Re: Idea - Distorting Sprites
Posted: Thu Apr 02, 2009 7:45 am
@hdon: I think that your "orthogonal API" thesis misses part of the point. As I see it, a solid, convincing reason not to expose OpenGL (besides some of the others posted here) is that it forces people using love to build in 2d. And this means that love is not a "3D game development platform," but rather an explicitly "2d game development platform." Honestly, if love hadn't been an explicitly 2d platform (as I understood it before downloading it), I probably wouldn't have picked it up. Don't get me wrong, I love messing around with OpenGL, etc., and I've done some reasonably difficult graphics coding. But I picked up love because it promised to be *simple*. And if you add 3D functionality, it can no longer make that promise. Suddenly you'd have a bunch of forum threads about "how do I do this or that in OpenGL?," and, as people have pointed out, love becomes just another mediocre 3D engine, as more and more resources get devoted to developing 3D features. Now, I understand that I'm exaggerating a bit, making a lot of assumptions, and running a slippery-slope argument. But I think that even taking those things into account, the point is valid. Love as an explicitly 2D engine is worth something.
Another reason not to go 3D (and potentially a reason for not implementing distortions): I've seen at least one place where the developers mentioned that they'd like to do some more optimization. It seems pretty obvious to me that there is almost certainly *huge* potential for optimization if everything is forced to be 2D: even within OpenGL, there are probably tricks available if you don't use the z coordinate that aren't if you do. This ties in with portability, which others have mentioned: love is *much* easier to port to a new system if you don't have to port a 3D API.
Those things said, I'm not sure how others feel about this, but in my opinion, there's nothing stoping you from forking the project. I don't think that Love 3D would be a bad thing (as long as it had a distinct community, website, etc. so that people interested in 2D stuff could focus on their own issues). But, for the reasons above, I don't see it as being a huge hit.
As for whether to implement distortions: I'm ambivalent. I agree that they would be cool, but as I just pointed out, if they're implemented using the 3D-ness of OpenGL (and you'd be stupid not to do so), they'll make the project as a whole more dependent on a 3D rendering system, and thus more difficult to port, and possibly more difficult to optimize.
Another reason not to go 3D (and potentially a reason for not implementing distortions): I've seen at least one place where the developers mentioned that they'd like to do some more optimization. It seems pretty obvious to me that there is almost certainly *huge* potential for optimization if everything is forced to be 2D: even within OpenGL, there are probably tricks available if you don't use the z coordinate that aren't if you do. This ties in with portability, which others have mentioned: love is *much* easier to port to a new system if you don't have to port a 3D API.
Those things said, I'm not sure how others feel about this, but in my opinion, there's nothing stoping you from forking the project. I don't think that Love 3D would be a bad thing (as long as it had a distinct community, website, etc. so that people interested in 2D stuff could focus on their own issues). But, for the reasons above, I don't see it as being a huge hit.
As for whether to implement distortions: I'm ambivalent. I agree that they would be cool, but as I just pointed out, if they're implemented using the 3D-ness of OpenGL (and you'd be stupid not to do so), they'll make the project as a whole more dependent on a 3D rendering system, and thus more difficult to port, and possibly more difficult to optimize.