Rhetoric aside, I'd like to see what makes you call RMS a "jackass", what harm he has caused, rather then just talking about it, have some real info. I have not read anything like this myself, so you can imagine I have no idea what you are talking about and why its so flamitory.
Plus I may have both operating systems. I started off by playing the game in windows, but I also have a Linux server and wanted to create a dedicated game server on there. By selling them both seperately I would have gotten the windows version and then downloaded the linux installer. This makes it seem like I had no interest in Linux at the retail stores.
> Of course, contributers could agree to let their contributions be dual-licensed as well, but it strikes me as unfair that you'd get paid by proprietary licensees for those contributions and the contributors get nothing.
The ideal situation is where most of your contributors dont depend upon making money from the contribution, but instead contribute for the benefit of influencing and stabilizing the project. That is where someone needs the software to be patched to work the way they need it to, as oposed to people patching and doing things that are unnecesary at a point in time. If a project makes money on the side, that money will more then likely keep the project alive, at the very least the mainter will see they have made money from this and either reinvest them money in the project or use that money to work more on the project, and incorporate patchs quicker (which a lot of us who do contribute like). Most of the code I have contributed to any project have been license free, that is I literally gave it *free* (license free) to the author so they can incorporate it under their license.
Anyway the purpose of GPL is not that people can't make money from your work, its to keep your work in the realm of free speech. That is users can see and modify your code, you can see what changes they have made to your code, and you can see how your code holds up when used by other people, etc.
> BZZZT Sorry, but unless the contributors specifically assign the copyright to the original author they are still copyright holders. This is one reason why the FSF requires the reassignment (so they can change license versions) and why it has been said closing linux would be impossible (because you could never get all the permisions you need).
It really depends on what you mean by "contribution". How big of a contribution. A bug fix is one form of a contribution, it is usually contributed in the form of a patch. In most cases people who contribute small things don't place their own copyrights on what they contribute, and no license to them. Contributing a driver for example is a big contribution, contributing a number of functions is also a big contribution, at least its easier to copyright. Copyrighting an simple algorithm is really pushing it, that is like trying to copyright a picture of a rectangle.
> And probably trademarked the name, too.
Well I am sure many people would not like it if another OS named itself Linux.
> "OpenRacer" is even dumber than "Tux Racer" and sounds even less like a kart game.
Well why can't the name be more creative?
> OpenRacer has so much intangible distance to make up, and this is still all because an open project closed.
Yes but who took it that distance?
> And the licensing/ownership does NOT change just because the forked version has made "enough" changes.
BS, if I as a single developer, took the code and used it as a starting base, but in the end recoded everything and removed all my dependancies on the original code, that is made huge enough changes, then ownership would change. This is about limitations in copyright law, not limitations in the (L)GPL. This is less likely to happen with something as large and complex as Linux, but in smaller less complex projects it should not be.
As long as the copyright notice still lists the owner(s) and your contribution does not add your name to its ownership it doesnt matter.
> If not, well, I know I'd be pissed to hell and back if I helped develop something and they stole the one reason for it existing, namely having a game to tweak/play with at the code level.
There still should be a game to tweak/play with at the code level. They can take the previous GPL versions off of their site, *but* they can not force any one to get rid of it, or not use it on their own. The previous GPL versions should be still free to use, modify, and even to create alternative/derivative versions of it on your own, but only as GPL, at least if you make a large enough changes ownership should change, and you should be able to do some of the same things.
This is the nth time I have heard someone use the word "harm" in describing RMS on OS, but they dont go into detail of this "harm". To me it was started by someone else who probably knew what they were talking about and then parrotted by those who don't.
I dont think you get black listed for *using* non-free software, you might *making* non-free software, but I dont even think he has a black list of people anyway, I believe he does have a black list of licenses. Also I dont see this "my-way-or-the-highway" bit either. He didnt say anything bad when the makers of Ogg Vorbis switched to BSD from LGPL, because they wanted hardware manufacturers to create players for Ogg Vorbis and they felt LGPL was to restrictive for this, I believe he said this was okay. I think he wants people to try creating free software and developing a business model (for survival) from it, but he is not asking people to commit business suicide, or to starve to death for the purpose of free software. Free software should be the ultimate (utopian kin of) goal of all software engineers, it is the ultimate feature you could possibly offer to your users, but in these times practicality benefits non-free, but things change over time as technology and society change over time.
I am not to sure about this. You only give one example and that is for illiteracy. In my school where that was a smaller problem, computers went unused except for once in a while, because they didnt have any good software. We had macs, they were used by science classes once in a while and only to use some funky software on them, and once for a chemistry test I had. We had apple2s and they had classes teaching basic things like word processors, etc, and were also used in the computer programming class, I took that class but it didnt last 1 semester because most of the students didnt care about it, I learned all my programming out side of school on my own from books I was even using C++ when I took that class, from my own computer. I would think more people would be interested in computers if they had them at home, since I had one at home I was interested. At the time I thought it would be cool if we had more computers, but when I look back at it now, they could have been using any computer they wanted and it would have made little diffrence except for the rare person like me and one other student who learned programming on his own outside of school. I mean this was 6 or so years ago in high school mostly.
Well there are 2 diffrent managements, automanaged where things snap to each others borders this can be anoying, and manual management. But they both dont account for zooming into an image. Like say if I zoom into an image its still nice to see more of the image rather then the image either disappear behind dialogs or be confined to a small portion of the screen.
I agree desktop transparency is not entirely useful, although as someone showed DirectFB displaying GIMP, its nice to be able to see the image you are working on through any dialogs you have open. What I think may also be interested in general in the hardware accelerated desktop, is to be able to scale and rotate windows around, this way you have more options about having multiple windows open. Or small stuff like, if you are looking at an upside down image on a website you can flip the browser upside down rather then sitting on your head.
In the visual candy area (not much better then transparency) is to add some type of physics to windows, so if you yank it in a direction it swings, or if you drop it, it bounces, if you try to push it off the desktop it squishes down instead, or some weird stuff like that.
My priority first is resolution/frequency(framerate), then when a video card surpasses my monitors abilities for resolution, then use anti-aliasing, when it surpasses my video frequency, then use motion blur. There is little to no reason to use those techniques if your slowing things down or making them pixelated and blurry.
For more realism, I'd love to see someone incorporate realtime distributed raytraced (soft) shadows (both soft specular and inter-diffuse)reflections (soft) transmissions and caustics. But right now those things take hours to render in software.
You should read the posts point count types, not the *last scored type*. As of right now she has only 1 troll, 1 insightful, and 5 funny, and 1 overrated, for a total of 8. So why waste time focusing on the 1 troll? I mean do you expect someone to get a perfect score or something? And 1 troll by far does not represent "everyone here". So just sit back, relax, and simmer down to a slow boil, before you explode.
Also I dont see the 'religious' few Microsoft users preventing Microsoft from being "a major desktop OS". And Linux is not about how bad Microsoft is, its more about a business model that allows freedom. You want to really know what makes Microsoft bad, its not linux, its Code Red, Nimda, etc, and these are not necesarily microsoft incompetence, more so then probability. The probability that the more simple minded users you attract the less respect they have for the under lying technology. IMO most of the people who waste their time hating microsoft, are microsoft users, not linux users.
I agree. If we act diffrently in any way after the terrorists attack we risk giving incentives to those who would like us to act diffrently, incentives that would lead them to provoke or do terrorists attacks themselves. So one of the risks is if we change laws and take away civil liberties we risk telling our government you can have more power while terrorists attack us. Think about it, what message does it say, when most people didnt like the president, what message does it send to the president that his aproval rating sky rocket after terrorists attack?
I hope not renderman. I'm finding that renderman shaders give to much freedom to the shader writer that it makes it hard to create a unified environment for realistic rendering such as inter-diffuse reflections and caustics. Special shaders have to be created in order to do this, but its hard to unify them to your settings for example in distributed ray tracing, bi-directional path tracing and photon mapping. A shader for example in Renderman simply returns color and transparency but with distributed ray tracing path tracing and photon mapping (advanced rendering techniques) you have to know the probability of a photon or ray being absorbed (transmitted) or reflected (speculary or diffusely) or the shader would at least have a second function which bounces photons or rays off it, in order to get high quality graphics with the least amount of calculations. This also has its issues because some shaders could never work in a realistic way, like cel shaders for non-photorealistic rendering, but these can be implemented as secondary shaders that sit on top of realistic shaders, so you do your calculations in a realistic scene then convert that data to look nonphotorealistic on a secondary shader/filter, but this again only works with a few shaders, a number of shaders are just not realistic at all. Phong highlights (dont confuse with phong shading) are another example of a shader not being real, for realism you would need a slightly rough surface which causes the reflection of a light source to appear soft and blurred, phong highlights are a fake approximation of this effect which will some day be replaced by real reflections of rough surfaces.
Obviously I am looking far into the future when real time graphics are going to start looking at real time photo-realistic lighting. Hopefully some day these kinds of things...
http://www.3dluvr.com/marcosss/
http://jackie.sf.net/
...will be done in real time, but at those times there will be problems as the ones I describe above with shaders.
The point of having extensions is so that it isnt so urgent. If they standardise it to quickly they risk making a mistake and have to back pedal and make changes like Microsoft has had to do on a number of occasions with Direct3D.
I dont know what he meant by hypocrisy, but I know that Microsoft has not only stimulated (as the previous poster notes) but also taken advantage of a competitive PC market. While its own market practice is exactly the same as IBMs was for a proprietary system, before their BIOS was reversed engineered and a competitive PC industry flourished. Imagine if microsoft's windows was reversed engineered and information contained in it was passed through a lawyer and then passed to a programmer working on an alternate windows operating system. I dont think microsoft would sit back and let that happen, even the Open Source projects that attempt to create an open alternative windows operating system say they are not going to reverse engineer nor use information from someone who has reversed engineered windows.
Well to draw a little more of the connection here, the pricing was lowered due to competition, the competition came from the IBM PC being opened by reverse engineering. Marketing is also a part of competition, if you have competitors you will want to out sell them, so you do things like marketing and lowering prices. The operating system was an equally important factor, but it is not more important then the platform. With out the IBM PC, Microsoft may not have never bought the OS and licensed it to IBM, and it would have taken a long time before an open platform was realized, its even conceivable that we would not be as advanced as we currently are in computer technology. The PC combined with Microsofts OSs have done a lot, but now it appears that microsoft is now the hold up, the IBM of the 21st century. Most people dont even try to reverse engineer windows, rather projects that have tried to create a Windows compatible OS are to afraid of the legal mess of using information that comes from reverse engineering. So most attempts at doing this are pretty weak, and focus primarly on the upper parts of windows, on the API levels.
Maybe it is for transparent materials for composites. Think about like a blue screen or green screen with a translucent white/grey piece of paper. The paper is going to have some blue (or green) in it, so you would want to reduce the blue more (not the over all color). Because you will get a white-blue reducing it down with a single scalar (alpha) would produce grey-blue or dark-gray-blue, instead you want to reduce the blue down more then the others so you have to use 2 or 3 scalars (red-alpha, green-alpha, blue-alpha).
Actually now that I think about it, it would be a very interesting way to have a sort of wireless keyboard, its a dummy keyboard that has a tracker attacked to it (no circutry for the keys needed so it can be lighter), and since your fingers are also tracked you can hit keys of the physical keyboard and your virtual fingers will be detected as colliding and pushing the virtual keyboard's keys. Hmmm, still I think this would be a more expensive solution.
I was thinking the most inexpensive thing to do is to let the user see the keyboard by simply looking down. It would be nice if you had 2 gloves and the keyboards position was tracked, then you can have 3d hands and a 3d keyboard to look at that is in the same position as your real hands and keyboard, but this sounds like a more expensive solution to me.
Dont forget computers will also be (have to be) more intelligent, in order to accomplish more complicated tasks. But the intelligence has to keep a track of the user (and has the gain the users trust), so that it can configure itself to best server the user. If the user does a lot of graphics, then it will optimize the hardware and software (including the interfaces) to meet those needs. Under these circumstances borrowing someone elses box will be almost impossible unless the user is able to carry data about their preferences around and inject them into the computer they wish to use, but that computer would still not be as optimal. The functions of a brain are similar in this reguard as well, its about deeper integration, higher productivity in the important areas. Having a common interface is not bad, as there would need to be a common interface for people to learn how to use a computer, and for the computer to learn about the user.
I agree with that, 3d would only be good if it extended your 2d desktop 360 degrees with motion captured 3d mouse/glove and HMD, with enough eye space so you can see the keyboard.
From there mixing actual 3d with 2d-floating-tablets, so you can grab a tablet/window and 3d objects can pop out of them (sort of like how holograms are used in sci-fi movies).
The future of interfaces will be controlled by the user. Not all users work best in the same way. Sure you can spend money researching to find the common interface that everyone is average in their productivity, but in the end the productivity is with in the users themselves, and the interface that works best for them. So the future of computers in general is adaptable interfaces.
Slashdot Mirror
Rhetoric aside, I'd like to see what makes you call RMS a "jackass", what harm he has caused, rather then just talking about it, have some real info. I have not read anything like this myself, so you can imagine I have no idea what you are talking about and why its so flamitory.
Plus I may have both operating systems. I started off by playing the game in windows, but I also have a Linux server and wanted to create a dedicated game server on there. By selling them both seperately I would have gotten the windows version and then downloaded the linux installer. This makes it seem like I had no interest in Linux at the retail stores.
> Of course, contributers could agree to let their contributions be dual-licensed as well, but it strikes me as unfair that you'd get paid by proprietary licensees for those contributions and the contributors get nothing.
The ideal situation is where most of your contributors dont depend upon making money from the contribution, but instead contribute for the benefit of influencing and stabilizing the project. That is where someone needs the software to be patched to work the way they need it to, as oposed to people patching and doing things that are unnecesary at a point in time. If a project makes money on the side, that money will more then likely keep the project alive, at the very least the mainter will see they have made money from this and either reinvest them money in the project or use that money to work more on the project, and incorporate patchs quicker (which a lot of us who do contribute like). Most of the code I have contributed to any project have been license free, that is I literally gave it *free* (license free) to the author so they can incorporate it under their license.
Anyway the purpose of GPL is not that people can't make money from your work, its to keep your work in the realm of free speech. That is users can see and modify your code, you can see what changes they have made to your code, and you can see how your code holds up when used by other people, etc.
> BZZZT Sorry, but unless the contributors specifically assign the copyright to the original author they are still copyright holders. This is one reason why the FSF requires the reassignment (so they can change license versions) and why it has been said closing linux would be impossible (because you could never get all the permisions you need).
It really depends on what you mean by "contribution". How big of a contribution. A bug fix is one form of a contribution, it is usually contributed in the form of a patch. In most cases people who contribute small things don't place their own copyrights on what they contribute, and no license to them. Contributing a driver for example is a big contribution, contributing a number of functions is also a big contribution, at least its easier to copyright. Copyrighting an simple algorithm is really pushing it, that is like trying to copyright a picture of a rectangle.
> And probably trademarked the name, too.
Well I am sure many people would not like it if another OS named itself Linux.
> "OpenRacer" is even dumber than "Tux Racer" and sounds even less like a kart game.
Well why can't the name be more creative?
> OpenRacer has so much intangible distance to make up, and this is still all because an open project closed.
Yes but who took it that distance?
> And the licensing/ownership does NOT change just because the forked version has made "enough" changes.
BS, if I as a single developer, took the code and used it as a starting base, but in the end recoded everything and removed all my dependancies on the original code, that is made huge enough changes, then ownership would change. This is about limitations in copyright law, not limitations in the (L)GPL. This is less likely to happen with something as large and complex as Linux, but in smaller less complex projects it should not be.
As long as the copyright notice still lists the owner(s) and your contribution does not add your name to its ownership it doesnt matter.
> If not, well, I know I'd be pissed to hell and back if I helped develop something and they stole the one reason for it existing, namely having a game to tweak/play with at the code level.
There still should be a game to tweak/play with at the code level. They can take the previous GPL versions off of their site, *but* they can not force any one to get rid of it, or not use it on their own. The previous GPL versions should be still free to use, modify, and even to create alternative/derivative versions of it on your own, but only as GPL, at least if you make a large enough changes ownership should change, and you should be able to do some of the same things.
What harm?
This is the nth time I have heard someone use the word "harm" in describing RMS on OS, but they dont go into detail of this "harm". To me it was started by someone else who probably knew what they were talking about and then parrotted by those who don't.
-
> It wasn't that RMS saw the points and then still disagreed. He simply could not be bothered to listen to a mere mortal.
Either that or he wasnt in the mood to talk about it, and your "smart friend" was anoying him at dinner.
I dont think you get black listed for *using* non-free software, you might *making* non-free software, but I dont even think he has a black list of people anyway, I believe he does have a black list of licenses. Also I dont see this "my-way-or-the-highway" bit either. He didnt say anything bad when the makers of Ogg Vorbis switched to BSD from LGPL, because they wanted hardware manufacturers to create players for Ogg Vorbis and they felt LGPL was to restrictive for this, I believe he said this was okay. I think he wants people to try creating free software and developing a business model (for survival) from it, but he is not asking people to commit business suicide, or to starve to death for the purpose of free software. Free software should be the ultimate (utopian kin of) goal of all software engineers, it is the ultimate feature you could possibly offer to your users, but in these times practicality benefits non-free, but things change over time as technology and society change over time.
Also I wanted to add, that we had tried to create a computer club and that failed as well, there just was not enough people interested.
I am not to sure about this. You only give one example and that is for illiteracy. In my school where that was a smaller problem, computers went unused except for once in a while, because they didnt have any good software. We had macs, they were used by science classes once in a while and only to use some funky software on them, and once for a chemistry test I had. We had apple2s and they had classes teaching basic things like word processors, etc, and were also used in the computer programming class, I took that class but it didnt last 1 semester because most of the students didnt care about it, I learned all my programming out side of school on my own from books I was even using C++ when I took that class, from my own computer. I would think more people would be interested in computers if they had them at home, since I had one at home I was interested. At the time I thought it would be cool if we had more computers, but when I look back at it now, they could have been using any computer they wanted and it would have made little diffrence except for the rare person like me and one other student who learned programming on his own outside of school. I mean this was 6 or so years ago in high school mostly.
Well there are 2 diffrent managements, automanaged where things snap to each others borders this can be anoying, and manual management. But they both dont account for zooming into an image. Like say if I zoom into an image its still nice to see more of the image rather then the image either disappear behind dialogs or be confined to a small portion of the screen.
I agree desktop transparency is not entirely useful, although as someone showed DirectFB displaying GIMP, its nice to be able to see the image you are working on through any dialogs you have open. What I think may also be interested in general in the hardware accelerated desktop, is to be able to scale and rotate windows around, this way you have more options about having multiple windows open. Or small stuff like, if you are looking at an upside down image on a website you can flip the browser upside down rather then sitting on your head.
In the visual candy area (not much better then transparency) is to add some type of physics to windows, so if you yank it in a direction it swings, or if you drop it, it bounces, if you try to push it off the desktop it squishes down instead, or some weird stuff like that.
My priority first is resolution/frequency(framerate), then when a video card surpasses my monitors abilities for resolution, then use anti-aliasing, when it surpasses my video frequency, then use motion blur. There is little to no reason to use those techniques if your slowing things down or making them pixelated and blurry.
For more realism, I'd love to see someone incorporate realtime distributed raytraced (soft) shadows (both soft specular and inter-diffuse)reflections (soft) transmissions and caustics. But right now those things take hours to render in software.
You should read the posts point count types, not the *last scored type*. As of right now she has only 1 troll, 1 insightful, and 5 funny, and 1 overrated, for a total of 8. So why waste time focusing on the 1 troll? I mean do you expect someone to get a perfect score or something? And 1 troll by far does not represent "everyone here". So just sit back, relax, and simmer down to a slow boil, before you explode.
Also I dont see the 'religious' few Microsoft users preventing Microsoft from being "a major desktop OS". And Linux is not about how bad Microsoft is, its more about a business model that allows freedom. You want to really know what makes Microsoft bad, its not linux, its Code Red, Nimda, etc, and these are not necesarily microsoft incompetence, more so then probability. The probability that the more simple minded users you attract the less respect they have for the under lying technology. IMO most of the people who waste their time hating microsoft, are microsoft users, not linux users.
I agree. If we act diffrently in any way after the terrorists attack we risk giving incentives to those who would like us to act diffrently, incentives that would lead them to provoke or do terrorists attacks themselves. So one of the risks is if we change laws and take away civil liberties we risk telling our government you can have more power while terrorists attack us. Think about it, what message does it say, when most people didnt like the president, what message does it send to the president that his aproval rating sky rocket after terrorists attack?
I hope not renderman. I'm finding that renderman shaders give to much freedom to the shader writer that it makes it hard to create a unified environment for realistic rendering such as inter-diffuse reflections and caustics. Special shaders have to be created in order to do this, but its hard to unify them to your settings for example in distributed ray tracing, bi-directional path tracing and photon mapping. A shader for example in Renderman simply returns color and transparency but with distributed ray tracing path tracing and photon mapping (advanced rendering techniques) you have to know the probability of a photon or ray being absorbed (transmitted) or reflected (speculary or diffusely) or the shader would at least have a second function which bounces photons or rays off it, in order to get high quality graphics with the least amount of calculations. This also has its issues because some shaders could never work in a realistic way, like cel shaders for non-photorealistic rendering, but these can be implemented as secondary shaders that sit on top of realistic shaders, so you do your calculations in a realistic scene then convert that data to look nonphotorealistic on a secondary shader/filter, but this again only works with a few shaders, a number of shaders are just not realistic at all. Phong highlights (dont confuse with phong shading) are another example of a shader not being real, for realism you would need a slightly rough surface which causes the reflection of a light source to appear soft and blurred, phong highlights are a fake approximation of this effect which will some day be replaced by real reflections of rough surfaces.
Obviously I am looking far into the future when real time graphics are going to start looking at real time photo-realistic lighting. Hopefully some day these kinds of things...
http://www.3dluvr.com/marcosss/
http://jackie.sf.net/
...will be done in real time, but at those times there will be problems as the ones I describe above with shaders.
The point of having extensions is so that it isnt so urgent. If they standardise it to quickly they risk making a mistake and have to back pedal and make changes like Microsoft has had to do on a number of occasions with Direct3D.
I dont know what he meant by hypocrisy, but I know that Microsoft has not only stimulated (as the previous poster notes) but also taken advantage of a competitive PC market. While its own market practice is exactly the same as IBMs was for a proprietary system, before their BIOS was reversed engineered and a competitive PC industry flourished. Imagine if microsoft's windows was reversed engineered and information contained in it was passed through a lawyer and then passed to a programmer working on an alternate windows operating system. I dont think microsoft would sit back and let that happen, even the Open Source projects that attempt to create an open alternative windows operating system say they are not going to reverse engineer nor use information from someone who has reversed engineered windows.
Well to draw a little more of the connection here, the pricing was lowered due to competition, the competition came from the IBM PC being opened by reverse engineering. Marketing is also a part of competition, if you have competitors you will want to out sell them, so you do things like marketing and lowering prices. The operating system was an equally important factor, but it is not more important then the platform. With out the IBM PC, Microsoft may not have never bought the OS and licensed it to IBM, and it would have taken a long time before an open platform was realized, its even conceivable that we would not be as advanced as we currently are in computer technology. The PC combined with Microsofts OSs have done a lot, but now it appears that microsoft is now the hold up, the IBM of the 21st century. Most people dont even try to reverse engineer windows, rather projects that have tried to create a Windows compatible OS are to afraid of the legal mess of using information that comes from reverse engineering. So most attempts at doing this are pretty weak, and focus primarly on the upper parts of windows, on the API levels.
Maybe it is for transparent materials for composites. Think about like a blue screen or green screen with a translucent white/grey piece of paper. The paper is going to have some blue (or green) in it, so you would want to reduce the blue more (not the over all color). Because you will get a white-blue reducing it down with a single scalar (alpha) would produce grey-blue or dark-gray-blue, instead you want to reduce the blue down more then the others so you have to use 2 or 3 scalars (red-alpha, green-alpha, blue-alpha).
Actually now that I think about it, it would be a very interesting way to have a sort of wireless keyboard, its a dummy keyboard that has a tracker attacked to it (no circutry for the keys needed so it can be lighter), and since your fingers are also tracked you can hit keys of the physical keyboard and your virtual fingers will be detected as colliding and pushing the virtual keyboard's keys. Hmmm, still I think this would be a more expensive solution.
I was thinking the most inexpensive thing to do is to let the user see the keyboard by simply looking down. It would be nice if you had 2 gloves and the keyboards position was tracked, then you can have 3d hands and a 3d keyboard to look at that is in the same position as your real hands and keyboard, but this sounds like a more expensive solution to me.
Dont forget computers will also be (have to be) more intelligent, in order to accomplish more complicated tasks. But the intelligence has to keep a track of the user (and has the gain the users trust), so that it can configure itself to best server the user. If the user does a lot of graphics, then it will optimize the hardware and software (including the interfaces) to meet those needs. Under these circumstances borrowing someone elses box will be almost impossible unless the user is able to carry data about their preferences around and inject them into the computer they wish to use, but that computer would still not be as optimal. The functions of a brain are similar in this reguard as well, its about deeper integration, higher productivity in the important areas. Having a common interface is not bad, as there would need to be a common interface for people to learn how to use a computer, and for the computer to learn about the user.
I agree with that, 3d would only be good if it extended your 2d desktop 360 degrees with motion captured 3d mouse/glove and HMD, with enough eye space so you can see the keyboard.
From there mixing actual 3d with 2d-floating-tablets, so you can grab a tablet/window and 3d objects can pop out of them (sort of like how holograms are used in sci-fi movies).
The future of interfaces will be controlled by the user. Not all users work best in the same way. Sure you can spend money researching to find the common interface that everyone is average in their productivity, but in the end the productivity is with in the users themselves, and the interface that works best for them. So the future of computers in general is adaptable interfaces.