 2505H              ;finished.
            int    21h
            ret
RESTINT    endp
            end

(10-11-2022, 04:29 AM)eoredson Wrote: My reasoning is thus:

If inline assembly could be injected into 'C' and inline 'C' could be injected (such as Mem.h) into QB64, then why not inject assembly into QB64!?
  (I know it doesn't work like that but it is a passing idea)..