Paku Paku -- 1.6 released 9 November 2011

[deathshadow60]

Should be ok, tested fine on the 1k... I'm assuming the problem was on the jr?

Is it possible to get a screencap of the issue? Double check that the copy you have isn't corrupted and you overwrote all the files.

[Trixter]

I just played on 5160 with CGA, joystick, adlib and it looked fine to me.

[Vorticon]

Yup. It turned out that my files were corrupted again... In any case, it runs just fine now. Really amazing conversion. As a matter of fact, the non-Jr version runs better than the Jr specific one, where in the latter the colors seem a bit oversaturated and the power pellets sometimes become flashing dashes.
One little bug: the high scores are not being saved, which is a pity given that I hit 139000 on my first try

[jmetal88]

I noticed something interesting - the game is in black and white when I hook up to a TV via composite out. That's only the standard version, the T1000/Jr version is in color.

[deathshadow60]

Version 1.6 is now out, and I've made room on my programming page to give it an official home.

http://www.deathshadow.com/pakuPaku

fun stuff... so in revamping the code, fine tuning EGA and VGA support, and adding MT-32/General Midi, what did I break on the Jr?

[Vorticon]

Looks great and plays great. The high score bug seems to be fixed. However, I did notice that the joystick calibration is a little off, but it is quite possible that my joystick is acting up. Fun game.

[deathshadow60]

Looks great and plays great. The high score bug seems to be fixed. However, I did notice that the joystick calibration is a little off, but it is quite possible that my joystick is acting up. Fun game.

Actually, and this is a laugh, it appears I'm not waiting long enough between joystick reads for the interface to 'normalize'... looking into that now. I might have to go back to the old joystick routines because of this -- or investigate other ways of handling it.

[Vorticon]

I replayed that game recently, and yup, the joysticks definitely are not behaving as they should. Are there plans still to fix this little issue? I can play it fine with the keyboard, but it's more arduous...

[Trixter]

I noticed that last week -- the fix is really easy, just wrap the joystick timing loops with CLI/STI. Works like a charm.

He used to have this in 1.5 but took it out for some reason in 1.6. Not sure why, since interrupts will definitely occur when you're trying to read a value from a joystick. So the old code here:

Code: Select all

function joyStick1Axis:longint; assembler;
asm
	xor   ax,ax
	xor   bx,bx
	mov   cx,$7FFF
	xor   di,di
	mov   dx,stickPort
	out   dx,al
@loop:
	shr   al,1
	adc   bx,0
	add   di,ax
	in    al,dx
	and   al,$03
	loopnz @loop
	mov   ax,bx
	mov   dx,di
end;

...should be changed to this:

Code: Select all

function joyStick1Axis:longint; assembler;
asm
	xor   ax,ax
	xor   bx,bx
	mov   cx,$7FFF
	xor   di,di
	mov   dx,stickPort
	cli
	out   dx,al
@loop:
	shr   al,1
	adc   bx,0
	add   di,ax
	in    al,dx
	and   al,$03
	loopnz @loop
	sti
	mov   ax,bx
	mov   dx,di
end;

[Vorticon]

Would you mind patching the executable and posting it here? Or if you could give me detailed instructions on how to do it, then I can take a stab at it. I assume one would have to use DEBUG on the Jr...