IntyBASIC Blast Junk to Screen?

[First Spear]

Is there a way to write random contents of ROM or RAM to the screen, so it shows "random junk"? I want to make the display look as it does when frying Star Strike (push reset and turn the game off and on at the same time rapidly and it shows a mess).

 

Thanks!

[mr_me]

The last 43 locations of grom is exec program/data. As graphics it should look like junk. What does gram look like before you write anything to it?

[intvnut]

If you just want random characters in BACKTAB, you could copy a portion of the EXEC to BACKTAB with a for loop.

 

e.g.

.

FOR I = $200 to $2EF
  POKE I, PEEK($1000 + I)
NEXT I

.

Try different starting locations than $1000. This particular loop actually copies $1200 - $12EF to BACKTAB ($200 - $2EF).

[carlsson]

I believe it should be #I because that is a variable holding a value larger than 255. Otherwise it works fine.

 

I tried to make a hack of the SCREEN function that also seems to work, except that for dynamic use you need to know on beforehand which variable is which:

 

  #a = $1000
  ASM MVI V1,R3 : REM V1 would be the address to #a and needs to be looked up
  ASM MVII #512,R2
  ASM MVII #20,R1
  ASM MVII #12,R0
  ASM CALL CPYBLK

It probably is not much faster than a FOR loop but introduces a couple levels of complexity. SCREEN as it is takes a label to DATA statements and while I tried to figure out if VARPTR in some way can be used, I failed to find a such solution so I duplicated the ASM output for SCREEN.

[intvnut]

I believe it should be #I because that is a variable holding a value larger than 255. Otherwise it works fine.

 

Oops. You are correct. I had originally written "0 to 239" and was adding $200 inside the loop. Then I "optimized" it. Clearly I didn't try to run it.

 

 

I tried to make a hack of the SCREEN function that also seems to work, except that for dynamic use you need to know on beforehand which variable is which:

 

  #a = $1000
  ASM MVI V1,R3 : REM V1 would be the address to #a and needs to be looked up
  ASM MVII #512,R2
  ASM MVII #20,R1
  ASM MVII #12,R0
  ASM CALL CPYBLK

It probably is not much faster than a FOR loop but introduces a couple levels of complexity. SCREEN as it is takes a label to DATA statements and while I tried to figure out if VARPTR in some way can be used, I failed to find a such solution so I duplicated the ASM output for SCREEN.

 

 

IntyBASIC 1.4 uses predictable names for user variables, solving your first issue. Also, or could implement this as a CALLable assembly function instead and receive the address in a register. Replace "CALL CPYBLK" with "B CPYBLK."

 

If you really do just want to throw a chunk of EXEC code on the screen, you can do something sleazy like this:

.

ASM __cur: QSET $
ASM  ORG $1000, $1000, "-RWBN"
exec: DATA 0
ASM  ORG __cur

.

Now you have an IntyBASIC label named "exec" that points to the EXEC.

[+Lathe26]

To add to the randomness, you could also add RAND to $1000 so that the pixels are not the same each time (or maybe add 16*RAND so the ranges overlap less). Alternatively, you could XOR the RAND value with the EXEC data.

[First Spear]

Thanks I did play with RAND a bit but there was no randomness, each time the program was run there was always the same junk on the screen, not random. Also adding a single WAIT inside and outside of the loop did not make a difference. Maybe I am misusing RAND?

#offSet =  RAND * 16
FOR #I = $200 to $2EF
  POKE #I, PEEK($1000 + #offSet + #I)
NEXT #I

To add to the randomness, you could also add RAND to $1000 so that the pixels are not the same each time (or maybe add 16*RAND so the ranges overlap less). Alternatively, you could XOR the RAND value with the EXEC data.

[+nanochess]

Maybe this could help:

 

	'
	' Generate a trash screen
	'
	' by Oscar Toledo G.
	'
	' Creation date: May/10/2019.
	'

	CLS
	MODE 0,0,0,0,0

	WHILE 1

		WAIT

		#d = RANDOM(256) * 8
		#e = RANDOM(256) * 4 + $1000

		FOR c = 20 TO 119
			#backtab(c) = PEEK(#e + c) + #d
		NEXT c
		
		WAIT
		
		#d = RANDOM(256) * 8
		#e = RANDOM(256) * 4 + $1000

		FOR c = 120 to 219
			#backtab(c) = PEEK(#e + c) + #d
		NEXT c
	
	WEND

[First Spear]

This is odd. On each repeated run of that code, I get the same screen every time. I wonder if my runtime environment is making a RAND issue.

 

#d = RANDOM(256) * 8
#e = RANDOM(256) * 4 + $1000
FOR c = 20 TO 119
	#backtab(c) = PEEK(#e + c) + #d
NEXT c
WAIT
Print at 0 Color 7 , <4>#e
#d = RANDOM(256) * 8
#e = RANDOM(256) * 4 + $1000
FOR c = 120 to 219
	#backtab(c) = PEEK(#e + c) + #d
NEXT c
Print at 20 Color 7 , <4>#e
Print at 40 Color 7 , <4>#d
For #spinWait = 0 to 6610 : Wait : Next #spinWait

Maybe this could help:

	'
	' Generate a trash screen
	'
	' by Oscar Toledo G.
	'
	' Creation date: May/10/2019.
	'

	CLS
	MODE 0,0,0,0,0

	WHILE 1

		WAIT

		#d = RANDOM(256) * 8
		#e = RANDOM(256) * 4 + $1000

		FOR c = 20 TO 119
			#backtab(c) = PEEK(#e + c) + #d
		NEXT c
		
		WAIT
		
		#d = RANDOM(256) * 8
		#e = RANDOM(256) * 4 + $1000

		FOR c = 120 to 219
			#backtab(c) = PEEK(#e + c) + #d
		NEXT c
	
	WEND

 

[+nanochess]

So you only need a frame of trash?

 

jzintv makes sure everything is set to zero, so it will show always the same trash.

 

I think it has a randomized memory flag that behaves like a real Intellivision.

 

Over a real Intellivision you would get always a slightly different "trash" screen because the pseudo-random values in RAM.

[intvnut]

The --rand-mem flag on jzintv will randomize memory at startup. FWIW, I recommend folks test their games with this flag if they intend to put them on cartridges.