Sylas ended the 64bit derail. It's done.
New article:
Development Update: Star Citizen Alpha 2.0 & Star Marine - Roberts Space Industries
New article:
Development Update: Star Citizen Alpha 2.0 & Star Marine - Roberts Space Industries
Because of exponential growth.I know the whole thing is bullshit because Chris Roberts has made outlandish statements like they can make zones that are "billions of kilometers across" just because they moved from 32 bit to 64 bit. He also said that standard 32 bit maps are just "a few kilometers across." Okay, Einstein, how do you go from a few kilometers to billions of kilometers by just doubling the bits. That's not what I call adding up your math right.
Uh, I thought my attempt to be ridiculous would garner more laughs. I guess this goes to show you guysreally dothink I'm a retard.Because of exponential growth.
The C++ standard is
FLT_MAX: 3.40282347E+38F
DBL_MAX: 1.7976931348623157E+308
The GNU C Library: IEEE Floating Point
float is usually considered a 32bit floating point number, double is a 64bit floating point number.
To give you some perspective, 1e2 is 100, 1e3 is 1000 and 1e4 is 10000. going from 38 to 308 is increasing the order of magnitude of resolution some 270 times.
Nah, that's cool, and to be honest I think when I first learned about 64 bit vs 32 bit, I thought, "Oh yeah, that's a lot better." I never really looked at how big the numbers got when you start going up those orders of magnitude. I don't remember where, but on some board where people were talking about SC (probably reddit) there was a math wiz that illustrated a lot more clearly how significant the difference was.Haha sorry =\
It's not that I think you or anyone is a retard for treating precision due to bit size intuitively (double your bits, double your precision), it's just that as a programmer I'm baffled that gamers are so focused on such a technical discussion. Nobody knows the details of how they are accomodating higher precision tracking of objects in space (It's not just simply compiling in 64 bit, as the above article indicates it's all about choosing which data structures to use 64bit with).
Yea. If you're only working with integer/fixed precision numbers, then going from 32 bits to 64 gives you 4 billion larger coordinates. Otherwise, if you're dealing with floating point, you go from plank-scale precision to beyond the edge of the visible universe, but it becomes more and more imprecise as objects recede...Nah, that's cool, and to be honest I think when I first learned about 64 bit vs 32 bit, I thought, "Oh yeah, that's a lot better." I never really looked at how big the numbers got when you start going up those orders of magnitude. I don't remember where, but on some board where people were talking about SC (probably reddit) there was a math wiz that illustrated a lot more clearly how significant the difference was.
I have a fondness for the Jaguar personally.COME ON MAN, DO THE MATH! 64 bit **ALWAYS** wins
(Yes I bought one of these lol.... Oh how sad 99% of the games were)
((Except Aliens Vs Predator )
Fired up my SC/AC for the first time in a year, and yeah its pretty simple actually. I was suprised...SC is really simple right now, I don't understand the complaint.