next »

[fruitfly]

Hi there,
As i am not the best artist and more intersted in the techology, over the christmas holidays i had the idea that it would be interesting to "clone" an old-school FPS like Duke Nukem 3D or Doom. By "cloning" it and thus reusing the original assets i could focus on the technology. Maybe at the end i have learned and a lot and i have a retro-style rendering engine that i can reuse in a future project.
I am planning on streaming my efforts via twitch (maybe 2-3 hours a week depending on what time permits). On the one hand this might keep me interested in case there is positive feedback from the community; and on the other hand it might be intersting for other begining game-devs to see a "full" game come to life.
I have no illusions that this require persitence and even then it is highly unlikely that i will come even close to writing a full clone of the game. That's why i always put the word clone in quotes. The goals is more to write a 10%/20% complete "clone" that implements the core features and concepts that are essential. And even then it should not be seen as a "clone" but as "implementing a retro-style 3d game engine" without worring to much about art- and content-creation.

What i would be intersted to get some feedback about is
- Is this something there might be interest for?
- Is there a preference for a target game? I currently tend to "duke 3d" as it was a great game and there seem to be a lot of doom-clones already; the community around duke 3d just seems smaller (maybe it can even be called dead by standards compared to doom).
- I will be using Java with either libgdx or raw lwjgl. Is there more a trend to "doing everything by yourself" or "relying on a small but solid framework". libgdx does not bring very much to the table for the project but it will here and there certainly get us up to speed quicker then having to code everything from scratch.

Thanks for your feedback already in advance,
Daniel

[carterza]

All the source code for for most id tech games is now available on github - https://github.com/id-Software - including Duke 3d.

There's also http://fabiensanglard.net/ who goes into depth on a lot of the technical details of the id tech engines. Forgive me if you already knew about these.

I think it would be interesting - I don't know the details of the Duke 3d renderer, but I'm not sure how exciting implementing it will be. Wolf 3d renderer would be fun. John Carmack mentioned in a Quake Con a few years back he was going to try to rewrite it in Haskell, no idea how far he got, he steered into a talk about functional programming.

[fruitfly]

yes, i know about all that. those articles by fabian helped a lot to understand the codebase of those different games.

to your last point: wolf3d i don't consider that much fun because really doom and duke where the games to me that had the perfect balance between gameplay and visual complexity. wolf 3d is on the lower end with very simple levels and geometry. i don't think it has aged well compared to doom. and i would like to build something in the tutorial series that could still be fun to play today.

[carterza]

I just don't understand why you'd want to write a tutorial series on a very early id tech engine / game, especially if you're going to write your tutorial series in Java.

Wolf3d had an interesting renderer at least - if you're just going to write a first person shooter tutorial then why even associate it with the id game at all? Why not bump opengameart or collab with someone on here?

I just don't think you're really going to touch on the interesting parts of those game engines when you're writing your game in Java using a high level library like that. If your point is that you just want to use content from those games, well okay but I think using another artists content who is looking for work could be more beneficial.

[fruitfly]

its not about art but that the whole content of the game is available; so the focus can really be on understanding how such a game works internally without worrying about content creation at all. think game's architecture.
your other points i dont really understand; wether what the choice of java has to do with it nor why wolf3d has a more interesting renderer then titles like doom.

[carterza]

its not about art but that the whole content of the game is available; so the focus can really be on understanding how such a game works internally without worrying about content creation at all. think game's architecture.
your other points i dont really understand; wether what the choice of java has to do with it nor why wolf3d has a more interesting renderer then titles like doom.

Okay fine... You wanted feedback, I'm sorry you didn't understand the feedback I provided, good luck with your tutorial / stream.

[metagrue]

Why have I never thought to stream development before? This is actually a pretty cool idea.

Have you set up a twitch link I can follow?