Geometry Wars Port?

[torpor]

e.g. glow

Well for now we can just use a very simple model (tri's would be best) with low poly counts, and a bit of texture.

Who has Beagleboards?

 

[Monk]

Tris it is then, and I'll hope somebody does a better job of it before I manage to. Textures are quite unlikely, from me at least, but I'll try to ensure I produce some workmanlike UV Mapping so that textures are practical from the get-go with anything I manage to produce.

I'm afraid all I own are PCs, a GP2X (F-200) and a Pandora (but the latter hasn't arrived just yet).

 

[bzar]

I agree that low-poly models is the way to go. That way you can have a whole lot of them . With enough effects you don't even notice it.

 

[Jan-Nik]

yes you do, but who needs shaders for a game like this?

e.g. glow

Ooh, glow:

Won't work that easily with 3d models. Shaders also provide more options (e.g. flashing glow, customizable glow, ...).

 

[zhasha]

Ooh, glow:

Won't work that easily with 3d models. Shaders also provide more options (e.g. flashing glow, customizable glow, ...).

It'll work like a charm with a 3D engine. I see 2 triangles and a texture with an 8bit alpha channel. That or you could draw it as series of vectors. You can easily draw 2D with a 3D engine designed for it. In fact, it's the preferred way to go. No new GPU has a 2D pipe or fixed functionality for that matter. It's all 3D primitives and shaders baybeh.

 

[Jan-Nik]

Ooh, glow:

Won't work that easily with 3d models. Shaders also provide more options (e.g. flashing glow, customizable glow, ...).

It'll work like a charm with a 3D engine. I see 2 triangles and a texture with an 8bit alpha channel. That or you could draw it as series of vectors. You can easily draw 2D with a 3D engine designed for it. In fact, it's the preferred way to go. No new GPU has a 2D pipe or fixed functionality for that matter. It's all 3D primitives and shaders baybeh.

I was talking about 3D models. As i understood it, he wanted to make the game with 3D models but with a top down view. Like you can choose in GTA 3.

Shaders will also be faster when rendering a lot of objects.

 

[torpor]

Well until we can actually run stuff on our Pandoras, any planning for games must be at the least a bit conservative about feature use. Sure, it'd be nice to know what sort of shaders we can make, but for a simple overhead-view 2.5D game that is intended to be a smashing shooter, we gotta be simple. For my needs, a GW clone need only consist of textures on 2 triangles-per-character .. for now. Unless someone comes up with a nicely detailed and well laid out 3D mockup, maybe I'll just use what objects I've got and work on setting up a scene that represents the gameplay, a little ..

EDIT: what we really need is for someone with the inclination and desire to sit down and do a bit of artwork. We need an artist. I think thats the most difficult part of all of this, really .. the gameplay itself is relatively banal to implement, but without a nice coherent set of art assets, it will still be crap looking, even if the gameplay is refined...

So, any artists here who can grok the concept and whip something together in blender? inkscape would be good too, in fact (SVG->GL ES model is another great idea..)

 

[Monk]

I'm trying, but I have several issues which will make my progress much slower than we may like. I'd rather not be doing it ALL but will make progress if we can confirm that the way I'm going is working well enough.

For example, I could further lower the poly count on this a bit, but not much.

 

[Yoyobuae]

Ooh, glow:

Won't work that easily with 3d models. Shaders also provide more options (e.g. flashing glow, customizable glow, ...).

It'll work like a charm with a 3D engine. I see 2 triangles and a texture with an 8bit alpha channel. That or you could draw it as series of vectors. You can easily draw 2D with a 3D engine designed for it. In fact, it's the preferred way to go. No new GPU has a 2D pipe or fixed functionality for that matter. It's all 3D primitives and shaders baybeh.

I was talking about 3D models. As i understood it, he wanted to make the game with 3D models but with a top down view. Like you can choose in GTA 3.

Shaders will also be faster when rendering a lot of objects.

What I mean is that a single image with alpha as texture is enough for a pure GW clone. Then the same sprites could be used. Plus with 3D hardware doing the rotations/morphing/etc it would be very fast (unlike a pure SDL version, as the main CPU is kinda limited).

IMO, just getting a GW port working would be a great start. Then taking that as a base to improve upon, real 3D or whatever could be introduced.

If you want artwork for a game, just get something functional and fun as quickly as possible. A working game attracts potential contributors better. (something I read from the art director at the Wesnoth project).

 

[dflemstr]

What I mean is that a single image with alpha as texture is enough for a pure GW clone. Then the same sprites could be used. Plus with 3D hardware doing the rotations/morphing/etc it would be very fast (unlike a pure SDL version, as the main CPU is kinda limited).

IMO, just getting a GW port working would be a great start. Then taking that as a base to improve upon, real 3D or whatever could be introduced.

If you want artwork for a game, just get something functional and fun as quickly as possible. A working game attracts potential contributors better. (something I read from the art director at the Wesnoth project).

So why make life difficult? Why use textures and quads when you can use shaders and simple lines? Textures require sophisticated loaders, a manager, a binding schedule, and they have aliasing artifacts and so on that you don't get by using pure vertices. So why bother?

 

[Monk]

IMO, just getting a GW port working would be a great start. Then taking that as a base to improve upon, real 3D or whatever could be introduced.

Agreed here, although we don't have anyone (as yet) offering to do an exact GW port - we have an offer of the use of a 3D engine along with, I assume, coding the game itself to use the engine.

If you want artwork for a game, just get something functional and fun as quickly as possible. A working game attracts potential contributors better. (something I read from the art director at the Wesnoth project).

From an artists perspective, I cannot agree with this more. It's hard to get too excited about putting work into a project when all one can see is the amount of potential work and none of where it's going. Even just seeing something where every "character" is the same model, but whizzing around and seeing some collision detection, would be greater motivation. IMHO

I'm going to start a thread in dev corner to continue this once I've got a couple more things together - if no-one beats me to it.

 

[Yoyobuae]

What I mean is that a single image with alpha as texture is enough for a pure GW clone. Then the same sprites could be used. Plus with 3D hardware doing the rotations/morphing/etc it would be very fast (unlike a pure SDL version, as the main CPU is kinda limited).

IMO, just getting a GW port working would be a great start. Then taking that as a base to improve upon, real 3D or whatever could be introduced.

If you want artwork for a game, just get something functional and fun as quickly as possible. A working game attracts potential contributors better. (something I read from the art director at the Wesnoth project).

So why make life difficult? Why use textures and quads when you can use shaders and simple lines? Textures require sophisticated loaders, a manager, a binding schedule, and they have aliasing artifacts and so on that you don't get by using pure vertices. So why bother?

If lines alone are used, then I guess it would be simpler. But I didn't get the impression that something like that was the plan.

 

[Monk]

So why make life difficult? Why use textures and quads when you can use shaders and simple lines? Textures require sophisticated loaders, a manager, a binding schedule, and they have aliasing artifacts and so on that you don't get by using pure vertices. So why bother?

Because, I gather, torpor is the only one offering to write a Gridwars-alike - and he's offering to do it using the 3d engine he's written (is writing?). Which does those things/needs to do those things anyway. This would also leave the way for a customizeable game, where the end user could (say) redesign all the 3D models in the game very easily without having to dig around in code or recompile and so on (although it seems Grid Wars is fairly customizable too - there's even a Millenium Falcon theme, I just discovered today!). Which I gather has some appeal to people who like 3D games (I like quite a few of them myself).

Oh - and quads have, sadly, been thrown out for the moment. Tris.

 

[Monk]

OK - I've created:

http://www.gp32x.de/board/index.php?/topic/50434-gridwars-inspired-game/

in the hopes that a productive thread on this topic can thrive there and perhaps be more likely to attract more useful aid than myself.

 

[torpor]

Great, lets move to that thread and maybe lock this one, since we're now in development territory ..