Powercore [WIP]

[Silvio Mogno]

Wow, this will be one of best homebrews ever!

[Seth A]

This looks amazing.

[LatchKeyKid]

Looks good!  Out of curiosity, how often are you scrolling the playfield?  Every fourth frame? 

[Thomas Jentzsch]

3 hours ago, LatchKeyKid said:

Out of curiosity, how often are you scrolling the playfield?  Every fourth frame? 

There is no playfield used.

 

Try ALT+. in Stella to see what is happening. 

[LatchKeyKid]

6 hours ago, Thomas Jentzsch said:

There is no playfield used.

 

Try ALT+. in Stella to see what is happening. 

 

Oh, wow.... that's even more impressive then!  I didn't switch to view the TIA object and am shocked!  

[Lazycow]

Update: Each stripe can have its own color scheme now. An the floors can move independently.

@Silvio Mogno Yes, it looks fine so far. But it's still a prototype, it's a lot of work to make a finished game out of this.

@LatchKeyKid Yes, no playfield registers are used. There're 20 independent subroutines and each of them is changing the background color at another horizontal position. To camouflage this, the ball sprite is set on top of this.

 

Any hints how moving floors could be used in the game?

[Bomberman94]

50 minutes ago, Lazycow said:

Any hints how moving floors could be used in the game?

At least as a bonus game or other part of the game: the typical „get at the top of the screen but don’t fall and die“. You start a the bottom and screen moves down while you must reach the top. 

[Thomas Jentzsch]

On 9/13/2024 at 12:39 PM, Lazycow said:

@LatchKeyKid Yes, no playfield registers are used. There're 20 independent subroutines and each of them is changing the background color at another horizontal position. 

Have you thought about less subroutines which are running within some cycle ranges?

On 9/13/2024 at 12:39 PM, Lazycow said:

Any hints how moving floors could be used in the game?

If they move at different speeds, they could form a changing pattern. And your goal would be to climb up without being pushed from the current platform from the platform above.

Edited September 16 by Thomas Jentzsch

[Greg2600]

On 9/13/2024 at 6:39 AM, Lazycow said:

Amazing even in this stage.

[Lazycow]

On 9/16/2024 at 6:37 PM, Thomas Jentzsch said:

Have you thought about less subroutines which are running within some cycle ranges?

Not until now... But how? I came up with this:

ldx delayValue ;(3)
lda $10FF,x    ;(4+)

...with delayValue=0, it will take 7 cycles, and with delayValue<>0, it should take 8 cycles.

But I don't have 7/8 cycles available in the kernel(s) right now. I couldn't find a way to use some existing opcode for a variable delay, to speed it up.

 

About the leveldesign with moving floors: I like the idea of repeating floor moving patterns. But including a challenge is difficult: You can be pushed down, but falling isn't harmful. A time limit would work, though, yes! Maybe not with a counter, but maybe a bouncing shot that's moving up or a robot that is following you. Cool! Anything else?

[Thomas Jentzsch]

21 minutes ago, Lazycow said:

Not until now... But how? I came up with this:

ldx delayValue ;(3)
lda $10FF,x    ;(4+)

...with delayValue=0, it will take 7 cycles, and with delayValue<>0, it should take 8 cycles.

But I don't have 7/8 cycles available in the kernel(s) right now. I couldn't find a way to use some existing opcode for a variable delay, to speed it up.

I am using indirect jumps. These are jumping to NOP #imm ($80) sequences (with a final standard NOP). That way you can delay in multiple, single cycle steps.

 

And my kernels are not using WSYNC. So I only have to setup (using a table) and execute this jump once per block (floor).

21 minutes ago, Lazycow said:

About the leveldesign with moving floors: I like the idea of repeating floor moving patterns. But including a challenge is difficult: You can be pushed down, but falling isn't harmful. A time limit would work, though, yes! Maybe not with a counter, but maybe a bouncing shot that's moving up or a robot that is following you. Cool! Anything else?

Maybe the floor could slowly scroll down. With some water, fire, spikes... below. That would act as a timer replacement. But maybe then the game would become to hard, depending on how much you can fall down.

[ZippyOasys]

Wow! Is this done with batariBasic?

[Thomas Jentzsch]

8 hours ago, ZippyOasys said:

Wow! Is this done with batariBasic?

Of course not!  

 

At least the kernel is hand crafted.

Edited September 22 by Thomas Jentzsch

[Lazycow]

Yes, the prototype is written in assembler. (assembler only)

 

Yes, jmp (x) is faster and more flexible, but I cannot use it, because the preparation of the jump vector would be too much overhead.

But your jmp trick inspired me, I came up with something else:

lda delayValue ;(3)
bne .continue  ;(2+)
.continue:

A 5/6 cycle delay might fit in some of the 20 kernel subroutines.

 

And I made a test, using a bonus level like timer seems to be challenging enough. I might use a bouncing shot instead of a timer.