Compileable version of Alien Breed 3D II - The Killing Grounds

[Karlos]

Originally Posted by Dunny

I think you're probably right tbh. Hey, would a true-colour RTG version be on the cards at some point? That might be a fun thing for PiStorm owners to experience.

Yes, it's on my hitlist since RTG was introduced. Some if it should be quite easy. For example, the game relies on taking a palette colour and looking up the brightness variation of that colour in a big shade table that currently maps the brightness variation of the colour to the nearest palette entry. That can just be replaced with a true colour lookup table containing the actual true colour variation of the brightness and written straight to the buffer.

That works for walls, floors and basic bitmap sprites. The only real challenge is transparency. Basic glare and shadow transparency works really well in 8-bit because all you have to do is to read the current frame pixel (which is a palette value) and look up the required shade variation again in the same big table. That isn't going to work in RGB mode. You have to do it properly. Read the RGB value from the framebuffer and adjust it accordingly. The other form of transparency is additive (plasma bolts and explosions). That may or may not be easier - it's not totally obvious how it works currently but as it only seems to work if you stick to the original asset palettes I think there's another table hidden away somewhere.

[Karlos]

I've updated the branch to incorporate Kalm's 030 optimised 1x1 C2P routines.

When using the generic builds (68020+), the runtime CPU detection determines if the 040 or 030 routines are appropriate.

For the 040 and 060 targeted builds, the 040 routines are always used and the 030 ones are not included in the executable.

[abu_the_monkey]

@Karlos

making great progress there, cool stuff indeed

sadly I will not be able to check this out (too many real life issues going on at the mo) hope to get back on it soon.

[Karlos]

There's still plenty to do there. We're only using kalm's 030 routines for fullscreen C2P on 020/030, but I can't see anything that precludes them from the 2/3 view (which uses 040 code), with the possible exception of alignment on screen.

[Karlos]

Originally Posted by abu_the_monkey

sadly I will not be able to check this out (too many real life issues going on at the mo) hope to get back on it soon.

Real life always comes first, sir. I know what that can like only too well.

[Karlos]

Anyone on AGA/68030 that can test the current "genetic" builds?

[vroom6sri]

Originally Posted by Karlos

Anyone on AGA/68030 that can test the current "genetic" builds?

I can try and have a look this evening but I am a bit limited with my tech skills. Please can you spell out what you want me to try and share the ready-to-go files and folders so I can just copy it over and try it?

I can test it on my A1200 ACA1230/56.

[Karlos]

Originally Posted by vroom6sri

I can try and have a look this evening but I am a bit limited with my tech skills. Please can you spell out what you want me to try and share the ready-to-go files and folders so I can just copy it over and try it?

I can test it on my A1200 ACA1230/56.

Do you have the original game?

The attached LHA file contains the two binaries and a RunAB3D2 launch script that sets up everything you need. Extract this archive using UnArc or your favourite extractor into the directory where your existing game is installed. The new binaries should be in a subdirectory called "bin" after this.

Double click the RunAB3D2 icon from workbench and it should present you with a file requester to select the engine from the bin directory to test. The one called "tkg" is the pure assembler version that has no RTG support, and the one called "tkgc" is the one that can run on RTG systems.

To test the latter, all you have to do is make sure you choose the PAL LoRes 320x256 screenmode from the mode requester when it opens.

All I am looking for at the moment is general stability and if possible any indication as to whether or not it's faster than before, especially in fullscreen mode.

The C2P routines used for 030 have changed from the ones optimised for 040/060 to the one optimised for 030. The guidance for the 040/060 routines indicate they do not work as well on 030.