Tag Info

Hot answers tagged

5

On tile based GPUs (commonly seen in mobile hardware) it is beneficial to clear the renderer, as it allows the GPU to dispose of the existing cached frame data. Here's a quote from OpenGL Insights, dealing with the benefits of clearing the framebuffer on a tile based renderer: "On a tile-based architecture, avoiding clears can be disasterous for ...


4

Java defaults to using GDI (AWT, Swing). JavaFX supposedly will be able to make use of OpenGL in future. Java + LWJGL (an OpenGL wrapper that accesses native opengl32.dll via JNI) provides more direct hardware-accelerated support for Java. Flash Player 11 onward made use of OpenGL via Stage3D. Without using Stage3D, it is using a software renderer built to ...


4

PyGame itself, being a wrapper around SDL, is unable to directly use hardware acceleration. You can, however, use PyGame to create an OpenGL context and then render into that via an OpenGL library like pyglet or PyOpenGL, which will let you then also use GLSL vertex and fragment shaders. This blog post does a good job of explaining the basics of using ...


3

Nick already gave a more specific answer, but I get the sense from your question that you'd benefit from a more generic answer. Different platforms have various ways of getting pixels to the screen. Software is written in layers. You can implement OpenGL on top of D3D (like Microsoft has done), or even on top of GDI as software rendering (like Microsoft ...


2

No you don't need to. As you said yourself, if you whole scene is drawn-over by the new rendered frame, then it makes no visual difference. And as comments rightfully mentionned, it might have a positive or a negative impact on performance: profile on your target platform(s) to make your decision. In the old days, it was common place to actually re-render ...


2

There are billion ways to do this. But here is my approach : To do this i would go with a more "particle" system approach. Rendering a line is quite easy with a quad, you just stretch the four vertex in the direction it´s traveling and then have a maximum length that it never can be longer than. the best way with this is that you can heayily optimize this ...


2

You should be able to use glViewport() to limit rendering to a specific portion of the screen. Just keep in mind that you might have to adjust your matrixes as well. In C++, the call would be as simple as this: glViewport(left, top, width, height); You should find a wrapper for this in LWJGL as well. glScissor() would be another alternative if you don't ...


1

Have different spritesheets for every weapon-component you load as separate image resources. When you want to draw a character, you first draw the character-sprite, and then the sprites of each weapon-component, one after another. Keep in mind that in some cases it might be necessary for some parts to change the drawing offsets of other parts. You might, ...


1

I can suggest the artist’s approach and the engine programmer’s approach. What an animator would do is rework the animation sequences in an application such as MotionBuilder, Maya, or even the free Blender. The animating features of these tools are quite tricky to use at first, but there are plenty of tutorial videos on the web to get started. Another ...



Only top voted, non community-wiki answers of a minimum length are eligible