Coming soon, you'll hate this :-)

[Julian Skidmore]

Hi folks,

Yes, a version of FIGgybird is in progress ;-)

-cheers from Julz

--
                             
                  The DIY 8-bit computer from nichemachines™

[David Bambrough]

I wondered if anyone would do it.
It was in the back of my mind as a new project.
Be interesting to see how close it is to the version on my iPad!
I think it will be a good example of coding a scrolling game in forth.

Cheers,
Dave

--
You received this message because you are subscribed to the Google Groups "FIGnition" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fignition+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[derrick welsh]

Like

[carl]

Looks like the killer app to me. 

[Si Brindley]

I hope you are ready for the viral spread, media storm, introspective anxiety about the nature of addiction, despair at the dichotomy of the human condition, followed by pulling the source code from the FIGnition site to save future potential addicts. 

[Julian Skidmore]

Hi guys,

Viral spread - that'd be great :-)

One of the basic jobs I do for a bitmapped game is to calculate how much work the blitter needs to do. My calculations from March 2013 estimated FIGnition 2blts 12622 tiles/s, which is 252 tiles per frame. These are 2blt timings (where a bitmap is xor-blitted with its old position to erase it and then (a possibly different bitmap) is xor-blitted into a new position).

FIGgyBird will need to blit 3 columns and a bird. Each column is 2 tiles wide and 14 tiles high (if we exclude the gap in the middle) = 28 tiles. Thus 3 columns require blitting 84 tiles. The bird needs 4 tiles, taking it up to 88 tiles. The background (if/when I add it) only needs blitting at the start.

84 tiles is only 33% of the frame blitting performance of FIGnition, which means we should achieve 50fps for this game; leaving 66% of CPU for playing the game, about 200*2/3 = 133KIPs or just over 2500 instructions per frame. So how much is needed?

The column loop needs about 12 instructions for each 8 pixels of the column and there are 12 of these in total. Also, 4 instructions are needed when displaying part of a column (which is done twice). Also, .pipe needs 57 instructions. So the total will be about: 209 per pipe or 627 for 3 pipes. So, the bulk of the computation is only 25% of remaining CPU. It will be a bit more than that for the animated version of the pipe display routine, let's say 33%. So, assuming that displaying the bird and doing other calculations takes up another half of that, then we're still only using 50% of CPU.

Conclusion, a 50fps version of FIGgyBird is easily within FIGnition's capabilities :-)

-cheers from Julz