Actually, from what I read, the only part of the XFree86 system that has been placed under the new licence are the server side components. Xlib and client side libraries, that you would actually link into your software application, are not and will not be placed under the new licence. These are the libraries that are included in your programs. The server side code that is under the new licence is not compiled into your programs, so there is no licencing conflict with GPL. Again, just the fact the GPL code is being shipped on the same CD with non-GPL code does not create a conflict, the code actually has to be linked into the same program with an incompatable licence for their to be a problem.
Also, from what I understand, the clause requires credit in places where other software credits are given, such as in documentation. It is not an advertising clause from what I understand, you do not have to plaster the credits whereever you refer to X, only in the documentation.
So there actually does not seem to be a conflict with GPL licencing in apps at all, so I dont understand why people are so concerned over the licence change. Its still an open source licence which you can freely use, modify, and redistribute.
For more information on the licence change, I really do recommend reading this page on XFree86s website: http://xfree86.org/legal/licenses.html There are potential issues as a result of this fork. A concern is that the driver API could become incompatable, between XFree86 and Xorg, so two sets of drivers would have to be maintained. Hopefully this can be avoided and a standard driver API/ABI maintained.
I really dont understand people who are complaining about network transparency and who wish to destroy one of Xs best features, and destroy Xfree86 as a X11-compliant implementation, by changing the X protocol. It seems to be working very well as it is. I use network transparency all the time to run applications running on a fast server across to several cheap, old PCs. It is a crucial feature, and provides much better performance than purely bitmap protocols. It has saved a lot of money, instead of having to buy a bunch of expensive workstations, you can simply use an application server, and some old ancient PCs as terminals which cost next to nothing, just try that with XP . To share applications on XP (a single application, not a display), you have to pay hundreds of dollars in licensing fees just for the products to do so. This feature has been avialable on X for 18 years, and freely avialable thanks to Xfree86.
It is important as well to maintain standards compliance with X11, so you can use a particular application with any X server without having to be concerned without running into compatability issues. Frankly, I dont think the need to change X11 protocol is there, we have DRI (and similar things) for apps that truly need high bandwidth, but web browsers, office suites, drawing programs, even animations and video, all work well under X11., i use them all the time under X and dont see a problem.
I really dont see this "bloat" or "cruft" people keep talking about in X11. X works fine for me, its great, its fast, and its small compared with XP! Its fast even with video applications. For those really demanding 3d or high bandwidth applications, we can have DRI and similar things which applications which truly need that kind of bandwidth can use. But most apps such as word processors, web browsers (animations and all), drawing programs, even realplayer, work excellent on top of X11 protocol.
No, we dont need to change the X11 protocol, and even on old computers I have found X protocol is more than fast and adequate in its display speed and capabilities. It seems to be working just fine, X protocol does not need to be changed and we do not need to enter a new era of incompatability, inflexibility, lack of versatility in unix GUIs. Standards compliant X11 protocol provides an excellent, time tested platform which is working very well and will continue to do so far into the future.
X11 itself is just fine and I see no need to change it, it is working great. Xfree86 as well actually has been doing a pretty good job I think of putting out a decent, stable, high quality, X11-compliant, versatile, flexible, system with many useful and beneficial features not removed. Perhaps it could do a little better yes, but it seems to be quite good already. One person commented on the size of a Xfree86 package being 70mb, however, the server itself is about 2 mb in size, most of the rest is drivers i am sure. Drivers should continue to be distributed with X, but it should now be very easy for OS distros (Redhat, etc) to offer drivers in seperate archives, so users can choose to install only the module they need. Of course, there can be seperate projects for video card drivers, especially with the new loadable module system, as those drivers become stable then they can always be included in the main Xfree86 distribution
Great, thats all we need, new incompatable versions of X. One of the great things about X is how it is so widely avialable across a wide range of platforms and operating systems, and interoperable and standardised at the same time. I can run a program on a a Sun workstation and display it to a PC running Linux, and vice versa. Few other window systems give you that kind of compatability, interoperability and flexibility. X is also avialable on nearly every modern operating system. Being a user of X, i think the X protocol works quite well as it is right now. As far as "cruft", I really dont see this, X seems to me to be a rather efficient window system, when compared with such hogs as Windows XP. X is in fact the best window system I have ever used, and I have used MacOS and many versions of Windows. I can run X with fvwm and Linux on old computers which things like Windows XP would never run on. X protocol seems to work quite well as it is right now.
I think Xfree86 does a pretty good job of providing a completely X compatable, full featured, flexible, versatile window system, not creating some incompatable, non-standard system with features that give it flexibility or usefulness missing or removed. I think the DRI concept is quite good as well as a solution to the need for high bandwidth graphics in X.
It would seem to me that the sensible, logical thing to do, instead of spending lots of time rewriting Star Office for the Mac OS X GUI API, would be to simply run Star Office under a rootless X server for Mac OS X. In this way, most Unix/X programs could be ported without touching any of the GUI code. It would save a lot of time, and provide great results. A rootless X server displays X programs to within the OS X GUI, so they can be used in windows right alongside other programs. This elimates the need to rework GUI code.
Slashdot Mirror
Actually, from what I read, the only part of the XFree86 system that has been placed under the new licence are the server side components. Xlib and client side libraries, that you would actually link into your software application, are not and will not be placed under the new licence. These are the libraries that are included in your programs. The server side code that is under the new licence is not compiled into your programs, so there is no licencing conflict with GPL. Again, just the fact the GPL code is being shipped on the same CD with non-GPL code does not create a conflict, the code actually has to be linked into the same program with an incompatable licence for their to be a problem.
Also, from what I understand, the clause requires credit in places where other software credits are given, such as in documentation. It is not an advertising clause from what I understand, you do not have to plaster the credits whereever you refer to X, only in the documentation.
So there actually does not seem to be a conflict with GPL licencing in apps at all, so I dont understand why people are so concerned over the licence change. Its still an open source licence which you can freely use, modify, and redistribute.
For more information on the licence change, I really do recommend reading this page on XFree86s website:
http://xfree86.org/legal/licenses.html
There are potential issues as a result of this fork. A concern is that the driver API could become incompatable, between XFree86 and Xorg, so two sets of drivers would have to be maintained. Hopefully this can be avoided and a standard driver API/ABI maintained.
I really dont understand people who are complaining about network transparency and who wish to destroy one of Xs best features, and destroy Xfree86 as a X11-compliant implementation, by changing the X protocol. It seems to be working very well as it is. I use network transparency all the time to run applications running on a fast server across to several cheap, old PCs. It is a crucial feature, and provides much better performance than purely bitmap protocols. It has saved a lot of money, instead of having to buy a bunch of expensive workstations, you can simply use an application server, and some old ancient PCs as terminals which cost next to nothing, just try that with XP . To share applications on XP (a single application, not a display), you have to pay hundreds of dollars in licensing fees just for the products to do so. This feature has been avialable on X for 18 years, and freely avialable thanks to Xfree86.
It is important as well to maintain standards compliance with X11, so you can use a particular application with any X server without having to be concerned without running into compatability issues. Frankly, I dont think the need to change X11 protocol is there, we have DRI (and similar things) for apps that truly need high bandwidth, but web browsers, office suites, drawing programs, even animations and video, all work well under X11., i use them all the time under X and dont see a problem.
I really dont see this "bloat" or "cruft" people keep talking about in X11. X works fine for me, its great, its fast, and its small compared with XP! Its fast even with video applications. For those really demanding 3d or high bandwidth applications, we can have DRI and similar things which applications which truly need that kind of bandwidth can use. But most apps such as word processors, web browsers (animations and all), drawing programs, even realplayer, work excellent on top of X11 protocol.
No, we dont need to change the X11 protocol, and even on old computers I have found X protocol is more than fast and adequate in its display speed and capabilities. It seems to be working just fine, X protocol does not need to be changed and we do not need to enter a new era of incompatability, inflexibility, lack of versatility in unix GUIs. Standards compliant X11 protocol provides an excellent, time tested platform which is working very well and will continue to do so far into the future.
X11 itself is just fine and I see no need to change it, it is working great. Xfree86 as well actually has been doing a pretty good job I think of putting out a decent, stable, high quality, X11-compliant, versatile, flexible, system with many useful and beneficial features not removed. Perhaps it could do a little better yes, but it seems to be quite good already. One person commented on the size of a Xfree86 package being 70mb, however, the server itself is about 2 mb in size, most of the rest is drivers i am sure. Drivers should continue to be distributed with X, but it should now be very easy for OS distros (Redhat, etc) to offer drivers in seperate archives, so users can choose to install only the module they need. Of course, there can be seperate projects for video card drivers, especially with the new loadable module system, as those drivers become stable then they can always be included in the main Xfree86 distribution
Great, thats all we need, new incompatable versions of X. One of the great things about X is how it is so widely avialable across a wide range of platforms and operating systems, and interoperable and standardised at the same time. I can run a program on a a Sun workstation and display it to a PC running Linux, and vice versa. Few other window systems give you that kind of compatability, interoperability and flexibility. X is also avialable on nearly every modern operating system. Being a user of X, i think the X protocol works quite well as it is right now. As far as "cruft", I really dont see this, X seems to me to be a rather efficient window system, when compared with such hogs as Windows XP. X is in fact the best window system I have ever used, and I have used MacOS and many versions of Windows. I can run X with fvwm and Linux on old computers which things like Windows XP would never run on. X protocol seems to work quite well as it is right now.
I think Xfree86 does a pretty good job of providing a completely X compatable, full featured, flexible, versatile window system, not creating some incompatable, non-standard system with features that give it flexibility or usefulness missing or removed. I think the DRI concept is quite good as well as a solution to the need for high bandwidth graphics in X.
It would seem to me that the sensible, logical thing to do, instead of spending lots of time rewriting Star Office for the Mac OS X GUI API, would be to simply run Star Office under a rootless X server for Mac OS X. In this way, most Unix/X programs could be ported without touching any of the GUI code. It would save a lot of time, and provide great results. A rootless X server displays X programs to within the OS X GUI, so they can be used in windows right alongside other programs. This elimates the need to rework GUI code.