Mis-use of wide customer base:
In the days of MS-DOS 4 it was said that if you wrote exactly the same program in Microsoft C and in Borland C and ran them on an MS-DOS machine, even if they both compiled into the same machine code, the one written in Borland C would run slower. This is becuase each of the compilers sign the executables they create differently. MS-Dos would simply look at the signature and decide whether to slow it down or not. The result would be that, since most people used MS-DOS, people would assume the MS C compiler was better.
Can you back this up? It's the first I've heard of any such thing.
I have heard of (more recently) Intel deliberately sabotaging their compiler on AMD processors (which runs Intel-compiled code better than any), but this is the first I've heard of MS-DOS detecting Borland-compiled code.
I do know, however, that Microsoft sabotaged Win3 on DR-DOS, and "encouraged" their users to switch to the (inferior) MS-DOS. But slowing down Borland code?
It doesn't seem you really understand how to make something secure. Me understands that probably not much either, but basically any professional would tell you that making single identity is a way in other direction. That's precisely what is called "single point of failure".
My post was made mostly tongue-in-cheek, as I had assumed most people would realize that I'm talking about some sort of widely-supported scheme generally similar to how PGP works. Something for global verification, and potentially identification. A carefully-generated RSA key pair is far more secure and reliable than OpenID, IMO, and, building on the existing GPG/PGP community, could be a far more reliable solution. I don't pretend to be an expert on security and encryption, but even I can see that OpenID has too many points of weakness, and is far too insecure and unreliable to be any kind of global passwordless identification. I'm far more willing to rely on my GPG signature and its 1024-bit key than I am on something that has a far greater potential for fraud.
In other words. What would be easier to crack: many of presumably belonging to single person identities or single one which strongly identifies single person?
A strawman argument, I think. OpenID is a sigular identity, just like PGP, but OpenID relies on your access to a certain account - which may be easily broken into or forged - while PGP relies on a privately-held, unique, and nearly impossible to forge encryption-grade key, which can be destroyed and invalidated by its owner at any time in favor of a new one. (I have two keys myself, the other I plan to invalidate as soon as I can get back into my old hard drive...)
I really need to back up my shit more often.
Conclusion: nobody in real world would try to snatch my public house card - though that would allow them to lift lots of precious books in my name. Same goes for about 20 other cards I have. But if I would have single card for everything, once it is forged I might be set back for a large amount of money. As long as there are 20 cards - not one - the trouble of forging them outweigh any gain. But having single card opens me up to possibly infinite number of problems: single point of failure make any failure fatal compromising everything you do and did.
(Emphasis mine.) Again, this is exactly the problem with OpenID, IMO. But like with your credit card, you can have it invalidated immediately if it's ever stolen, and you can have it replaced with a new one very easily - just like with PGP.
Now, I don't claim to unerstand the terms of the XNA license, but I got the definite impression that you couldn't self-publish games either onto the marketplace or for free distribution - it has to be published through Microsoft.
You can't even share a game YOU wrote with a friend unless they also have a developer account, and even then, it has to be done over the XBL network.
So not only is this bad for developers who want to release their work for free or under their own license, but it forces you into a position of relying on Microsoft to publish your work regardless of your own wishes.
Am I simply misunderstanding something here, or is XNA really as idiotic as it looks to me?
For example, none of the OpenPGP signatures in this thread verify under PGP Desktop 7.1 ("ascii armor input incomplete").
That's because the/. comment code forcefully breaks up long words. Look closely and you'll notice that the signature itself has a spurious space partway. Delete those two spaces and they validate just fine.
Indeed. OpenID also seems too unreliable. What's to say the server my blog is on won't get hacked again? What's to keep the crackers from using that to forge my identity? There's no signing mechanism, no challenge/response, and it doesn't even bother to protect my "identification" from interception or duplication! All it does is prove that I have access to the blog I linked to.
What I want is a complete solution that allows me to protect my identification by a strong encryption schema and use that everywhere - maybe have a Firefox extension (or a user.js in Opera) that handles the legwork for me. I don't know, it probably doesn't exist. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux)
The inverse square root is used to calculate a normal vector. Normal vectors, are, in turn, used in lighting calculations (to determine, for example, the light intensity for a given vertex), among other things.
AMD's 3D Now! instruction set includes a set of instructions (TFSQRTS/TFSQRTD) for approximating the inverse square root.
I'm actually very wary of an upgrade to 2.6.19 (I still haven't tried any of the new Gentoo patch bundles for 2.6.18!) because of the trouble I had with 2.6.17.
I swear, all of the odd-numbered releases have sucked lately.
When it's in that state, it's entirely in kernel mode, pulling things off the swap that it shouldn't have put there to begin with. It just drops back to the idle process while it waits for the hard drive to scratch the magnetic particles off the platter.
Honestly, I know of no other OS which WASTES so much time and tries so hard to overwork hardware than Windows' bad virtual memory manager.
Re:The main problem are graphics drivers
on
More Bioware For Linux?
·
· Score: 2, Informative
Not true. NVIDIA's driver supports OpenGL 2.1 with full support for every feature of the GLSL - provided you've got the hardware for it. The latest version even adds a new GLX extension that the X server can use for hardware-accelerated compositing. ATi's driver, while it still mostly sucks, is at a similar level of advancement. OpenGL 2.0 support (last I heard) with GLSL support matching the Windows driver.
The only drivers that would have problems running games like NWN2 (should it ever be ported) are things like Intel's GMA and crappy hardware like SiS chipsets and so forth.
Then again, if that's all you've got, I've got to wonder why you're trying to make a fancy 2D graphics chip do 3D. Then again, you'd have an ATi or NVIDIA card anyway... Right?
I really think anyone in law enforcement needs to be held accountable. A zero-tolerance policy. If there is any evidence that the behavior of an officer is not in line with his duties (such as with brutality), then all duties should be revoked, all benefits ended, pay terminated, and all ties cut.
It is the job of an officer to carry out the enforcement of the letter of the law, is it not? It is his job to ensure that the law is being obeyed. So, if he disobeys that law, then the law should be even less lenient to him than to the common citizen. Taken to the logical extreme: If a police officer kills an innocent man under any circumstances - then the fullest extent of the punishment for such a crime should be carried out.
Too many police officers believe themselves to be above law. They need a deterrent far stronger than those over whom they enforce it.
The CPU is an IBM PowerPC 750CXI, nearly identical to the GameCube's PowerPC 750CX. (It's clocked higher and has a few more advanced features over the older model.) This, in addition to the higher RAM capacity, more powerful GPU, and DVD-style drive, makes little difference to the OS. It doesn't surprise me at all that it already runs GCLinux.
Why is Linux being lambasted here for poor backwards compatibility?
I've heard stories of running code from linux 1.0 on linux 2.0, and I have recently used software compiled for linux 2.2 on 2.6. So why is the backwards incompatibility of GlibC and GCC and other common libraries whose APIs have changed as they mature suddenly the fault of linux?
Blame the GNU guys for GlibC's versioning scheme if you really must complain. Blame them for the change in binutils' ABI changes in recent versions.
After all, this is why most linux distros have libstdc++.5 (GCC 3.3 and prior) installed next to libstdc++.6 (GCC 3.4 and later)!
Anyway, even somewhat ancient games that use OSS, an old MesaGL build, and were statically linked and compiled with GCC 2.95 still run. So what's the problem?
I've been wondering recently if such a functionality is available in Linux. One of my clients is a health center that would like to migrate toward a thin-client solution. We'd like to keep people from storing, or worse carrying out, "protected health information," so being able to block USB storage devices would be a good feature.
Easy. In the kernel configuration, disable everything except HID under USB. Keyboards and Mice will work, but nothing else will. Don't pass out the root passwords and practise good security and the users can't set up a new kernel or insmod anything.
They're quite stable, actually. In a couple months use of Compiz and a few weeks with Beryl, I've had virtually no crashes that were not directly my fault. The only issues I've run into is with Beryl in KDE with the screen saver and OpenGL games. (The console never seems to wake up beyond the mouse cursor - I have to drop to a console and kill the screen saver and password lock processes before it's usable again. This may simply be a problem with the way KDE locks the console.) Occasionally, Beryl will actually crash an OpenGL app for no apparent reason - It may be my configuration, but I haven't traced the problem yet. Compix doesn't seem to have the problem.
Beryl does have a few minor performance problems, which can be quickly solved by disabling unneeded plugins. (Honestly, who needs the water plugin?)
Beyond that, my only complaint is that a full half of the 3D effects are useless and downright annoying. The cube (or desktop plane in Beryl) and window deformation is really all that is worth having. (Some of the skins are really pretty though.)
I've been begging for an answer to my ever-so-simple question for months now. It's so trivial a thought that I'm amazed that i have yet to actually get a response, much less an answer:
Why buy Vista?
DirectX 10 is hardly a selling point, because, as a developer, I see very little that makes it special enough to consider as a target API, and with such a small audience (i.e., one solitary platform with the possibility of porting to XBox 360), I fail to see the point of even giving it a passing glance. Aero Glass is only merely interesting. I use AIGLX on Linux at home and actually turn it off from time to time because Beryl's effects get kinda old after a while (and Compiz is just plain boring). MacOS X has similar OpenGL-accelerated windowing effects, and has had them for quite a while, so it seems little more to me than a passing fad.
All the other benefits and new features that Vista will actually include elude me. In researching the product, I see no reason to even bother looking at Vista for any computer. Yet, I see so many people getting so excited about Vista and all the new stuff it will bring to their systems, that I have to stop and wonder.
So, I am asking, why does anyone want to buy Vista? Why should I consider an upgrade path to Vista? What's so exciting about it? I really want to know! Someone, PLEASE tell me!
All this talk about all the glitz and glamor that is Aero Glass, and not one word said about why someone would actually bother!
I'm still waiting for an answer to the question I've asked here and elsewhere at least half a dozen times: Why would I (as a user or sysadmin, not a developer) bother upgrading from XP to Vista?
(Disclosure: I'm a programmer and I use both XP and Linux. I'm just trying to understand WHY anyone would get Vista.)
Slashdot Mirror
Can you back this up? It's the first I've heard of any such thing.
I have heard of (more recently) Intel deliberately sabotaging their compiler on AMD processors (which runs Intel-compiled code better than any), but this is the first I've heard of MS-DOS detecting Borland-compiled code.
I do know, however, that Microsoft sabotaged Win3 on DR-DOS, and "encouraged" their users to switch to the (inferior) MS-DOS. But slowing down Borland code?
I submit that the estimate is 95% accurate.
/d/, however, has a slightly greater impact on IQ.
Perhaps if the SNES wasn't a 16-bit platform...
Months for compiling OOo? That's being generous.
My post was made mostly tongue-in-cheek, as I had assumed most people would realize that I'm talking about some sort of widely-supported scheme generally similar to how PGP works. Something for global verification, and potentially identification. A carefully-generated RSA key pair is far more secure and reliable than OpenID, IMO, and, building on the existing GPG/PGP community, could be a far more reliable solution. I don't pretend to be an expert on security and encryption, but even I can see that OpenID has too many points of weakness, and is far too insecure and unreliable to be any kind of global passwordless identification. I'm far more willing to rely on my GPG signature and its 1024-bit key than I am on something that has a far greater potential for fraud.
A strawman argument, I think. OpenID is a sigular identity, just like PGP, but OpenID relies on your access to a certain account - which may be easily broken into or forged - while PGP relies on a privately-held, unique, and nearly impossible to forge encryption-grade key, which can be destroyed and invalidated by its owner at any time in favor of a new one. (I have two keys myself, the other I plan to invalidate as soon as I can get back into my old hard drive...)
I really need to back up my shit more often.
(Emphasis mine.) Again, this is exactly the problem with OpenID, IMO. But like with your credit card, you can have it invalidated immediately if it's ever stolen, and you can have it replaced with a new one very easily - just like with PGP.
"Comparing apples to oranges" is perfectly valid. Please use a different silly analogy.
Now, I don't claim to unerstand the terms of the XNA license, but I got the definite impression that you couldn't self-publish games either onto the marketplace or for free distribution - it has to be published through Microsoft.
You can't even share a game YOU wrote with a friend unless they also have a developer account, and even then, it has to be done over the XBL network.
So not only is this bad for developers who want to release their work for free or under their own license, but it forces you into a position of relying on Microsoft to publish your work regardless of your own wishes.
Am I simply misunderstanding something here, or is XNA really as idiotic as it looks to me?
My pr0n collection, for starters?
That's because the /. comment code forcefully breaks up long words. Look closely and you'll notice that the signature itself has a spurious space partway. Delete those two spaces and they validate just fine.
They should fix that.
-----BEGIN PGP SIGNED MESSAGE-----
x 9H8y8oK3NRNOsym+Ofu8XBgCfQ52+e M6qE=
Hash: SHA1
Indeed. OpenID also seems too unreliable. What's to say the server my blog is on won't get hacked again? What's to keep the crackers from using that to forge my identity? There's no signing mechanism, no challenge/response, and it doesn't even bother to protect my "identification" from interception or duplication! All it does is prove that I have access to the blog I linked to.
What I want is a complete solution that allows me to protect my identification by a strong encryption schema and use that everywhere - maybe have a Firefox extension (or a user.js in Opera) that handles the legwork for me. I don't know, it probably doesn't exist.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
iD8DBQFFdbVCi1yS1BuzIvgRAnMxAJ9qG+
jj6A/Oyo3ez/9QGuwL
=IaLD
-----END PGP SIGNATURE-----
The inverse square root is used to calculate a normal vector. Normal vectors, are, in turn, used in lighting calculations (to determine, for example, the light intensity for a given vertex), among other things.
AMD's 3D Now! instruction set includes a set of instructions (TFSQRTS/TFSQRTD) for approximating the inverse square root.
I'm actually very wary of an upgrade to 2.6.19 (I still haven't tried any of the new Gentoo patch bundles for 2.6.18!) because of the trouble I had with 2.6.17.
I swear, all of the odd-numbered releases have sucked lately.
When it's in that state, it's entirely in kernel mode, pulling things off the swap that it shouldn't have put there to begin with. It just drops back to the idle process while it waits for the hard drive to scratch the magnetic particles off the platter.
Honestly, I know of no other OS which WASTES so much time and tries so hard to overwork hardware than Windows' bad virtual memory manager.
Not true. NVIDIA's driver supports OpenGL 2.1 with full support for every feature of the GLSL - provided you've got the hardware for it. The latest version even adds a new GLX extension that the X server can use for hardware-accelerated compositing. ATi's driver, while it still mostly sucks, is at a similar level of advancement. OpenGL 2.0 support (last I heard) with GLSL support matching the Windows driver.
The only drivers that would have problems running games like NWN2 (should it ever be ported) are things like Intel's GMA and crappy hardware like SiS chipsets and so forth.
Then again, if that's all you've got, I've got to wonder why you're trying to make a fancy 2D graphics chip do 3D. Then again, you'd have an ATi or NVIDIA card anyway... Right?
I really think anyone in law enforcement needs to be held accountable. A zero-tolerance policy. If there is any evidence that the behavior of an officer is not in line with his duties (such as with brutality), then all duties should be revoked, all benefits ended, pay terminated, and all ties cut.
It is the job of an officer to carry out the enforcement of the letter of the law, is it not? It is his job to ensure that the law is being obeyed. So, if he disobeys that law, then the law should be even less lenient to him than to the common citizen. Taken to the logical extreme: If a police officer kills an innocent man under any circumstances - then the fullest extent of the punishment for such a crime should be carried out.
Too many police officers believe themselves to be above law. They need a deterrent far stronger than those over whom they enforce it.
The CPU is an IBM PowerPC 750CXI, nearly identical to the GameCube's PowerPC 750CX. (It's clocked higher and has a few more advanced features over the older model.) This, in addition to the higher RAM capacity, more powerful GPU, and DVD-style drive, makes little difference to the OS. It doesn't surprise me at all that it already runs GCLinux.
Why is Linux being lambasted here for poor backwards compatibility?
I've heard stories of running code from linux 1.0 on linux 2.0, and I have recently used software compiled for linux 2.2 on 2.6. So why is the backwards incompatibility of GlibC and GCC and other common libraries whose APIs have changed as they mature suddenly the fault of linux?
Blame the GNU guys for GlibC's versioning scheme if you really must complain. Blame them for the change in binutils' ABI changes in recent versions.
After all, this is why most linux distros have libstdc++.5 (GCC 3.3 and prior) installed next to libstdc++.6 (GCC 3.4 and later)!
Anyway, even somewhat ancient games that use OSS, an old MesaGL build, and were statically linked and compiled with GCC 2.95 still run. So what's the problem?
Easy. In the kernel configuration, disable everything except HID under USB. Keyboards and Mice will work, but nothing else will. Don't pass out the root passwords and practise good security and the users can't set up a new kernel or insmod anything.
For God's sake, get that guy some Valium.
I'd laugh if a variation on the old AIDS virus was brought to Vista. That'd just be hillariously ironic.
They're quite stable, actually. In a couple months use of Compiz and a few weeks with Beryl, I've had virtually no crashes that were not directly my fault. The only issues I've run into is with Beryl in KDE with the screen saver and OpenGL games. (The console never seems to wake up beyond the mouse cursor - I have to drop to a console and kill the screen saver and password lock processes before it's usable again. This may simply be a problem with the way KDE locks the console.) Occasionally, Beryl will actually crash an OpenGL app for no apparent reason - It may be my configuration, but I haven't traced the problem yet. Compix doesn't seem to have the problem.
Beryl does have a few minor performance problems, which can be quickly solved by disabling unneeded plugins. (Honestly, who needs the water plugin?)
Beyond that, my only complaint is that a full half of the 3D effects are useless and downright annoying. The cube (or desktop plane in Beryl) and window deformation is really all that is worth having. (Some of the skins are really pretty though.)
I've been begging for an answer to my ever-so-simple question for months now. It's so trivial a thought that I'm amazed that i have yet to actually get a response, much less an answer:
Why buy Vista?
DirectX 10 is hardly a selling point, because, as a developer, I see very little that makes it special enough to consider as a target API, and with such a small audience (i.e., one solitary platform with the possibility of porting to XBox 360), I fail to see the point of even giving it a passing glance. Aero Glass is only merely interesting. I use AIGLX on Linux at home and actually turn it off from time to time because Beryl's effects get kinda old after a while (and Compiz is just plain boring). MacOS X has similar OpenGL-accelerated windowing effects, and has had them for quite a while, so it seems little more to me than a passing fad.
All the other benefits and new features that Vista will actually include elude me. In researching the product, I see no reason to even bother looking at Vista for any computer. Yet, I see so many people getting so excited about Vista and all the new stuff it will bring to their systems, that I have to stop and wonder.
So, I am asking, why does anyone want to buy Vista? Why should I consider an upgrade path to Vista? What's so exciting about it? I really want to know! Someone, PLEASE tell me!
AIM works. It's the only way I can get files from windows users without resorting to IRC or some other protocol.
All this talk about all the glitz and glamor that is Aero Glass, and not one word said about why someone would actually bother!
I'm still waiting for an answer to the question I've asked here and elsewhere at least half a dozen times: Why would I (as a user or sysadmin, not a developer) bother upgrading from XP to Vista?
(Disclosure: I'm a programmer and I use both XP and Linux. I'm just trying to understand WHY anyone would get Vista.)
Tried this?