Re: [stella] More flicker, please

Subject: Re: [stella] More flicker, please
From: Manuel Polik <cybergoth@xxxxxxxx>
Date: Mon, 07 Oct 2002 10:44:43 +0200
Hi Thomas!

>Manuel wrote:
>>No, reflections can be handled. The main problem I see 
>>with ANDing and most of the other solutions you 
>>mentioned (set aside the cycles I don't have for the 
>>moment) is that it'd reduce the the visible space for 
>>another 16 pixels on both sides. So it'd be reduced from 
>>128 to 96 Pixels.

>No, see below.

I try to, but I don't get it :-)

>>>- Or use the trick with all HMOVE-blanks on the left 
>>>side. They would mask 8 bits. 
>>But not 16.

>Yes, but it might save you some ANDs. And you could disable 
>completely invisible object halves by making that player black. 
>Then 8 pixel should be enough.

Ah, here is where you don't know the whole problem.

The stuff my engine can display is allowed to have 128 different
horizontal positions.

It's a miracle that this is working at all and that I managed to get
these centered.

Expanding that system to both sides for another 15 Pixel would certainly
break the repositioning code. Plus it'd complicate other game / kernel
inside logic a lot, because all is set up, planned and prepared for 128
Pixels width, even the radar.

But, I was wrong on another thing, it's not reduced to 96 Pixel width,
but 96+16, since I forgot to add the ships width.

I might give this a try...

>I meant a *visible* window border that masks the objects and 
>doesn't reduce the size at all.

Oops, sorry.  I got you wrong here. But still it has to reduce the size
for 15 Pixels on each side. Or am I not getting yet another thing? We're
still just ANDing, not RORing, right? :-)

>>I can reinstal the code I had so far and you can give it 
>>a try, if you really like to do that. I didn't get it, 
>>neither with try & error nor with thinking about it.

>If you comment the important parts... ;-)

Ok, I'll do my best :-)

>>Hm... it requires 2 extra lines for the kernel to 
>>determine wether the next ship is ready for 
>>repositioning. I assume these are part of the "blind 
>>spot" problem I mentioned. But, all I could do was 
>>shifting it :´-(

>I still see no technical reason, why you can't show objects 
>that startt before the top. You can setup all variables outside 
>the kernel. 
>Where is the problem? Has it something to do with the sorting 

No, that works fine. Just I don't get a smooth transition done between
the point where still the half sprite is done and where finally the
kernel kicks in. Any object starting at a certain range inbetween these
two points get's invisible. So there's a blind spot for any ship of any
size. That's what I meant, I hope this time I'm making myself more

Archives (includes files) at
Unsub & more at

Current Thread