C/C++ is used to program general-purpose processors for a huge variety of operations, where flexibility and expressiveness of the language is paramount. Cg is used to program the highly specialized processor on the graphics card for a relatively narrow range of operations where performance is paramount. They really aren't the same thing.
As for the library idea... Chances are that it will soon become feasible to implement what we would call GL's or DX's feature set entirely on the shader processor, and the CPU-based library would only be charged with shuffling these shaders between system RAM and VRAM and providing bitmaps, models, and other static data. When this happens, it will be feasible to customize the library itself, but today there are still reasons for writing shaders directly.
If you have far more RAM than you are using for your processes, you could have saved some money. Also remember that quite often you are using a good bit more RAM than you have, due to virtual memory.
It would be difficult to use RAM to cache disk writes, since dirty areas would have to be flushed when a process wanted more memory.
This idea provides a neat solution to the problem of how the humans can outperform the computers in the Matrix, suggested above...
Consider this suggestion of running the Matrix process on the human brain as if it was a node in a distributed cluster. There's a great deal of Matrix information stored in the brain, but there's also a human consciousness alongside it in there, unaware that there's data flowing through your unused neurons. "Freeing your mind" could consist of gaining the ability to allow your consciousness to attach to the Matrix simulation the same way a debugger attaches to an existing process (or an aimbot attaches to CS), gain access to its data, and start poking values. The AIs would have to allow individual nodes to be authoritative to realize any net gain, so any changes you imagine to your own Matrix node would be propagated to others as reality, and you would be able to "will" your strength to increase the same way your aimbot can "will" perfect headshots at 100m. This would also explain why hacking the Matrix involves so much activity that resembles meditation/concentration techniques.
The classic reductio ad absurdam: A truly realistic game would, upon your first death, delete itself from your hard disk and refuse to ever run again:P
I accept that ultra-realism is just as much fun as real life. But who says there's nothing funner?
I expect any performance gain from that would be overshadowed by network latency; distributed systems are not meant for real-time processes. It would be more worthwhile to design an engine to take better advantage of multiple processors.
The deeper problem is that a lot of the time, ultra-realism is simply not fun. The technology to blast chunks out of the wall arrived with Red Faction, but it was used very little in the game. It was too easy for a player to blow up the only route to a critical part of the level, or punch a hole in a random wall and completely screw up the scripts and level flow. As for limiting motion to that of a human, I expect most people play games to do things they cannot do in real life, and sprinting for 10 minutes while jumping six feet in the air and waving around a high-powered rifle like a toy satisfies that nicely.
Nonophotorealistic games like JSR or XIII. Abstract games like that thing for the PS2 whose name escapes me right now. Innovation in art design (those graphics still have to be made by someone, no matter how much work the GPU takes on).
You say "end of the world" like it's a bad thing...
Why would this be so bad? What makes us a better or more important form of life than this collective mind, or gives us more of a right to exist than it does?
Look at medical care today: Treatments that save lives in immediate danger are available to pretty much everyone. Treatments for non-critical conditions are available on the open market at prices determined by economics. Elective/cosmetic surgery is available only to the rich, but it's not like those without it are suffering horribly. If genetics turns out the same way it won't be too bad.
Slashdot Mirror
Judg1ng fr0m 7h3 r357 0f 7h3 1n73rn37, 1'd 54y y0u'r3 n07 410n3. 31337 h4X0r1ng!
How is this any different from current piracy?
Don't forget Amdahl's(sp?) Law either. Everything gets optimized sooner or later, if only because everything else has already been optimized.
C/C++ is used to program general-purpose processors for a huge variety of operations, where flexibility and expressiveness of the language is paramount. Cg is used to program the highly specialized processor on the graphics card for a relatively narrow range of operations where performance is paramount. They really aren't the same thing.
As for the library idea... Chances are that it will soon become feasible to implement what we would call GL's or DX's feature set entirely on the shader processor, and the CPU-based library would only be charged with shuffling these shaders between system RAM and VRAM and providing bitmaps, models, and other static data. When this happens, it will be feasible to customize the library itself, but today there are still reasons for writing shaders directly.
Half-Life 2: Quarter-Life.
If you have far more RAM than you are using for your processes, you could have saved some money. Also remember that quite often you are using a good bit more RAM than you have, due to virtual memory.
It would be difficult to use RAM to cache disk writes, since dirty areas would have to be flushed when a process wanted more memory.
Can't wait for a bluetooth tasp! Make my day, baby!
I bet this story gets duped *after* the announcement next week is posted.
Then it's a good thing I didn't read it before posting :)
This idea provides a neat solution to the problem of how the humans can outperform the computers in the Matrix, suggested above...
Consider this suggestion of running the Matrix process on the human brain as if it was a node in a distributed cluster. There's a great deal of Matrix information stored in the brain, but there's also a human consciousness alongside it in there, unaware that there's data flowing through your unused neurons. "Freeing your mind" could consist of gaining the ability to allow your consciousness to attach to the Matrix simulation the same way a debugger attaches to an existing process (or an aimbot attaches to CS), gain access to its data, and start poking values. The AIs would have to allow individual nodes to be authoritative to realize any net gain, so any changes you imagine to your own Matrix node would be propagated to others as reality, and you would be able to "will" your strength to increase the same way your aimbot can "will" perfect headshots at 100m. This would also explain why hacking the Matrix involves so much activity that resembles meditation/concentration techniques.
Yeah, we have standards, dammit! We demand only the best from our stolen movies!
No, the top passwords are LOVE, SEX, SECRET, and GOD.
Nah, let's just get everyone in China to jump off a chair at the same time. And make sure they're aiming away from the sun.
Try beating the game in less than (IIRC) 3 hours.
The classic reductio ad absurdam: A truly realistic game would, upon your first death, delete itself from your hard disk and refuse to ever run again :P
I accept that ultra-realism is just as much fun as real life. But who says there's nothing funner?
E) CowboyNeal.
I expect any performance gain from that would be overshadowed by network latency; distributed systems are not meant for real-time processes. It would be more worthwhile to design an engine to take better advantage of multiple processors.
The deeper problem is that a lot of the time, ultra-realism is simply not fun. The technology to blast chunks out of the wall arrived with Red Faction, but it was used very little in the game. It was too easy for a player to blow up the only route to a critical part of the level, or punch a hole in a random wall and completely screw up the scripts and level flow. As for limiting motion to that of a human, I expect most people play games to do things they cannot do in real life, and sprinting for 10 minutes while jumping six feet in the air and waving around a high-powered rifle like a toy satisfies that nicely.
Nonophotorealistic games like JSR or XIII. Abstract games like that thing for the PS2 whose name escapes me right now. Innovation in art design (those graphics still have to be made by someone, no matter how much work the GPU takes on).
You say "end of the world" like it's a bad thing...
Why would this be so bad? What makes us a better or more important form of life than this collective mind, or gives us more of a right to exist than it does?
Look at medical care today: Treatments that save lives in immediate danger are available to pretty much everyone. Treatments for non-critical conditions are available on the open market at prices determined by economics. Elective/cosmetic surgery is available only to the rich, but it's not like those without it are suffering horribly. If genetics turns out the same way it won't be too bad.
Because you don't die when your computer crashes?
Yes, the discovery of warp drive was a major influence on our modern world. I can't imagine what the 20th century would have been like without it.
Oh, wait...
Well, it's the present right now... Where the hell is my flying car already?
We have a vaccien for measles.