| Sqrl said: I'm kinda tired atm (heading to bed shortly) so if this comes off wrong, I apologise. But you kind of missed what I was saying and then restated it as your position. I don't think you have a problem with my first paragraph so I'll skip it since the WYSIWYG editors many "Level Designers", "Scripters", etc.. use now days is most certainly classified as a very high level, above C. In my second paragraph I say almost verbatim that I wouldn't want to build an engine in Assembly (for the exact reason you mention) and then I go on to say that a lot of engine work is done in assembly. Which is true, major optimizations are absolutely made at this level because its where you have the most control over the hardware. And optimization is a massive part of engine building. Also I think you might be a bit mistaken in your impression of how big an engine really is these days in terms of code length...they aren't short by any means but they aren't insanely large either. Its definitely feasible to build one entirely in Assembly but like you said, its uneccessary to do so with the right tools.
|
I agree with you, that I wouldn't want to write one. I just took it that you thought a large portion was written in assembly, which I would say only a very small portion (a few lines here, few there, but hardly any at all) is written in assembly.
As to the point of being able to write one completely in assembly, I agree, but of course it would be hugely unmanagable. I've worked on a 3D engine before that mapped real time position of objects on a 3D map, with a goal of 1000ks of objects at a time. I mostly built in the ray tracing and bounding box code along with a few other things, but the project was fairly large compared to the engineering tools I typically write.
