[STJFPOST]Love3D

[qaisjp]

STJFPOST = "STUPID TOPIC JUST FOR PONDERING OVER STUPID THINGS" (meaning, please dont flame or rage, just give your thought)

What would you think about Love3D? What features would you expect? Would you stay in Love? Would you continue with Love3D or stay to old builds (love2d)? Would you prefer 2D to be retained (2D and 3D framework, both are equally worked on) OR just 3D?

... im going to regret this post.

[Inny]

I would think the problem with moving up to 3d is that there are far too many ways to do 3d for a simple API to be made, without making major assumptions about what kind of 3d to support.

For instance: in 2d, you have basically two ways to do it. Draw pixels onto the screen, or give opengl the textures and tell it to put them on the screen. And where you pull those textures from, well, jpg and png are more or less it.

3d, things get a little complicated, assuming you stick with opengl, because now you have to deal with projection matrices. Even doing basic 2d things with these matrices is a pain in the ass. And to make 3d usable, a certain degree of polygon removal has to be done in the language closer to the metal (C++), which means assumptions would have to be made in love3d about what kind of data structures to use. I.E. do we use octrees or other bsps? Finally, we'd need 3d analogous formats. Do we use the formats that Maya stamps out, or Blender?

I personally think that, if 3d is what you want, jumping 20 years of progress to the front of the line might be too much cognitive overload to jam into the existing love2d. The first 3d engines were Ray Casters, so, whip out a love.graphics.newCanvas and see what kind of raycasters you can pull off.

[tv_user]

When I was a PC (and Windows) user, I played with a library called Truevision3D (VB.NET) for a while. It was pretty fun, but 3D apps require a lot more effort. Not just because of the math, but mainly because of the resources. It's not as easy to find a good 3d model, textures, etc, as it is to find a good sprite (well, at least for free), and its no joke trying to build your own models (Blender, Milkshape, Maya, 3d studio max...omg just thinking about it still makes me shiver). And because I also know a little bit about "photoshoping-my-way-out-of-misery", I like 2d game development much more, so I started fiddling with IndieLib (C++) for sometime, which was also a 2d framework, and finally, after going Mac, I found the LOVE of my life
IMO, 2D game logic and potential is just as great as 3D games'.
If LOVE went 3d I would probably keep using it, but solely for 2d.
What do you guys think?

[Nixola]

I'd like a bit of 3d: distorting images in a non-parallelogram form
Right, Jasoco?

[tv_user]

I'd like a bit of 3d: distorting images in a non-parallelogram form
Right, Jasoco?

oh, you mean 2.5D?

[Nixola]

I don't know, Jasoco's the expert here

[substitute541]

viewtopic.php?f=5&t=10777

Apparently I am working on that now... We also have the thread of the same name (Mine is, The Love3D Project...)

[ivan]

By the way, there are already a few OpenGL/Lua bindings out there.

[kikito]

The math required for even basic 3d stuff, like aligning and texturing, gives me headaches.

The only way I can see a love3d thingie with a similar level of complexity to our current 2d system is by abstracting away like crazy. I'm talking about no using polygons, but meshes, and maybe levels, instead. But then it would not be a framework any more; it would be more similar to a game engine than to a framework (which can be good or bad).

And even so, that's the programming part only. Content production would be still 2 orders of magnitude more difficult (more expensive hardware, more complex tools).

[coffee]

I don't know, Jasoco's the expert here

I would say Jasoco's is the expert in 2.5d and fake 3d. Retrotails have also some touchable work in that domain. Patalo It's perhaps more connoisseur of a more really oriented 3D field.