I read this news item assuming that it would be informative on what sort of video game violence was being kept from those under the age of 18, or at least would talk about video game violence.
Instead, it takes the point of the hypocrisy of a government.
While I am a fan of games of this genre (not especially SoF, but many others), allow me to point out a flaw in your logic:
When an individual kills something in a game, it's for fun. You can argue whether it's a "killing simulation" (I'd disagree) or merely a game of skill, but either way, you're finding pleasure in it. To compare this with the slaughter of animals for food is completely unfounded. Unless you're talking about some twisted individual who takes joy in killing cows in a meat packing plant, then you're off on your analysis. The people who work in those establishments aren't having fun, or killing for the sake of killing. It's a job.
If you want to analyze whether it desensitizes people or makes them like killing, then post under that topic. This has nothing to do with video game violence.
While a lot of users tend to download mp3s off of Napster a lot of ripping "groups" still exist in forums such as IRC. The use of files that list the results of a hashing algorithm such as md5 on an album, for example, is often distributed with mp3s.
When downloading mp3s off of another person with Napster or gnutella, you're bound to get a bunch of incorrect, incomplete, or otherwise flawed audio files. They may play, but this is the sort of thing bands are using as an argument- this doesn't accurately portray their original work. If widespread distribution of formats like mp3 is going to work, either on a commercial *or* illegal basis, some sort of checksum system needs to be in place.
It's sort of interesting that they're using the four-head Matrox boards to power these things. While consumers are now seeing boards that have output for two monitors from Matrox, according to a friend of mine, Matrox makes a lot of specialty boards like the one mentioned. Some of the four screen models are used in financial institutions or somesuch.
As for the technology driving it, it's a massive board (or combination of boards) powered by the G200 chipset. Matrox may be making these based on the G400 (or even G450) by now, but I'm not sure.
IBM must be using some sort of tiling scheme to display the stuff. xinerema in hardware?:)
While Linux-based systems are infiltrating the server world, I question whether they are replacing current commercial unix systems or creating a new niche. While SCO's apparent demise (and possible upcoming resurrection?) may be due to a surge of Linux use, I would doubt it.
More than likely, their business was merely not as profitable anymore in a developing industry. I'd like to see statistics, to know whether people were replacing SCO systems with other commercial unix systems or Linux/*BSD.
It may be a bit pretensious to assume it was Linux pushing SCO out of the market.
Exactly. But the point is, since Rob is so addicted to Diablo II, he's going to play it eventually, so can't he wait and just watch the clip before he plays Diablo II? For things that aren't a high priority, your OS may not have to do every task. As for the five minutes, how many months is he going to have to wait for somebody to port the Sorenson codec anyway? What if someone were to hand him a VHS tape with a movie trailer? I'd bet it'd take a good five minutes for him to walk over to his TV, turn on the VCR, pop in the tape and get everything going. We're living in a world of convenience, and we want every tool to do every job. It just isn't the most efficient that way because we end up with bloated mismanaged tools. Perspective.
The real question here is whether Microsoft is going to keep this technology Windows-only, if they're going to make the minimal showing (a Mac version, possibly even HPUX or Solaris like IE), or if they're going to go full out and either port it to a lot of platforms. Even in that case, it's a toss-up whether they port it themselves, allow others to under NDA (like some Java VMs), or allow the source to the compiler/interpreter to be open enough to allow others to do so.
If they wanted to truly standardize, they could get the language ANSI certified or somesuch. While APIs aren't going to be the same, the language would have a basis, and Microsoft would be able to say they have a language controlled by an independant standards body, while Sun would monopolize Java. This would be sort of similar to the approach they took with their web browser
While this is filed under the "from the what'd-you-expect-from-AOL-and-Microsoft dept.", I'm sort of doubting that AOL and Microsoft are purely to blame for this. Ironically, "the Center for Democracy and Technology" is credited in the press release. But what I'm wondering is this: Are upcoming specifications that the W3C are going to release public? Is there a period for public review of upcoming technologies? I would think problems like this, and the flaws pointed out in the article, would have to be addressed. It really sounds from the press release that unless you're a corporation in on the development, your input doesn't count. Should the W3C's drafts have to undergo public review? Or do they already, and I'm missing a step...
When will these people catch on that tracking an IP or username means amazingly little? If users on Napster had to register a username by giving personal information, then maybe this would be possible. But there's no such correlation. Tracking by IP, on the other hand, is even more useless. If I'm a modem user and you know my ISP, how the hell does that help? Do you sue the ISP? I think not. They don't really have any control. However, some college campuses exert an insane amount of control over local networks. Most are completely lax, but some (I've heard UIUC does this) actually require you to register your MAC address with the computing center to obtain an IP from DHCP. So if a college ever cooperated with any sort of agency, however unlikely, they could trace you. This brings up a question: how long will IP addresses not be traceable to a physical location? Sure, a lot of modem pools will still assign IPs fairly randomly to dialup customers, but static IPs, or at least IPs that seldom change, are increasingly common.
Indeed, there is a proliferation of network operating environments at the moment. To claim that it's a pretty saturated market is accurate as well, but the fact that few people are catching on to is that Inferno is in no way new. I read about it and downloaded the environment to toy with around two years ago. It's made an appearance in several phones with an interface, among other places. This isn't an up-and-coming product, but rather one that is already in place.
Interesting that you've decided to compare a couple authors I have an interest in I've read a great deal of RAW and he's just as guilty of every "sin" that you claim Adams is. While you could claim Wilson had some grand scheme to put the events in a certain order in works like the Illuminatus Trilogy, you're going to have a hard time convincing me of that. Other than the fact that both authors have the ability to somewhat connect disjointed thoughts (although Wilson's are much more disjointed), you're comparing apples and oranges. Adams has been a writer of humorous British fiction. RAW was American and wrote about a variety of topics, exploring everything from social taboos to the mingling of philosophies to his views on death. And as for the opiate overdose, no offense, but that's most likely to occur in Wilson's camp, not Adams'.
I'm not completely sure how this amazing protocol Microsoft has cooked up works, but if it's anything like the WMA files that required authorization, it's something like this:
User downloads file. When they attempt to play it, the program connects to the server and grabs the key if you're allowed to have it. You can then play that file freely on your system.
As far as I could tell, it didn't actually unlock the file, per se. I would guess it stores the key in a special little registry section so you can play it later.
So the question becomes this: How do they keep the user from sending this registry bit to someone else? I would bet it's generated with the system GUID (that global unique identifier key that caused some controversy a while back). It's a fairly common thing for programs to use. So the client sends the server its GUID, the server returns with a key that works to unlock for that user's GUID, and it works seamlessly.
Where they ran into problems before was with programs like Winamp. When using an output plugin that could write to a file, Winamp could get around this restriction by removing the authentication through playback.
If a single application is primarily responsible for the copyright-violating actions, it could possibly be attacked in the way the MPAA has pursued DeCSS. But what if I develop a distributed protocol? The strength of a system like gnutella isn't the program itself, but the existence of a protocol made for distributed filesharing, with no central server. If I publish a specification and anyone can make a client for it, who can you attack? Do you try to stop the specification from being spread, or do you attack the myriad clients that pop up? And if you start filtering by packets, what's to stop someone from releasing a trivial change to the packet format that makes it untraceable again? Unless there is complete control of the internet, where you have to have explicit permission for every media source you access, then stopping this type of system is troublesome. And even Big Brother wasn't that powerful.
Actually, you couldn't really up the mp3 bitrate to make the quality higher for the discerning listener. All comparisons I've heard or seen using encoders to create mp3s at any bitrate above 192 hasn't really been that much better than 192kb/s. It's mostly the law of diminishing returns; beyond that bitrate, you're somewhat maxing out the effectiveness of mp3 quality. I'm sure an encoder that optimized for absolute maximum quality could be made to encode at really high bitrates, but by that point you'd be better off using a different format.
While I'm an advocate of sticking Linux everywhere it can work well (and a few other places, for fun), there's no way I'd pick Linux over Digital Unix on one of those systems, unless there was some monumental task that worked under Linux that just couldn't be ported to Digital Unix. Digital Unix performs amazingly well in things like Apache benchmarks, partially due to their amazingly optimized tcp/ip code. The OS is written completely for the Alpha, and the optimizations shows. It's a work of beauty, don't kick it out just because you like Linux.:)
(I'm not saying Digital Unix is not without faults, I'm just saying it has some great strengths on the Alpha.)
While this book may beat Programming Perl in the examples department, that doesn't surprise me at all. O'Reilly's book covering code examples would be the Perl Cookbook, which does a fairly good job demonstrating how to ping a machine (section 18.7 in the book). Programming Perl is more of a reference for how functions work.
The fact that you can find the IP of a person requesting a file has nothing to do with the fact that it's "truely distributed," as there are very few file sharing networks where there is not a direct connection between the person sharing the file and the person getting it. Unless you want a lot of servers that have massive amounts of traffic, or you have a group of people that don't mind if files are cached on their systems, you'll have to do direct transfers. The other problem with this IP posting strategy is that this in no way reflects people who are actively searching for child pornography. You could log searches for things such as "kiddie porn", but searches are anonymous. Only file connections are direct, and there are some people who go crazy and do things like download every possible jpg image.
I'd be careful there. Shouldn't use the Geek Windex on that there monitor. According to my monitor manual, it contains some non-Geek ammonia that can be harmful to your screen.
Actually, some universities require you to register your mac address in order to get an IP from the university DHCP server. There would be ways around this, but it makes the majority of network traffic traceable to a person. My university (Iowa State) has no such policy, but I know University of Illinois at Urbana-Champaign does.
So, because no one has implemented a way that can be both open source and effective in watching for cheating, it can't be done? That's ludicrous. It's possible someone could have created a cheating proxy before, it's just become easier now that a hacked client can be made instead. In the current situation, if I want to fix a bug in Slade's version, or add on to it, I have to become a member of a group he decides on. This wasn't the intent of Quake's source release, to create little cabals that control their versions and fight for dominance. It's petty and idiotic, and kills the intention of the source release. Security through obscurity doesn't work, at least not forever.
I remember at one point that I enjoyed reading slashdot, for some reason or another. At that point, it had a small but interesting readership, in that the readers were knowledgeable. The articles were fairly interesting, because they often gathered info from sites I didn't normally visit. That day is apparently over. While slashdot may present itself as a "news" site, this is in no way unbiased news. Instead of reporting events as they occur, while maintaining an objective viewpoint, the editors seem to throw up anything vaguely inflammatory in hopes of starting up a big discussion. Were this article objective, someone at id Software would have been contacted, in hopes of getting the facts straight. Instead, this was posted without any sense of perspective. And instead of allowing the readership to decide on their own viewpoint, the person posting the article chose to express their own, in depth. Please, seperate news from editorials. They aren't the same damned thing, and I'm not going to stand idly by while someone attempts to add their comments on to every news item I read. There's a term for what slashdot has become: yellow journalism. It's become a place for news that is created, not that which happens by itself. If all parties were contacted before posting this story to begin with, the entire story might have been "id forgot to include one line in the documentation" instead of what it has become. slashdot's editors really need to start doing fact checking, because by reporting things in this light, it opens *this* site up to lawsuits.
What no one is evaluating is if this information qualifies as valuable. How many casual computer users realize that their web browser sends out information on the browser version among other things? I have never authorized slashdot to collect information on the OS I'm using to access the site, yet I know this information is collected. This "feature" is not easily turned off in most browsers, either.
This comparison isn't completely vaild. While the memory may be an issue, you're comparing an ISA card that has hardware support for wavetable\midi with a PCI card that has no hardware acceleration for these features. The reason you were getting different performance isn't because of a lack of onboard memory, it's because the card doesn't even support hardware acceleration of that media. If you used xmp to compare the AWE64 with a PCI card that was also supported by xmp, and ran them both under xmp to test, then you'd come out with a valid comparison. As it is, you're not even testing using the same mod player, even though xmp has a software mode.
As for the entire AGP thing, the way AGP has been implemented in every card I've seen is that the card still mainly relies on onboard memory while the AGP bus allows for quick transfers from main memory when you've used up the board's memory. The same thing occurs with a PCI video card, except for the fact the PCI bus is slower and has less bandwidth. And as for the comment that there is no honorable future for cards that make the CPU do the work, you're right. The only things I've seen that use this idea tend to sell to the lower-end market. New 3D accelerators are actually taking more work *away* from the processor as they're now doing transform and lighting, a function that was previously left to the CPU.
Please make vaild comparisons, or at least check your facts.
Incorrect. The beta version of slackware that has been available for quite a while (I've been running it for probably close to two months) indeed uses glibc2.1. This fact was posted on slashdot about that long ago, actually.
While there are plenty of legitimate reasons to not allow derivative works, this can only make me wonder if Microsoft is attempting to insulate themselves from the type of thing that they do on an almost daily basis. In the past they've added either "special extensions" or created a proprietary standard out of a commonly used format out of HTML and XML. My first thought when I heard they had introduced a standard was that this would be a prime time to do the same back to them, but they've apparently thought of that.
Slashdot Mirror
I read this news item assuming that it would be informative on what sort of video game violence was being kept from those under the age of 18, or at least would talk about video game violence.
Instead, it takes the point of the hypocrisy of a government.
While I am a fan of games of this genre (not especially SoF, but many others), allow me to point out a flaw in your logic:
When an individual kills something in a game, it's for fun. You can argue whether it's a "killing simulation" (I'd disagree) or merely a game of skill, but either way, you're finding pleasure in it. To compare this with the slaughter of animals for food is completely unfounded. Unless you're talking about some twisted individual who takes joy in killing cows in a meat packing plant, then you're off on your analysis. The people who work in those establishments aren't having fun, or killing for the sake of killing. It's a job.
If you want to analyze whether it desensitizes people or makes them like killing, then post under that topic. This has nothing to do with video game violence.
While a lot of users tend to download mp3s off of Napster a lot of ripping "groups" still exist in forums such as IRC. The use of files that list the results of a hashing algorithm such as md5 on an album, for example, is often distributed with mp3s.
When downloading mp3s off of another person with Napster or gnutella, you're bound to get a bunch of incorrect, incomplete, or otherwise flawed audio files. They may play, but this is the sort of thing bands are using as an argument- this doesn't accurately portray their original work. If widespread distribution of formats like mp3 is going to work, either on a commercial *or* illegal basis, some sort of checksum system needs to be in place.
It's sort of interesting that they're using the four-head Matrox boards to power these things.
:)
While consumers are now seeing boards that have output for two monitors from Matrox, according to a friend of mine, Matrox makes a lot of specialty boards like the one mentioned. Some of the four screen models are used in financial institutions or somesuch.
As for the technology driving it, it's a massive board (or combination of boards) powered by the G200 chipset. Matrox may be making these based on the G400 (or even G450) by now, but I'm not sure.
IBM must be using some sort of tiling scheme to display the stuff. xinerema in hardware?
While Linux-based systems are infiltrating the server world, I question whether they are replacing current commercial unix systems or creating a new niche. While SCO's apparent demise (and possible upcoming resurrection?) may be due to a surge of Linux use, I would doubt it.
More than likely, their business was merely not as profitable anymore in a developing industry. I'd like to see statistics, to know whether people were replacing SCO systems with other commercial unix systems or Linux/*BSD.
It may be a bit pretensious to assume it was Linux pushing SCO out of the market.
Exactly. But the point is, since Rob is so addicted to Diablo II, he's going to play it eventually, so can't he wait and just watch the clip before he plays Diablo II?
For things that aren't a high priority, your OS may not have to do every task. As for the five minutes, how many months is he going to have to wait for somebody to port the Sorenson codec anyway? What if someone were to hand him a VHS tape with a movie trailer? I'd bet it'd take a good five minutes for him to walk over to his TV, turn on the VCR, pop in the tape and get everything going. We're living in a world of convenience, and we want every tool to do every job. It just isn't the most efficient that way because we end up with bloated mismanaged tools.
Perspective.
Iowa State just implemented a proxy server within the last six months, so it wouldn't have been in place when he was there.
The real question here is whether Microsoft is going to keep this technology Windows-only, if they're going to make the minimal showing (a Mac version, possibly even HPUX or Solaris like IE), or if they're going to go full out and either port it to a lot of platforms. Even in that case, it's a toss-up whether they port it themselves, allow others to under NDA (like some Java VMs), or allow the source to the compiler/interpreter to be open enough to allow others to do so.
If they wanted to truly standardize, they could get the language ANSI certified or somesuch. While APIs aren't going to be the same, the language would have a basis, and Microsoft would be able to say they have a language controlled by an independant standards body, while Sun would monopolize Java. This would be sort of similar to the approach they took with their web browser
While this is filed under the "from the what'd-you-expect-from-AOL-and-Microsoft dept.", I'm sort of doubting that AOL and Microsoft are purely to blame for this. Ironically, "the Center for Democracy and Technology" is credited in the press release. But what I'm wondering is this:
Are upcoming specifications that the W3C are going to release public?
Is there a period for public review of upcoming technologies? I would think problems like this, and the flaws pointed out in the article, would have to be addressed. It really sounds from the press release that unless you're a corporation in on the development, your input doesn't count. Should the W3C's drafts have to undergo public review? Or do they already, and I'm missing a step...
When will these people catch on that tracking an IP or username means amazingly little?
If users on Napster had to register a username by giving personal information, then maybe this would be possible. But there's no such correlation.
Tracking by IP, on the other hand, is even more useless. If I'm a modem user and you know my ISP, how the hell does that help? Do you sue the ISP? I think not. They don't really have any control. However, some college campuses exert an insane amount of control over local networks. Most are completely lax, but some (I've heard UIUC does this) actually require you to register your MAC address with the computing center to obtain an IP from DHCP. So if a college ever cooperated with any sort of agency, however unlikely, they could trace you.
This brings up a question: how long will IP addresses not be traceable to a physical location?
Sure, a lot of modem pools will still assign IPs fairly randomly to dialup customers, but static IPs, or at least IPs that seldom change, are increasingly common.
Indeed, there is a proliferation of network operating environments at the moment. To claim that it's a pretty saturated market is accurate as well, but the fact that few people are catching on to is that Inferno is in no way new. I read about it and downloaded the environment to toy with around two years ago. It's made an appearance in several phones with an interface, among other places. This isn't an up-and-coming product, but rather one that is already in place.
Interesting that you've decided to compare a couple authors I have an interest in
I've read a great deal of RAW and he's just as guilty of every "sin" that you claim Adams is. While you could claim Wilson had some grand scheme to put the events in a certain order in works like the Illuminatus Trilogy, you're going to have a hard time convincing me of that. Other than the fact that both authors have the ability to somewhat connect disjointed thoughts (although Wilson's are much more disjointed), you're comparing apples and oranges. Adams has been a writer of humorous British fiction. RAW was American and wrote about a variety of topics, exploring everything from social taboos to the mingling of philosophies to his views on death.
And as for the opiate overdose, no offense, but that's most likely to occur in Wilson's camp, not Adams'.
I'm not completely sure how this amazing protocol Microsoft has cooked up works, but if it's anything like the WMA files that required authorization, it's something like this:
User downloads file. When they attempt to play it, the program connects to the server and grabs the key if you're allowed to have it. You can then play that file freely on your system.
As far as I could tell, it didn't actually unlock the file, per se. I would guess it stores the key in a special little registry section so you can play it later.
So the question becomes this:
How do they keep the user from sending this registry bit to someone else? I would bet it's generated with the system GUID (that global unique identifier key that caused some controversy a while back). It's a fairly common thing for programs to use. So the client sends the server its GUID, the server returns with a key that works to unlock for that user's GUID, and it works seamlessly.
Where they ran into problems before was with programs like Winamp. When using an output plugin that could write to a file, Winamp could get around this restriction by removing the authentication through playback.
If a single application is primarily responsible for the copyright-violating actions, it could possibly be attacked in the way the MPAA has pursued DeCSS. But what if I develop a distributed protocol?
The strength of a system like gnutella isn't the program itself, but the existence of a protocol made for distributed filesharing, with no central server. If I publish a specification and anyone can make a client for it, who can you attack?
Do you try to stop the specification from being spread, or do you attack the myriad clients that pop up? And if you start filtering by packets, what's to stop someone from releasing a trivial change to the packet format that makes it untraceable again?
Unless there is complete control of the internet, where you have to have explicit permission for every media source you access, then stopping this type of system is troublesome. And even Big Brother wasn't that powerful.
Actually, you couldn't really up the mp3 bitrate to make the quality higher for the discerning listener. All comparisons I've heard or seen using encoders to create mp3s at any bitrate above 192 hasn't really been that much better than 192kb/s. It's mostly the law of diminishing returns; beyond that bitrate, you're somewhat maxing out the effectiveness of mp3 quality. I'm sure an encoder that optimized for absolute maximum quality could be made to encode at really high bitrates, but by that point you'd be better off using a different format.
While I'm an advocate of sticking Linux everywhere it can work well (and a few other places, for fun), there's no way I'd pick Linux over Digital Unix on one of those systems, unless there was some monumental task that worked under Linux that just couldn't be ported to Digital Unix. Digital Unix performs amazingly well in things like Apache benchmarks, partially due to their amazingly optimized tcp/ip code. :)
The OS is written completely for the Alpha, and the optimizations shows. It's a work of beauty, don't kick it out just because you like Linux.
(I'm not saying Digital Unix is not without faults, I'm just saying it has some great strengths on the Alpha.)
While this book may beat Programming Perl in the examples department, that doesn't surprise me at all.
O'Reilly's book covering code examples would be the Perl Cookbook, which does a fairly good job demonstrating how to ping a machine (section 18.7 in the book). Programming Perl is more of a reference for how functions work.
The fact that you can find the IP of a person requesting a file has nothing to do with the fact that it's "truely distributed," as there are very few file sharing networks where there is not a direct connection between the person sharing the file and the person getting it. Unless you want a lot of servers that have massive amounts of traffic, or you have a group of people that don't mind if files are cached on their systems, you'll have to do direct transfers.
The other problem with this IP posting strategy is that this in no way reflects people who are actively searching for child pornography. You could log searches for things such as "kiddie porn", but searches are anonymous. Only file connections are direct, and there are some people who go crazy and do things like download every possible jpg image.
I'd be careful there. Shouldn't use the Geek Windex on that there monitor. According to my monitor manual, it contains some non-Geek ammonia that can be harmful to your screen.
Actually, some universities require you to register your mac address in order to get an IP from the university DHCP server. There would be ways around this, but it makes the majority of network traffic traceable to a person. My university (Iowa State) has no such policy, but I know University of Illinois at Urbana-Champaign does.
So, because no one has implemented a way that can be both open source and effective in watching for cheating, it can't be done? That's ludicrous. It's possible someone could have created a cheating proxy before, it's just become easier now that a hacked client can be made instead. In the current situation, if I want to fix a bug in Slade's version, or add on to it, I have to become a member of a group he decides on. This wasn't the intent of Quake's source release, to create little cabals that control their versions and fight for dominance. It's petty and idiotic, and kills the intention of the source release. Security through obscurity doesn't work, at least not forever.
I remember at one point that I enjoyed reading slashdot, for some reason or another. At that point, it had a small but interesting readership, in that the readers were knowledgeable. The articles were fairly interesting, because they often gathered info from sites I didn't normally visit.
That day is apparently over.
While slashdot may present itself as a "news" site, this is in no way unbiased news. Instead of reporting events as they occur, while maintaining an objective viewpoint, the editors seem to throw up anything vaguely inflammatory in hopes of starting up a big discussion. Were this article objective, someone at id Software would have been contacted, in hopes of getting the facts straight. Instead, this was posted without any sense of perspective. And instead of allowing the readership to decide on their own viewpoint, the person posting the article chose to express their own, in depth.
Please, seperate news from editorials. They aren't the same damned thing, and I'm not going to stand idly by while someone attempts to add their comments on to every news item I read.
There's a term for what slashdot has become: yellow journalism. It's become a place for news that is created, not that which happens by itself.
If all parties were contacted before posting this story to begin with, the entire story might have been "id forgot to include one line in the documentation" instead of what it has become.
slashdot's editors really need to start doing fact checking, because by reporting things in this light, it opens *this* site up to lawsuits.
What no one is evaluating is if this information qualifies as valuable. How many casual computer users realize that their web browser sends out information on the browser version among other things?
I have never authorized slashdot to collect information on the OS I'm using to access the site, yet I know this information is collected. This "feature" is not easily turned off in most browsers, either.
This comparison isn't completely vaild. While the memory may be an issue, you're comparing an ISA card that has hardware support for wavetable\midi with a PCI card that has no hardware acceleration for these features. The reason you were getting different performance isn't because of a lack of onboard memory, it's because the card doesn't even support hardware acceleration of that media. If you used xmp to compare the AWE64 with a PCI card that was also supported by xmp, and ran them both under xmp to test, then you'd come out with a valid comparison. As it is, you're not even testing using the same mod player, even though xmp has a software mode.
As for the entire AGP thing, the way AGP has been implemented in every card I've seen is that the card still mainly relies on onboard memory while the AGP bus allows for quick transfers from main memory when you've used up the board's memory. The same thing occurs with a PCI video card, except for the fact the PCI bus is slower and has less bandwidth.
And as for the comment that there is no honorable future for cards that make the CPU do the work, you're right. The only things I've seen that use this idea tend to sell to the lower-end market. New 3D accelerators are actually taking more work *away* from the processor as they're now doing transform and lighting, a function that was previously left to the CPU.
Please make vaild comparisons, or at least check your facts.
The beta version of slackware that has been available for quite a while (I've been running it for probably close to two months) indeed uses glibc2.1. This fact was posted on slashdot about that long ago, actually.
dillinger:~$ ldd /bin/bash libtermcap.so.2 => /lib/libtermcap.so.2 (0x4001a000) libc.so.6 => /lib/libc.so.6 (0x4001e000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
dillinger:~$ ls -l /lib/libc.so.6 lrwxrwxrwx 1 root root 13 Aug 27 07:54 /lib/libc.so.6 -> libc-2.1.1.so*
Looks like 2.1 to me.
While there are plenty of legitimate reasons to not allow derivative works, this can only make me wonder if Microsoft is attempting to insulate themselves from the type of thing that they do on an almost daily basis. In the past they've added either "special extensions" or created a proprietary standard out of a commonly used format out of HTML and XML.
My first thought when I heard they had introduced a standard was that this would be a prime time to do the same back to them, but they've apparently thought of that.