What I meant was that if it went to court, the defendant could point out that you lied in your initial correspondance with him, and this could assist in turning the judge's opinion against you.
Honest. The first MP3 I ever downloaded was purchased from BLEEP (mentioned on Slashdot 2 days ago). I actually bought over $15 worth of tracks and I bought another $9 last night. I am going to wait for the credit card statement to make sure the charge is correct before I buy any more.
No DRM + previews + knowledge that the download will really be what you want: made it easily worth it (though if a similar non-DRM site for.99 shows up I would probably abandon BLEEP immediately).
No. You cannot force them to release the code, therefore stating that is going to immediately get you in trouble for lying in your first communication.
The only thing you can force them to do is stop distributing their version until they either remove the code from it or they comply with the GPL by releasing their code.
You also can sue them for copyright violation. I have heard you almost have to do this for money, not for any other remedy. I certainly don't think there is any case where the remedy for a copyright violation is that the guilty party is forced to give up all copyrights to a piece of their own work, especially if their work is significantly larger than the violated work.
So do not say anything stupid about them being forced to release their code. That is false, no matter how much Billy at Microsoft wants people to think otherwise.
Linus is confused. There is no reason to seperate 2D and 3D. If all the verticies in your 3D object have z=0 and you look at it flat on the screen, you have 2D. Thus it is trivial to make 3D accelerated hardware accelerate 2D.
So no matter where the 3D and 2D reside (kernel, user-space, with the ability to remotly connect or not...) they will reside in the same place in any intelligently designed system.
This was implemented over 15 years ago in IrisGL. These machines ran at about 5Mhz and were perfectly capable of using OpenGL to update a window system just as good as X does today. So speed has not ever been an issue.
Most of the problem is that OpenGL (as opposed to IrisGL) lacks some operations. There is no good handling of fonts and text. There is no easy way to create a complex clip region. There is no efficient way to take an image in local memory and create a rectangle texture-mapped with it (everything is designed to reuse textures, not do one-off copies). You also have to negotiate an ugly and complex glX (or WGL on Windows) interface before you can start drawing with OpenGL. Since these are things that 2D interfaces such as Xlib or GDI32 do, people are forced to use that for 2D, and then tend to segregate their programs into a "window" in which 3D is displayed, surrounded by 2D. It also does not help that Windows, and until recently X, had stupid design decisions so it is impossible to draw with both OpenGL and 2D into the same window, thus enforcing this segregation at the API level.
I would like to see a design where the necessary missing drawing functions are added to OpenGL (perhaps copy Cairo). And then make a library to replace xlib where you create a window and you immediately have a current OpenGL context that you can draw. And do the Keith Packard thing with compositing managers: the "window" can in fact be a texture map and this is then mapped to the screen through arbitrary transformation and shading, to produce Quartz-like effects.
If one was to win, I would have to bet on Gnome now. However as little as 6 months ago I would have said KDE.
What does appear to be happening is that this competition is figuring out a basic layer of functionality, far simpler than either KDE or Gnome, that is really necessary for interoperability. Mostly the freedesktop.org stuff like standards for drag & drop, and "window manager standards" which are really kludgy methods of storing process information in the X server so other programs can see it.
This could be a very good result, as it could cause "which desktop" to become as irrelevant to the programs that are running as which text editor you use is relevant to a C compiler, and make it as easy to experiment with a new desktop as it is to experiment with a new text editor.
The big problem is some of the basic underlying stuff, especially the COM-like stuff, is per-desktop. The only way it is going to be unified is if one of the projects abandons it's lower-level implementation and starts using the other one. This is going to suck for the engineers working on the abandoned one, but it has to happen.
Damn, you are right! It even works in tcsh. I have no idea why I thought you had to type "start" before the filename. In fact in cmd.exe "start" will hide error messages and otherwise screw up, while just the filename seems to work quite well.
I certainly thought this is how it *should* work but for some reason I got convinced that the "start" was necessary. Possibly I tried it once and it did not work.
Look at the shadows! They are all pointing in the wrong direction! And the reflections on his head don't line up with the overhead lights! And if you analyze the echos of the applause, you can tell that the room it was recorded in has wood panelling and not the plaster that you see on tv! All of this proves the speech was faked!
Absolutely agree. The name is backwards, and I can see no possible reason to do so. The reversed name sounds equally good and neither makes one system sound better than the other. I can't see any plausable insidious plot by Microsoft with the backwards name.
Well maybe: they want to make it really difficult for a future Microsoft to release Windows emulation software for Linux, by taking up the obvious name. Ie by using this name they are trying to prevent business decisions in the future that they don't like now?
Nah, that is too far fetched. I think it is just somebody in marketing being stupid.
Minimizing to an icon on the desktop is how the very first X window managers worked. I also think it is how things earlier than that worked, such as XViews and Andrew. So this is a very old idea. Updating the icon was definately supported in the first versions of X with window managers, it was not easy, but they definatley thought that programs would want to change their icons to show status.
I actually think the biggest innovation to come out of Microsoft is the "toolbar" where the icons exist even if the window is open. Yes there were a lot of earlier schemes for arranging icons in windows or grids but all of them had the idea that the icon was only visible when the window was not visible. This meant that to find a window you had to look both for the icon and for the already-opened (but perhaps buried) window. Trust me this is an innovation and I do think Microsoft thought of it. And it is pretty obvious that all new desktop environments are copying this, even OSX (which shows an icon in the dock for all running applications, whether they are open or not).
So in a lot of ways I feel that anything that reduces a window to a desktop icon is a throwback to an idea that seems to have been discredited.
You don't have to use Windows+R, you can run command.exe and type the commands in there, which is a lot faster, as long as you are able to click on the terminal window.
No, you have to type "open " before the name of the document.
Windows has this and calls it "start". At least tcsh on Windows does, I eventually learned that the intuitive command in command.exe is "rundll32.exe url.dll,FileProtocolHandler ", I would guess that a shortcut like the initial poster described could be used to make "start" work there.
Of course Linux is far behind either, as there is *no* command that duplicates what a double-click does in a file browser. This is really bad! The KDE and Gnome people should both be ashamed. Though in some ways I think they fear that they will make it too easy to write file browsers, and their Microsoft-like control over the Linux desktop would be eroded, so they don't provide something that would seem obvious to any Unix user.
While the mainland, especially Germany, has been big on OSS, it seems the UK is more of a Windows shop than the USA is, and they appear to be targeting the UK. Of course they may not have anybody at SCO who know any languages other than english...
What happened to "Atheos" (it was called something like that). It was also a one-person effort to make a Unix-like system designed for the desktop, with integrated GUI. A few years ago that sounded very interesting, but nothing ever came of it. Not a good precedent for this project (unless this is the same project, it is hard to tell).
But I think one thing that killed Atheos is the same thing that killed almost any alternative to X: inability to support any modern graphics cards at any resolution higher than VESA. Unfortunatley this information is locked up in X drivers that are so tightly integrated with internal complexities of X that it is impossible to extract and reuse it, despite the open source nature. Perhaps XDrive will help here by making the driver interface cleaner. In any case this project sounds like it has some hardware acceleration, so maybe they will escape this trap.
Personally I am not thrilled with putting widgets into the OS. My feeling is that this locks GUI design and innovation. I would prefer a design where there were powerful graphics and event handling calls, so it is easy to write a widget, but the interface is designed so that it is obvious that you can write different widgets that have not yet been invented.
Installing or upgrading Windows will always hose your boot sector so that you lose Linux. Recent versions will try really hard to repartition your drive or reformat the unknown sectors, though I have been told you can stop this by hitting No enough times. So Microsoft has already released their version of this.
A battery (even a rechargable one) has negative energy costs too. It takes energy to manufacture them, plus all the energy that you get out has to be put in.
From what I have heard they are using up about 1/4 or so of the disk to store WMA equivalents of the CD tracks. It would seem extremely sensible to put an autorun program on there that reads the CD tracks and creates the WMA (or something equivalent) files, and also does whatever messing they want with the poor user's Windows box.
Contrary to the other posters here, I don't think this would be any more or any less evil. It still won't work on Linux, but the same old workarounds will work on both Linux and Windows.
And they can put 10 more minutes (or whatever amount) more music on the copy-prevented disk. And there would be more assurance that you can't miss any music when you play the disk in the computer, that worry is probably more of an incentive to work around this than a desire to make an MP3.
Because such actions reinforce stupid decisions by legal departments such as yours.
If it becomes standard that you cannot buy or use code without "indemnity" then you have made it impossible for an independent code writer to write a program and have people use it. This is irrelevant to open source, what you have pretty much said is that writing code is the realm of big business and that start-ups and free enterprise and just the desire to tell people about solutions you have figured out have become illegal.
This goes far beyond Linux. Linux is now being protected by big money interests in the same way Windows is. But the next genius grad student who writes the "Linux-killer" operating system is going to be out of luck, as everybody who he shows it to is going to say "but you have no idemnity, so I'm going to have to stick with Linux..."
Even people who think Linux is the end-all of software should realize that this means the death of applications for Linux from independent authors, since they cannot "idemnify" their code. And Windows fans should realize this means the death of all the Shareware and Freeware and open source software, and also the few remaining tiny software vendors, all of who cannot afford "idemnification" either.
I don't care if this action causes SCO to go bankrupt tomorrow, this is a bad precedent for Novell to do this.
IMHO the damning evidence is the scanf strings, though I would really need to look at the source code and the documentation for how DVD menu information is stored, to see if there is another reason. My guess is that anybody writing this themselves could come up with different scanf strings, or multiple ones, or use something other than scanf to accomplish the same thing. But then again, if the documentation for the menu system on DVD's says "use a scanf string like this to parse this field" then it is pretty obvious why both programs use the same strings.
Obviously the list of names means they looked at the MPlayer code, but I think this can be excused: engineer A says to engineer B: "Hey we need to implement this list of stuff" and then cut & pastes the string list from the MPlayer source. I think that is fair use of the code. And of course the resulting list would end up in the same order. Notice that they changed the capitalization and changed "dunnowhat" to "unknown".
KISS does not have to open their source code. Even if they violated the GPL, all they have to do is remove that code and replace it. MPlayer could sue for copyright damages but I don't think they would. If in fact KISS did not steal the source code, they could probably prove it reasonably by revealing a copyrighted chunk of their code that can be easily seen to be unrelated, despite the fact that it produces the same strings.
Part of the problem is that it is not default, so an application cannot assumme it.
I'm using KDE and I have it set this way, but there are still problems. First of all you can't move or resize the window without raising it, they really should restore the old FVWM behavior where if you click it raises, but if you drag it just moves/resizes.
More of a concern is that KDE has a bug that also exists on Windows and makes even using child windows to enforce stacking impossible. This is that if you raise a "child" window it will also raise it's parent to just below it. This makes overlapping windows with floating toolbars impossible. Most window managers that obey TRANSIENT_FOR have this bug, including old ones. I'm not sure why but perhaps they are all copying the same code base.
Changing the user interface automatically has been proven to be bad and users do not like it.
However as other posters have pointed out, you can certainly add actions the user can do to cause the interface to change. Since the user actively does these things there is no confusion when the interface changes (well, in theory...)
Slashdot Mirror
What I meant was that if it went to court, the defendant could point out that you lied in your initial correspondance with him, and this could assist in turning the judge's opinion against you.
Honest. The first MP3 I ever downloaded was purchased from BLEEP (mentioned on Slashdot 2 days ago). I actually bought over $15 worth of tracks and I bought another $9 last night. I am going to wait for the credit card statement to make sure the charge is correct before I buy any more.
.99 shows up I would probably abandon BLEEP immediately).
No DRM + previews + knowledge that the download will really be what you want: made it easily worth it (though if a similar non-DRM site for
No. You cannot force them to release the code, therefore stating that is going to immediately get you in trouble for lying in your first communication.
The only thing you can force them to do is stop distributing their version until they either remove the code from it or they comply with the GPL by releasing their code.
You also can sue them for copyright violation. I have heard you almost have to do this for money, not for any other remedy. I certainly don't think there is any case where the remedy for a copyright violation is that the guilty party is forced to give up all copyrights to a piece of their own work, especially if their work is significantly larger than the violated work.
So do not say anything stupid about them being forced to release their code. That is false, no matter how much Billy at Microsoft wants people to think otherwise.
That copy of Windows was customized for that laptop, idiot.
If you installed a plain Windows the little blue buttons would not work.
Linus is confused. There is no reason to seperate 2D and 3D. If all the verticies in your 3D object have z=0 and you look at it flat on the screen, you have 2D. Thus it is trivial to make 3D accelerated hardware accelerate 2D.
So no matter where the 3D and 2D reside (kernel, user-space, with the ability to remotly connect or not...) they will reside in the same place in any intelligently designed system.
This was implemented over 15 years ago in IrisGL. These machines ran at about 5Mhz and were perfectly capable of using OpenGL to update a window system just as good as X does today. So speed has not ever been an issue.
Most of the problem is that OpenGL (as opposed to IrisGL) lacks some operations. There is no good handling of fonts and text. There is no easy way to create a complex clip region. There is no efficient way to take an image in local memory and create a rectangle texture-mapped with it (everything is designed to reuse textures, not do one-off copies). You also have to negotiate an ugly and complex glX (or WGL on Windows) interface before you can start drawing with OpenGL. Since these are things that 2D interfaces such as Xlib or GDI32 do, people are forced to use that for 2D, and then tend to segregate their programs into a "window" in which 3D is displayed, surrounded by 2D. It also does not help that Windows, and until recently X, had stupid design decisions so it is impossible to draw with both OpenGL and 2D into the same window, thus enforcing this segregation at the API level.
I would like to see a design where the necessary missing drawing functions are added to OpenGL (perhaps copy Cairo). And then make a library to replace xlib where you create a window and you immediately have a current OpenGL context that you can draw. And do the Keith Packard thing with compositing managers: the "window" can in fact be a texture map and this is then mapped to the screen through arbitrary transformation and shading, to produce Quartz-like effects.
If one was to win, I would have to bet on Gnome now. However as little as 6 months ago I would have said KDE.
What does appear to be happening is that this competition is figuring out a basic layer of functionality, far simpler than either KDE or Gnome, that is really necessary for interoperability. Mostly the freedesktop.org stuff like standards for drag & drop, and "window manager standards" which are really kludgy methods of storing process information in the X server so other programs can see it.
This could be a very good result, as it could cause "which desktop" to become as irrelevant to the programs that are running as which text editor you use is relevant to a C compiler, and make it as easy to experiment with a new desktop as it is to experiment with a new text editor.
The big problem is some of the basic underlying stuff, especially the COM-like stuff, is per-desktop. The only way it is going to be unified is if one of the projects abandons it's lower-level implementation and starts using the other one. This is going to suck for the engineers working on the abandoned one, but it has to happen.
Damn, you are right! It even works in tcsh. I have no idea why I thought you had to type "start" before the filename. In fact in cmd.exe "start" will hide error messages and otherwise screw up, while just the filename seems to work quite well.
I certainly thought this is how it *should* work but for some reason I got convinced that the "start" was necessary. Possibly I tried it once and it did not work.
Look at the shadows! They are all pointing in the wrong direction! And the reflections on his head don't line up with the overhead lights! And if you analyze the echos of the applause, you can tell that the room it was recorded in has wood panelling and not the plaster that you see on tv! All of this proves the speech was faked!
Absolutely agree. The name is backwards, and I can see no possible reason to do so. The reversed name sounds equally good and neither makes one system sound better than the other. I can't see any plausable insidious plot by Microsoft with the backwards name.
Well maybe: they want to make it really difficult for a future Microsoft to release Windows emulation software for Linux, by taking up the obvious name. Ie by using this name they are trying to prevent business decisions in the future that they don't like now?
Nah, that is too far fetched. I think it is just somebody in marketing being stupid.
You are allowed to compile non-GPL code with GCC, idiot.
However I am suprised that Microsoft would include a C compiler for free. That would cut into sales of VC++ pretty badly.
Minimizing to an icon on the desktop is how the very first X window managers worked. I also think it is how things earlier than that worked, such as XViews and Andrew. So this is a very old idea. Updating the icon was definately supported in the first versions of X with window managers, it was not easy, but they definatley thought that programs would want to change their icons to show status.
I actually think the biggest innovation to come out of Microsoft is the "toolbar" where the icons exist even if the window is open. Yes there were a lot of earlier schemes for arranging icons in windows or grids but all of them had the idea that the icon was only visible when the window was not visible. This meant that to find a window you had to look both for the icon and for the already-opened (but perhaps buried) window. Trust me this is an innovation and I do think Microsoft thought of it. And it is pretty obvious that all new desktop environments are copying this, even OSX (which shows an icon in the dock for all running applications, whether they are open or not).
So in a lot of ways I feel that anything that reduces a window to a desktop icon is a throwback to an idea that seems to have been discredited.
You don't have to use Windows+R, you can run command.exe and type the commands in there, which is a lot faster, as long as you are able to click on the terminal window.
No, you have to type "open " before the name of the document.
Windows has this and calls it "start". At least tcsh on Windows does, I eventually learned that the intuitive command in command.exe is "rundll32.exe url.dll,FileProtocolHandler ", I would guess that a shortcut like the initial poster described could be used to make "start" work there.
Of course Linux is far behind either, as there is *no* command that duplicates what a double-click does in a file browser. This is really bad! The KDE and Gnome people should both be ashamed. Though in some ways I think they fear that they will make it too easy to write file browsers, and their Microsoft-like control over the Linux desktop would be eroded, so they don't provide something that would seem obvious to any Unix user.
While the mainland, especially Germany, has been big on OSS, it seems the UK is more of a Windows shop than the USA is, and they appear to be targeting the UK. Of course they may not have anybody at SCO who know any languages other than english...
What happened to "Atheos" (it was called something like that). It was also a one-person effort to make a Unix-like system designed for the desktop, with integrated GUI. A few years ago that sounded very interesting, but nothing ever came of it. Not a good precedent for this project (unless this is the same project, it is hard to tell).
But I think one thing that killed Atheos is the same thing that killed almost any alternative to X: inability to support any modern graphics cards at any resolution higher than VESA. Unfortunatley this information is locked up in X drivers that are so tightly integrated with internal complexities of X that it is impossible to extract and reuse it, despite the open source nature. Perhaps XDrive will help here by making the driver interface cleaner. In any case this project sounds like it has some hardware acceleration, so maybe they will escape this trap.
Personally I am not thrilled with putting widgets into the OS. My feeling is that this locks GUI design and innovation. I would prefer a design where there were powerful graphics and event handling calls, so it is easy to write a widget, but the interface is designed so that it is obvious that you can write different widgets that have not yet been invented.
Installing or upgrading Windows will always hose your boot sector so that you lose Linux. Recent versions will try really hard to repartition your drive or reformat the unknown sectors, though I have been told you can stop this by hitting No enough times. So Microsoft has already released their version of this.
A battery (even a rechargable one) has negative energy costs too. It takes energy to manufacture them, plus all the energy that you get out has to be put in.
Mensa indeed.
From what I have heard they are using up about 1/4 or so of the disk to store WMA equivalents of the CD tracks. It would seem extremely sensible to put an autorun program on there that reads the CD tracks and creates the WMA (or something equivalent) files, and also does whatever messing they want with the poor user's Windows box.
Contrary to the other posters here, I don't think this would be any more or any less evil. It still won't work on Linux, but the same old workarounds will work on both Linux and Windows.
And they can put 10 more minutes (or whatever amount) more music on the copy-prevented disk. And there would be more assurance that you can't miss any music when you play the disk in the computer, that worry is probably more of an incentive to work around this than a desire to make an MP3.
Already done. It is built, and it is secret.
Look at Plan9 for ideas!
In theory it fixes all the problems with Unix, while still keeping to the original design philosophy.
Because such actions reinforce stupid decisions by legal departments such as yours.
If it becomes standard that you cannot buy or use code without "indemnity" then you have made it impossible for an independent code writer to write a program and have people use it. This is irrelevant to open source, what you have pretty much said is that writing code is the realm of big business and that start-ups and free enterprise and just the desire to tell people about solutions you have figured out have become illegal.
This goes far beyond Linux. Linux is now being protected by big money interests in the same way Windows is. But the next genius grad student who writes the "Linux-killer" operating system is going to be out of luck, as everybody who he shows it to is going to say "but you have no idemnity, so I'm going to have to stick with Linux..."
Even people who think Linux is the end-all of software should realize that this means the death of applications for Linux from independent authors, since they cannot "idemnify" their code. And Windows fans should realize this means the death of all the Shareware and Freeware and open source software, and also the few remaining tiny software vendors, all of who cannot afford "idemnification" either.
I don't care if this action causes SCO to go bankrupt tomorrow, this is a bad precedent for Novell to do this.
IMHO the damning evidence is the scanf strings, though I would really need to look at the source code and the documentation for how DVD menu information is stored, to see if there is another reason. My guess is that anybody writing this themselves could come up with different scanf strings, or multiple ones, or use something other than scanf to accomplish the same thing. But then again, if the documentation for the menu system on DVD's says "use a scanf string like this to parse this field" then it is pretty obvious why both programs use the same strings.
Obviously the list of names means they looked at the MPlayer code, but I think this can be excused: engineer A says to engineer B: "Hey we need to implement this list of stuff" and then cut & pastes the string list from the MPlayer source. I think that is fair use of the code. And of course the resulting list would end up in the same order. Notice that they changed the capitalization and changed "dunnowhat" to "unknown".
KISS does not have to open their source code. Even if they violated the GPL, all they have to do is remove that code and replace it. MPlayer could sue for copyright damages but I don't think they would. If in fact KISS did not steal the source code, they could probably prove it reasonably by revealing a copyrighted chunk of their code that can be easily seen to be unrelated, despite the fact that it produces the same strings.
Proof positive that the trolls here do not even bother reading the other comments.
Part of the problem is that it is not default, so an application cannot assumme it.
I'm using KDE and I have it set this way, but there are still problems. First of all you can't move or resize the window without raising it, they really should restore the old FVWM behavior where if you click it raises, but if you drag it just moves/resizes.
More of a concern is that KDE has a bug that also exists on Windows and makes even using child windows to enforce stacking impossible. This is that if you raise a "child" window it will also raise it's parent to just below it. This makes overlapping windows with floating toolbars impossible. Most window managers that obey TRANSIENT_FOR have this bug, including old ones. I'm not sure why but perhaps they are all copying the same code base.
Changing the user interface automatically has been proven to be bad and users do not like it.
However as other posters have pointed out, you can certainly add actions the user can do to cause the interface to change. Since the user actively does these things there is no confusion when the interface changes (well, in theory...)