by TBone » Aug 15, 2001 @ 5:10pm
Not necessarily harder. Depends on the task. Because you're directly manipulating display memory using the GAPI, you're basically writing the entire screen yourself, which means that writing a 3D engine might require more code, but has the potential to be a lot more flexible. DirectX will do your sound, your video, your 3D, and your input for you, but it also adds a degree of overhead.<br><br>I personally prefer working with the GAPI. I like being able to do the work myself, especially since all engines are software on current PPC hardware. It's likely that any implementation you write will be faster because you know exactly what you need instead of using Microsoft's all-purpose code.<br><br>The downside is that you'd have to write your own poly-fill routines, your own rotation functions, and your own blit. These are NOT hard, in fact they're much easier than you think. They are time-consuming to code, but what you gain is speed and efficiency.<br><br>I guess the best way to compare the difficulty is that between C and assembly coding. Assembly is very low-level, making it faster in execution but more time-consuming to write. C is higher-level, so you can have certain common code practices handled for you by the compiler, but you lose some of that execution speed. Now imagine GAPI as assembly and DirectX as C and you've got it.
3V1L L337 H3150