Blackhawk2-
"I doubt that last part after the comma very much, even if 3dfx would have let him do some benchmarks I really doubt they would let him keep technology that could have fallen into the enemies hands at the time, aka Nvidia."
I don't think you understand that nVidia has not had any interest in tilers to date. They could have made one themselves if they wanted to. Even ignoring that, what are the main advantages to defered rendering-
1) Higher effective fillrate
2) Reduced local memory bandwith needs for video cards
With those in mind, let's think about UT for a minute. With my well over year old GeForce DDR I have to crank the resolution to 1280x960 32bit color to get more then a 1-2FPS drop from 640x480 16bit. Using a V5, Radeon, GF2MX, GTS or any other more current board there is nearly no drop all the way up to 1600x1200. What does this tell us about the UT engine? It isn't too concerned with either local memory bandwith or raw fillrate.
Perhaps a great deal of his distaste for the technology comes from the fact that he uses his own "HSR" in software to deal with the issues that tiling addresses. Will a tiler be more effective? Probably, but not by much, and it likely won't matter anyway as it still will be CPU limited. Given the reality of the Unreal engine his comments make perfect sense.
"Touching a graphics card and doing some benchmarks doesn't mean Sweeny understands the technology or can build one."
When did I say he could? I said I think Carmack likely could, but not Tim. Understands the technology? Graphics cards execute instructions, Tim wrote the instructions for those graphics cards to execute. Without at least a minimal understanding of how things work I find it highly unlikely that he would have as much success as he does.
"It would be like me saying I know how Intel's processors work and can build one because I use them when I program in Visual C++ or Visual Basic. Does that make any sense whatsoever? If you say yes I will have to recommend a mental health expert."
In terms of an overall design if you couldn't come up with one with the experience you have programming then you need more experience. I'm not saying that you become an expert in electronic engineering by learning how to program, it is however an excellent way to learn the shortcomings of a particular design(mainly if you do a lot of optimization, churning out a generic app isn't going to do much). You mentioned the Athlon previously, I had absolute full faith after I saw the design for the first time, because of the experience that I have with coding(which is extremely meager, and not something I care to expand on either). Understanding the basic principles of how processors work isn't too terribly difficult, though creating one from scratch is clearly a very daunting task that takes dozens, if not hundreds, of people many years. If you have worked on optimizing pretty much any piece of software, you quickly learn about the shortcomings of various processors(last I knew AMD and Intel would send you free optimization guides for their particular processors).
Soccerman-
"guys, I think many people praise Carmack too much. he's not a god. there are plenty of great 3d game engines out there, his happens to have the famous name 'Quake' on it, and he happens to be pretty good. but he's NOT a diety."
No, he certainly isn't all mighty, but he is however far beyond the majority. The name "Quake" means something because of Carmack, not the other way around. How many people work on the Quake engines compared to the other offerings? How many can comete not only in terms of graphics, but also network code and particularly stability? None that I can think of, even though they have larger groups of people working on them.
"he may have been an advisor, because he spoke for the programmers in the industry, advising 3dfx on what features they need, and how to get around issues that would come up in doing these features in hardware (because software solutions always come out before hardware solutions)."
The fact is that they thought enough of him to have him on their board. Carmack pushes the envelope of 3D technology with every new game he(and the rest of id) releases. Doom3 looks to be a leap the likes of which we haven't seen since GLQuake, Quake3 was hands down the best looking game to date when it shipped as was Quake2. He takes the initiative of having new additions to both hardware and software APIs(he also is part of the OpenGL "board".
He pretty much by himself forced a second major(third at the time) API to the gaming world and assured that both Linux and Mac users would be able to play at the very least a handful of good games. Carmack has a had a larger influence as a single person then anyone else that I can think of in PC gaming.
Sleepy Tim-
"Hi Ben!"
Hey Tim, how's it going
e-phex
one word(and some numbers)..... Lithtech 2.5
Know of any games on the 2.5 engine that aren't buggy Tis a very good engine, also is quite a bit newer then te Quake3 engine. Not to mention that from what I have seen to date, Quake3's is capable of quite a bit more(seen the Wolfenstein shots).
Mikewarrior2-
Congrats on Elite and thank you
Sunner-
English is my first language and you still spell better and have less typos then I do
ReMeDy-
"Hey Ben why don't you give Dave a position in the basement, i'm pretty sure there is some room left down there!"
The site that Dave built is quite a bit bigger then the Basement