I have a question.  I'm the horrible flavor of coder who doesn't fare well with manipulating the stack perfectly, so in case I run into difficulties with some code, I tried writing two small routines that would get the location of the stack (with an extra pop beforehand) and the value at the top of the stack before I even touch it myself, so when I do a final exit from the app, the stack looks as though it was never touched to the OS.  Here is my code for this:

;start stack check initializer
;keeps track of app start location
;followed up later in case of emergency exits

StartStackCheckInit
pop hl
ld (OP6),hl  ;load emergency resources into OP6
ld hl,0
adc hl,sp
ld (OP6+2),hl
ld hl,(OP6)
push hl
ret
StartStackCheckInit_end


;emergency exit call
;used only when an exit is REALLY REALLY REALLY needed

ExitUsingEmergencyResources
ld sp,(OP6+2)
ld hl,(OP6)
push hl
END_APP_RUN  ;just a small stand-in for that bcall(_JForceNoCMD...) thing I always forget the name frequently
ExitUsingEmergencyResources_end

Looks like it'll work (from what I saw in Wabbit, it seemed to work fine there too -- however, I can't test on hardware ATM) but is it safe to sexually abuse the stack this much?  Or is it even safe in general to attempt such a thing?

Bcall(_JForceNoCmd) already does this for you. It fixes the hardware stack, the floating point stack, and whatever else, so even if you're stack is leaking horribly you should still be able to call it and exit safely to the OS

Yeah thats why its ok to bcall it, even though it was technically meant to be a bjump.

You know me too well

Don't worry I still like him