You can use SSE math (though it's not enabled by compilers by default), you just can't pass scalar floating-point arguments in SSE registers.
Also this is a bit irrelevant to what the message I replied to said. The poster asked whether multiple x32 processes could still use more than 4GB. I replied that it is already the case with i386 processes running with an amd64 kernel (actually, it's also even possible with a 32-bit kernel with PAE enabled).
I see, so the whole system exists for sandboxing. But then why not do a simpler approach where each module runs in its own process and communicates with the master through an IPC mechanism (sockets, shared memory, named pipes...)?
From a quick reading, I don't get the point of this system. Couldn't they just compile their C code to a dll normally and then load that dll and call vmMain?
I don't see what those "virtual machines" are bringing.
For VS 2010, C++11 (then 0x) implementation was actually better than g++
MSVC10 is very bad at standard conformance, even with C++98. It has so many bugs it's not even funny, notably in name lookup and overload resolution. When bugs are reported to them they refuse to fix them. GCC, on the other hand, has had very good C++ support for a long time and is only getting better.
What's the link between technology and D&D? Roleplaying (which involves real-life social interaction) in a medieval setting (which has no technology) is pretty far from IT or software development.
It is true that software engineers often fancy board and role playing games. But it's not because of technology, it's because of their mindset.
Then it wasn't really Computer Science. Computer Science is a theoretical field akin to Meta-Mathematics and Logic.
True, its main application is to solve problems, and the most complex and interesting problems are usually in Physics. But that's just an application, it's not what the discipline itself is.
The point he made is that at current world population growth we'll be forced to kill off most of our population to survive. I pointed out that we could feed 20 billion people just fine, which, with the growth of the past 20 years, gives us at least 50 years.
World population growth is however expected to drop dramatically in the next 20 years.
You mean, as long as we don't mind eating pesticide-infused GMO crops, foregoing beef, and hoping that fisheries magically don't collapse?
We're talking about manufacturing food for a huge quantity of people as opposed to killing off four fifth of the world population. Beef? Fish? You're dreaming. They don't yield enough protein for the space they use.
Slashdot Mirror
So essentially the advantage of this virtual machine is that it forces mod providers to provide a retargetable source of their mods?
You can use SSE math (though it's not enabled by compilers by default), you just can't pass scalar floating-point arguments in SSE registers.
Also this is a bit irrelevant to what the message I replied to said. The poster asked whether multiple x32 processes could still use more than 4GB. I replied that it is already the case with i386 processes running with an amd64 kernel (actually, it's also even possible with a 32-bit kernel with PAE enabled).
I'm not missing anything. Your comment, albeit replying to mine, has nothing to do with it.
What kind of advantage does that have over recompiling the mods for a new architecture?
Linking a DLL does not require recompilation.
This was decided later. It still doesn't explain why they'd think their system would be faster than inter-process communication.
Whatever overhead an IPC mechanism would have would be insignificant compared to the overhead of running interpreted code as was initially desired.
I see, so the whole system exists for sandboxing.
But then why not do a simpler approach where each module runs in its own process and communicates with the master through an IPC mechanism (sockets, shared memory, named pipes...)?
I wrote this very comment *because* I read that article.
From a quick reading, I don't get the point of this system.
Couldn't they just compile their C code to a dll normally and then load that dll and call vmMain?
I don't see what those "virtual machines" are bringing.
You've always been able to run 32-bit applications with a 64-bit kernel. x32 is not required for this.
You don't need to use x32 for this. It works fine with the usual i386/i686 ABI.
MSVC10 is very bad at standard conformance, even with C++98. It has so many bugs it's not even funny, notably in name lookup and overload resolution. When bugs are reported to them they refuse to fix them.
GCC, on the other hand, has had very good C++ support for a long time and is only getting better.
You mean being considered then discarded?
On Linux, Nvidia is still clearly the best option as long as you're ok with proprietary drivers.
Pretty much. They use special intrinsics mapped to instructions specific to the floating-point units that ARM processors have (VFP, NEON).
A global law? You mean from the world government?
You think no one tried that one before?
What's the link between technology and D&D? Roleplaying (which involves real-life social interaction) in a medieval setting (which has no technology) is pretty far from IT or software development.
It is true that software engineers often fancy board and role playing games. But it's not because of technology, it's because of their mindset.
Then it wasn't really Computer Science. Computer Science is a theoretical field akin to Meta-Mathematics and Logic.
True, its main application is to solve problems, and the most complex and interesting problems are usually in Physics. But that's just an application, it's not what the discipline itself is.
Software design is part of Software Engineering, not Computer Science. Those two are about as different as Electrical Engineering and Physics.
The point he made is that at current world population growth we'll be forced to kill off most of our population to survive. I pointed out that we could feed 20 billion people just fine, which, with the growth of the past 20 years, gives us at least 50 years.
World population growth is however expected to drop dramatically in the next 20 years.
I suggest you read the parent message by Taco Cowboy.
http://slashdot.org/comments.pl?sid=2911939&cid=40303213
We're talking about manufacturing food for a huge quantity of people as opposed to killing off four fifth of the world population. Beef? Fish? You're dreaming. They don't yield enough protein for the space they use.
Everything said in that summary is so obvious I hope the authors were given a super hero medal for figuring it out.
With current agriculture technologies, the Earth can provide sustenance for 20 billion people.