"VM = interpreter. No matter how much marketing people may want to convince you otherwise. If it's not native code, it's not run by native CPU, but by a virtual machine, it means it's interpreted."
Nonsense. A just-in-time compiler compiles the bytecode to native code at runtime. Claiming this is "not native code" is BS.
This is not true whatsoever of game code, which is the primary matter at hand. Rarely does a game company write a new engine from scratch, even when they want a very different version. E.g. Quake III is based on Quake II is based on Quake I, UT3 is based on UT2k4 is based on UT2k3 (can't remember the UTE version numbers).
When you're talking about game engines, however, recycling code isn't just an option, it's a necessity. You're not going to rewrite half a million lines of code if you don't have to -- it makes no sense. You update the engine, add new logic, and package it with new assets to release a new game. It's the only way that makes sense.
The code is used for other projects by the same company. Few companies release their old/failed code -- id being the only game company I know of that does so (GPLing their old code).
Where in either of these documents did it say that the routine was written in assembly? I believe you were confused by the PDF's use of a disassembler.
Assembly will only be used for small, high-cost operations. These pieces are small enough that if they malfunction, it's in a way that will be immediately visible.
I'm no DRM fan (I've been working against it for years, e.g. El Tunes and PyMusique), but there's no reason that it'd cause these problems, outside of the authorization problems. Once the game has started, the most the DRM will be doing is decrypting game code, if it's not decrypted entirely at loadtime.
This is bullshit. They're not making money off of the copyrighted works of students, they're making money off of their algorithms. Google is only good because they have sites to index, but it's their indexing code that makes that possible. Google indexes copyrighted materials, should they be compensating siteowners?
Many, many engines support OpenGL. id tech 4 (Doom 3, Quake 4, Prey, etc) was pure OpenGL, Unreal's engine is always D3D and OGL, etc. The only big engine that doesn't support OGL is Source; even stock Gamebryo supports it, although many games opt not to ship with it (e.g. Oblivion, Morrowind (back when it was NetImmerse)) since they make internal changes and don't care about maintaining OGL support.
Once you've reverse-engineered a hardware interface, it's rarely difficult to reverse the next iteration. The basic functionality is usually the same, it's just a slightly different version. The bigger issues are jailbreaking the new device, getting to the point where you can replace the bootloader, etc.
It's being reverse-engineered, like everything else on the iPhone. None of the specs for iBoot, the baseband, etc are public either and the iPhone hackers have done just fine. I'm confident they'll succeed.
There's no reason you can't renegotiate your contract or go elsewhere and get overtime. Companies will use overtime as a way of competing with others for employees and better jobs will result, as they tend to.
You have the choice of either renegotiating your employment contract or going elsewhere, where overtime is paid. Companies will use overtime as a way to compete with other companies for employees -- we'll end up with better jobs in the end.
I disagree entirely. Overtime legislation is akin to forcing employers to pay for health care. I believe that employers of full-time workers should be handled by the companies, but it's not a government issue.
You have to realize that there's not exactly a surplus of good IT workers. Companies have to compete for the good employees, and you can be certain that overtime is going to be one of the ways they draw people in.
You see it as a step backwards, I see it as a step in the right direction.
Sure you can, if your employment contract says you get overtime. Most companies are still going to pay for overtime regardless of whether the government tells them to or not.
Audio drivers, yes (the PMP), but that doesn't come into play here. When it comes to USB, if it/is/ there (which I don't know), it'd be in the specific drivers, although it'd make more sense to handle this at the app level when the copy takes place, since the file would be unusable without transferring license info anyway. Regardless, this can't have any effect since the crash is in an Apple driver, as was specified elsewhere in the comments.
Why would you want a new kernel in it? I do kernel development on a daily basis and the NT kernel is by far the best in popular use.
XNU is an ungodly slow mess of code with so many redundant APIs at every level that it's not even funny. Take a look at L4 to see microkernels done right -- hint, it was created to be a less retarded Mach.
Linux is poorly documented, has little to no code reuse, no real design (leading to modules being rewritten to fix bugs and design flaws while introducing even more), a ton of race conditions (causing stability and security issues), and scales very poorly in an SMP setting (the BKL is a joke).
Solaris is quite nice, but it's not used nearly enough. I've heard good things about the kernels in the BSDs, but I don't have enough experience to talk about them.
A lot of Windows is horrendous, but people talking about the kernel as this terrible thing need to learn what the NT kernel actually does and assign blame where blame it's deserved -- base services and shittacular drivers.
I moved to Assembla ( http://www.assembla.com/ ) after Google removed the MPL and I haven't looked back. It now hosts all of my projects, open and closed source and I really couldn't be happier with it. Well designed (better than Google code, which is rare) and faster than SF ever was, despite the huge strides they've made lately -- it's definitely better than 5 or 6 years ago where CVS would die twice a week:P
Refer to my response to the previous comment in this thread. I'm not saying that the extra memory use doesn't affect performance, I'm saying that the/reason/ it exists has nothing to do with performance.
Slashdot Mirror
"VM = interpreter. No matter how much marketing people may want to convince you otherwise. If it's not native code, it's not run by native CPU, but by a virtual machine, it means it's interpreted." Nonsense. A just-in-time compiler compiles the bytecode to native code at runtime. Claiming this is "not native code" is BS.
This is not true whatsoever of game code, which is the primary matter at hand. Rarely does a game company write a new engine from scratch, even when they want a very different version. E.g. Quake III is based on Quake II is based on Quake I, UT3 is based on UT2k4 is based on UT2k3 (can't remember the UTE version numbers).
When you're talking about game engines, however, recycling code isn't just an option, it's a necessity. You're not going to rewrite half a million lines of code if you don't have to -- it makes no sense. You update the engine, add new logic, and package it with new assets to release a new game. It's the only way that makes sense.
Many application vendors have released their code GPL'd or otherwise open source, but it's very rare in the gaming world.
The code is used for other projects by the same company. Few companies release their old/failed code -- id being the only game company I know of that does so (GPLing their old code).
Where in either of these documents did it say that the routine was written in assembly? I believe you were confused by the PDF's use of a disassembler.
Ugh, sad to hear. When will people learn that this nonsense only hurts them?
Assembly will only be used for small, high-cost operations. These pieces are small enough that if they malfunction, it's in a way that will be immediately visible.
I'm no DRM fan (I've been working against it for years, e.g. El Tunes and PyMusique), but there's no reason that it'd cause these problems, outside of the authorization problems. Once the game has started, the most the DRM will be doing is decrypting game code, if it's not decrypted entirely at loadtime.
This is bullshit. They're not making money off of the copyrighted works of students, they're making money off of their algorithms. Google is only good because they have sites to index, but it's their indexing code that makes that possible. Google indexes copyrighted materials, should they be compensating siteowners?
Wine doesn't support x64 code, so the only benefit you can get is having the full address space enabled (because the kernel can properly map memory).
Many, many engines support OpenGL. id tech 4 (Doom 3, Quake 4, Prey, etc) was pure OpenGL, Unreal's engine is always D3D and OGL, etc. The only big engine that doesn't support OGL is Source; even stock Gamebryo supports it, although many games opt not to ship with it (e.g. Oblivion, Morrowind (back when it was NetImmerse)) since they make internal changes and don't care about maintaining OGL support.
Once you've reverse-engineered a hardware interface, it's rarely difficult to reverse the next iteration. The basic functionality is usually the same, it's just a slightly different version. The bigger issues are jailbreaking the new device, getting to the point where you can replace the bootloader, etc.
It's being reverse-engineered, like everything else on the iPhone. None of the specs for iBoot, the baseband, etc are public either and the iPhone hackers have done just fine. I'm confident they'll succeed.
Erg, "I believe that health-care of full-time workers..." not "I believe that employers of full-time workers..."
There's no reason you can't renegotiate your contract or go elsewhere and get overtime. Companies will use overtime as a way of competing with others for employees and better jobs will result, as they tend to.
You have the choice of either renegotiating your employment contract or going elsewhere, where overtime is paid. Companies will use overtime as a way to compete with other companies for employees -- we'll end up with better jobs in the end.
I disagree entirely. Overtime legislation is akin to forcing employers to pay for health care. I believe that employers of full-time workers should be handled by the companies, but it's not a government issue.
You have to realize that there's not exactly a surplus of good IT workers. Companies have to compete for the good employees, and you can be certain that overtime is going to be one of the ways they draw people in.
You see it as a step backwards, I see it as a step in the right direction.
Sure you can, if your employment contract says you get overtime. Most companies are still going to pay for overtime regardless of whether the government tells them to or not.
This type of thing should be left to employers and not mandated by the government. Thank you for putting some control back where it belongs.
Audio drivers, yes (the PMP), but that doesn't come into play here. When it comes to USB, if it /is/ there (which I don't know), it'd be in the specific drivers, although it'd make more sense to handle this at the app level when the copy takes place, since the file would be unusable without transferring license info anyway. Regardless, this can't have any effect since the crash is in an Apple driver, as was specified elsewhere in the comments.
Mod this bullshit down. The iTunes DRM is 100% inside quicktime.qts.
(I'm the original author of PyTunes, the base for Pymusique -- I know a bit about Fairplay)
Why would you want a new kernel in it? I do kernel development on a daily basis and the NT kernel is by far the best in popular use.
XNU is an ungodly slow mess of code with so many redundant APIs at every level that it's not even funny. Take a look at L4 to see microkernels done right -- hint, it was created to be a less retarded Mach.
Linux is poorly documented, has little to no code reuse, no real design (leading to modules being rewritten to fix bugs and design flaws while introducing even more), a ton of race conditions (causing stability and security issues), and scales very poorly in an SMP setting (the BKL is a joke).
Solaris is quite nice, but it's not used nearly enough. I've heard good things about the kernels in the BSDs, but I don't have enough experience to talk about them.
A lot of Windows is horrendous, but people talking about the kernel as this terrible thing need to learn what the NT kernel actually does and assign blame where blame it's deserved -- base services and shittacular drivers.
I moved to Assembla ( http://www.assembla.com/ ) after Google removed the MPL and I haven't looked back. It now hosts all of my projects, open and closed source and I really couldn't be happier with it. Well designed (better than Google code, which is rare) and faster than SF ever was, despite the huge strides they've made lately -- it's definitely better than 5 or 6 years ago where CVS would die twice a week :P
Refer to my response to the previous comment in this thread. I'm not saying that the extra memory use doesn't affect performance, I'm saying that the /reason/ it exists has nothing to do with performance.