Re: [stella] RPG progress

Subject: Re: [stella] RPG progress
From: "Paul Slocum" <paul-stella@xxxxxxxxxxxxxx>
Date: Mon, 23 Feb 2004 11:18:48 -0600

So maybe you can reduce the time to two frames, right?

I originally did it in two frames using some of the visible screen for processing time, but I decided it wasn't worth wasting screen space. And adding a few more features to the tile system made it take a bit longer, requiring even more visible space. You could still see the tile generation delay with two frames, it just wasn't as pronounced.

I really think the best solution would be to set aside 5 more bytes and generate it offscreen. I was originally planning on having the monsters displayed on the map. Since I'm not doing that now, I may actually be able to spare those 5 bytes.

>Or you could just use *two* large swap tables and make the code
>completely simple.

That's a lot of space. I'll think about it. Another option would be to continue drawing the playfield graphics while doing the calculations between tiles. This would have similar visual results to closing the vertical gap between rows.

I am still far from understanding, but couldn't you use the cmdBuffer?
Isn't that solely used for filling the new row? And while the new row
is filled the content of the cmdBuffer is getting more and more

The cmdBuffer area is destroyed during the kernal (it is used for pointers to tiles while drawing the screen).

Makes sense. But where exactly are you using the advantages of BCD?

It's used for some of the calculations that determine the location of tiles in the 10x10 screen window, and it's used in the random tree/rock generation section.


Archives (includes files) at
Unsub & more at

Current Thread