Updated Alien Breed 3D 2

[Karlos]

For those that aren't aware of it, pipper on EAB has gotten the AB3D2 source code into a buildable state. I remember trying this over 20 years ago and remember it being a real pain in the rear.

Please see the original thread on EAB here https://eab.abime.net/showthread.php?t=111090

[Karlos]

maybe one day 1x1 AB3D

I think you are out of luck there. AB3D renders 12-bit RGB directly, relying on the peculiarities of it's chunky copper screen. Only an RTG version using an RGB display mode is going to reproduce it properly at higher resolution.

[Karlos]

@BigD

Nope. If you want 1x1 AB3D, and you want it to look correct, you need to use RGB. The original game doesn't use C2P and uses 12-bit RGB. Your closest match is 15-bit RGB on a graphics card. If you can also generate the extra 1 bit of precision per channel, it would look pretty nice.
 

[Karlos]

@BigD

Perhaps because not everyone that wishes for an improved AB3D experience has, or wants, an NG system.

[Karlos]

My interest in AB3D2 is on the RTG side actually. One thing I am keen to investigate is whether or not it can be adapted to render directly to an RGB display without the reduction to 256 step. Given all the shading and lighting effects the game uses, it doesn't seem an unreasonable objective.

[Karlos]

@utri007

Is it possible to compile it to OS4 with current sources?

No. Moreover, if you want to get really technical it's not possible to compile it, period. It's written entirely in 68020 integer mode assembler, so there's no compilation to be done (having said that, pipper's build process uses the linker from a GCC cross compiler).

The way to make it NG friendly is to make it OS friendly and remove direct dependency on AGA, Paula, etc. then let emulation do the rest.

[Karlos]

Blasting like it's 1996!

https://www.youtube.com/watch?v=1qoA2IykmBc

[Karlos]

Drifting floor texture bug fixed.

Playthrough of modded level B https://youtu.be/JH6HYQ49604

[Karlos]

Update pushed, now includes the latest build of the engine binary from pipper's repo, so all you need to do is download the mod and you're good to go.

[Karlos]

Brings back many happy memories

Thanks for the update.

I've often felt that game got a bum rap from history. On release in 1996 it was clear to me that the engine was the best of all the "doom clones" of the day but had hardware requirements to match. It was playable for me, on a 68040 at 2/3 size, but not fullscreen. I could only wonder how it fared on the 68030 boards of the day. The game itself had issues. It felt rushed and unfinished. It didn't live up to the level design of its predecessor. All of these different factors combined to take the shine off what was a remarkable technical achievement, especially given that most of the software was written in assembler by one person.

[Karlos]

When I download from Git, I get 50mb files with lots of files.

Where is actual game?

Which repo are you downloading? It sounds like you may be downloading the engine repo.

This is now the only repo you need to play the mod: https://github.com/0xABADCAFE/karlos-tkg

The game is located in the Game directory. All you need to do is double click the KarlosTKG icon.

All the editing assets and original editors are in the Editors directory. Feel free to ignore those. This is an active project (after a 25 year break), having these in GitHub is extremely handy.

[Karlos]

@utri007

Does it guru during initial load up of the application or does it happen randomly during play?

There is a bug in the engine already identified that appears to be an illegal memory access relating to opening the screen. This is totally reproducible in UAE where the binary won't start without allocating "32-bit chip ram" (a new class of chip ram not related to the original 2-8MB setting, lives in a different address range). Allocating that memory likely prevents a bus error from the access.

[Karlos]

@Karlos

It does guru before menus.

Ok, I think this is the same issue as UAE then.

[Karlos]

You could try SystemPatch: https://eab.abime.net/showthread.php?p=1562755#post1562755a

Other than the processor cards, what else is different between the two machines?

Could you include a showconfig output for each one?

[Karlos]

Thanks. The showconfig output supports the hypothesis that a memory access in a particular address space is the culprit