The Open Group (who took over X11 from the former X Consortium) had changed the license terms on the X11R6.4 release of the "Sample Implementation" of the X11 windowing system so that it no longer qualified as Open Source Free Software. This made it effectively unusable by any of the "free" OS projects such as Linux, NetBSD, FreeBSD, and OpenBSD.
This was all quite disappointing and could be argued to be a betrayal of the many students and professionals that have contributed to the design and construction of this implementation with the expectation and understanding that it would continue to be freely available.
This meant that some Linux distributions, most notably Debian could not ship this new version. Commercial vendors such as Caldera, Red Hat, and SuSE may be able to make arrangements with the Open Group to allow distribution, but the license would encumber them in subtle and surprising ways.
In September, 1998, a "new license" was announced, that essentially the same as the one formerly used, which is once again compatible with the past and continuing efforts of The XFree86 Project group, allowing continuing integration of X11R6.4 technologies into XFree86 and vice-versa.
The remainder of this page I have left unchanged from when the "old, bad" license was in effect; the information may still be of some historical interest.
The XFree86 Project is the nonprofit organization that has been primarily responsible for the releases of X available for use with various free operating systems such as Linux, NetBSD , FreeBSD , OpenBSD , and GNU Hurd.
Note that the X11R6.4 license has changed to a "friendlier" one since the writing of this material. This section is of historical interest; it no longer represents a true statement of affairs.
The XFree86 Project announces its position regarding the recently announced licensing of X11R6.4 by The Open Group (TOG) of Cambridge MA. "As a direct response to the new licensing terms made by TOG a week ago, we have found it necessary to state publicly the position and future direction with which The XFree86 Project will proceed," said David Dawes, President of The XFree86 Project, Inc.
The X11R6.4 sample implementation (SI) from TOG comes with a new license. The new licensing reduces the freedom with which their SI may be used. This is because a license fee must now be paid by those who generate revenue from distributing their SI or code derived from it. TOG has stated that the X11 standards will, however, remain "open."
The TOG licensing change is incompatible with the goals of XFree86 and most of the free software community that have been in communication with The XFree86 Project directors over the last week. Some have made the point that the change would make it extremely difficult for their projects to ship physical media which contains TOG derived product, and they've made their preference for unencumbered X11R6.3-based source releases overwhelmingly clear. A fundamental part of XFree86's mission is service to the free software community. As a result, The XFree86 Project is opposed to TOG's new licensing terms.
The XFree86 Project will continue its development based on the freely available X11R6.3 SI and, where appropriate, attempt to implement future developments to the X11 standards independently of TOG. "It is our intention to beef up our problem tracking system and request that people sending bug fixes or enhanced source code to TOG also send them to us, so that the free version of X can also benefit from them," said Dirk Hohndel, Vice President of The XFree86 Project.
The XFree86 Project remains committed to providing a freely redistributable (in source and binary form) implementation of X11. To this end, the copyright and licensing of XFree86 code will continue to be of the BSD style (as used by the X11 SI up to and including X11R6.3). These terms allow TOG and others to integrate XFree86 code into their distributions without a license fee or other financial consideration.
Linus Torvalds, principal author of the Linux operating system, said " The XFree86 Project is doing the only reasonable thing in this context. They have my full support in keeping the idea of free software alive for the X Window System. I hope that the public will recognize this important task and that many people and corporations will continue to support XFree86. I will personally work with them to help with their goals."
Garry Paxinos, Vice President of Metro Link Incorporated said " In keeping with the original philosophy of the X Consortium, Metro Link believes that the sample implementation of the X Window System should remain unencumbered. The decision reached by the XFree86 Project to fork the tree at R6.3 will ensure unrestricted access to the ongoing development and enhancements to the core X technology. Metro Link endorses this decision, and we will continue to contribute to and support the XFree86 Project. "
Theo de Raadt, leader of the OpenBSD group said " The OpenBSD group is massively pleased to know the X11 subsystems we ship on our 8 main platforms will continue to be available freely. The fact that XFree86 runs on all our platforms and the TOG's X11R6 SI runs on two or so shows that XFree86 has been doing a wonderful job of support for free operating systems on multiple platforms. "
Note that the base protocols of X have been stable for rather a long time now. The addition of gratuitous incompatibilities is not likely. It is of critical importance to continue to maintain a high degree of compatibility with existing X11R5 and X11R6 servers and clients, as there is a lot of hardware and software targeted at those operating platforms.
Interesting additions that could result include:
Integrating in True Type font support
Display Postscript is an extension that is permissible; since there was previously no "libre" implementation, it was not heavily supported.
As DGS uses the GPL, it will not be possible to deeply integrate it into XFree86 , but the "increasing modularization" of XFree86 should make it easier to bring DGS alongside...
Integrating in some sort of "portable execution engine" (perhaps a Java Virtual Machine) so that applications could run within an X server, if supported by the server.
OpenGL is a portable environment for deploying 2D and 3D graphical applications.
Far less likely are scenarios that involve rapid divergence away from X. Many person years have been invested in constructing and designing X clients, servers, and overall infrastructure. These efforts would have to be substantially replicated in order to build a new alternative to X.
Moving wholesale to GNUStep or Berlin, building these atop GGI are all wonderfully interesting ideas that would require years of further effort in order to provide anywhere near the functionality and interoperability already provided by X.
This looks to be a very important project; the plan is to provide a mechanism for communication particularly of the rendering of fonts.
A severe problem over the years has been in the handling of fonts. In particular, X has not historically provided a standard way to communicate font metrics, that is, metadata about the size and appropriate relative spacing of glyphs.
The result has been that any program that needs to be aware of glyph sizing has needed to have its own independent scheme for managing font metrics. Notable examples have included:
Ghostscript with Adobe Font Metrics (.afm)
Having an increasing number of applications going off implementing their own "font data registries" makes for a Bad Situation, so that whatever the Render framework can do to eliminate the need to do this should be helpful.
In any case, it is appropriate to solidify further support behind The XFree86 Project, including those of a financial nature.