![]() |
Quote:
Already an A500 with OCS can handle the game in C or C++. :agree |
never use SDL in action games. It crawls.
|
@Vlad
Footman is debatably a better version of Pac-Man and it’s been on Amiga for many many years. Ms Pac-Man should be doable too. |
Had a look at the arcade sprites and put them into Dpaint - only use 12 colours for all graphics including mazes and sprites are only 3 colours as far as I can see so would be perfect for Amiga; it would seem that you could do all the moving graphics in sprites so probably any language would be good enough?
As mentioned previously, getting the ghost movement right is the biggie but if it's like the original then there are just three different ghost states and very simple AI which again would not be beyond any langauge. Can we expect a Tiny Ms Pacman next... |
I’m hopeful for a Tiny Ms. Pac-Man.
:) |
Quote:
You got me curious. The AI logic is about 1,200 lines of ASM code from my brief check earlier this morning. The C logic would easily expand to 3x-4x that (unless the compiler is really good). Perhaps just the outer layer was written in C and innards were ASM ? |
Quote:
But, for some quick prototyping and experimenting SDL is great, unless one has an established ASM codebase already... Still, experimenting with AI in ASM is lunacy. Best to try it all out quickly in higher level language and only when it works port it to ASM... |
Quote:
Quote:
Why don't you go translate the Z80 logic into something readable by everybody ? You don't even have to translate it to C, just some pseudocode that lists all FSM states, end conditions and all decision spots with exact behavior description (if.then.else). If you do that, I promise I will rewrite it in my Higgs language in one day, ok :great ? |
This is quite interesting - you don't necessarily need to dissemble 1200 lines of code?
|
Quote:
Go to https://github.com/BleuLlama/GameDoc...mble/mspac.asm And the first portion of AI starts at Line 5503 and continues for roughly ~700 lines. I suspect the identical behavior is repeated for all 4 colors (red,orange,pink,inky) so in theory once you document behavior of one color, it should be the same for others. I certainly don't recall different behavior per enemy color, but it's been few decades since I played it... Second portion starts at 6353 (ghost house release). I forgot where is the third part now, but should be pretty obvious from the comments, if you scroll a bit further down. |
|
Quote:
Quote:
|
Quote:
Even during combat, if some more complex logic caused overflow to next frame and fps dropped (and is it even VSYNC'ed in the first place?), you wouldn't be able to notice it. If the fps halved from 60 to 30 (or 30 to 15, whichever it is) in pac-man, you sure would notice. But, clearly, this proves that Amiga must have had some amazing C compilers back in the day! Now contrast that with Jaguar :laughing Quote:
I might have noticed that quarter century ago, just didn't remember that today. |
Quote:
|
Quote:
Great video! |
@Havie
Thanks for the video! Can anyone use this to programme the behaviour in an Amiga port? I could give it a go in Blitz Basic. Seems a bit of a challenge. |
All times are GMT +2. The time now is 16:55. |
Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.