Slashdot Mirror
XFree86 Alters License
kinema writes "According to the XFree86 announcement starting with XFree86 v4.4.0-RC3 there will be a new license. There are some worries that these changes might be incompatible with the GPL." The FSF has a good page about the problems with BSD-style advertising clauses, which ironically uses XFree86's old license as an example of one to emulate.
the source is still out there, worst case scenario - the license isn't gpl, and they don't change it to be so - some of the developers split off and recruit noobs, and we end up with a renamed X that everyone uses - that takes a little while to ramp back up to full speed.
It's not the end of the world, but it (could) be annoying, that's for sure. I think thorough investigation is needed (e.g. try reading the license)
cyn, free software and *nix operating systems enthusiast.
Will this affect the popularity of the GPL?
XFree86 is using a different license, as is Apache... will this put off others using the GPL, and encourage them to use a license of their own creation that best suits their needs?
This isn't a troll, but seriously, why should it be GPL compatable? The only way to be GPL compatable is to have a licence where the software can re-licensed under the GPL. The GPL states it can only be linked with other software under the GPL (or under a licence which basically means the same thing)
If the GPL is unwilling to be compatable with anyone else, why should anyone be too worried about being compatable with the GPL.
Remember. Open source =\= GPL.
Combination - fun iPhone puzzling
Programming can be fun again. Film at 11.
This only seems to concern documentation, not ads, so the problems described on the FSF page are not quite as bad - full page of credits on ad may be bad, but full page of credits among hundreds of other pages of documentation isn't nearly as big deal.
Anyway this seems to be rather stupid move, XFree86 seems to have enough problems (infighting, resulting diverting and forks...) already without any license trouble. If it ends up incompatible, all the more reason to concentrate on those, if the XFree86 folks want to shoot themselves to foot and slowly bleed to death, it's their choice - very stupid one but it's their nevertheless.
The requirement to add an acknowledgement (often called the "advertising clause") is something that's not present in the GPL. The GPL states that GPLed code may only be distributed if further requirements or restrictions above and beyond the GPL are not made. It's therefore impossible to satisfy the terms of both licenses simultaneously, and therefore the extra clause makes the new XFree license GPL incompatible.
Now, why's this a problem? The XFree project doesn't include GPLed code itself, so there are no concerns from that point of view. However, if any of the code in the X libraries falls under this new license, then the FSF's interpretation of the GPL means that you wouldn't be able to link any GPLed program against the X libraries and distribute it. That's fairly bad.
On the plus side, the freedesktop.org x libraries were branched from XFree before the license change - as a result, we can pretty much guarantee that there will be libraries available that can be used with GPLed code. The end result would probably be to reduce the amount of XFree code in a distribution, rather than to increase the credit that the XFree project wants. It's almost certainly a counter-productive move.
Version 1.1 of XFree86 Project License.
Copyright (C) 1994-2004 The XFree86 Project, Inc.
All rights reserved.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
1. Redistributions of source code must retain the above copyright notice, this list of conditions, and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution, and in the same place and form as other copyright, license and disclaimer information.
3. The end-user documentation included with the redistribution, if any, must include the following acknowledgment: "This product includes software developed by The XFree86 Project, Inc (http://www.xfree86.org/) and its contributors", in the same place and form as other third-party acknowledgments. Alternately, this acknowledgment may appear in the software itself, in the same form and location as other such third-party acknowledgments.
4. Except as contained in this notice, the name of The XFree86 Project, Inc shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization from The XFree86 Project, Inc.
THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE XFREE86 PROJECT, INC OR ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Is there a way to post a babelfished link? Does google do tranlation stuff yet?
Anyway... enjoy.
=-=-=
Does XFree86 GPL become incompatible?
Sent of demon at the Fr, 30 January 2004 around 10:21
Durch a change of the license regulations will will become the coming version of XFree86 incompatible to the well-known GPL and a linking of GPL applications to XFree86 "problematic".
Hardly the turbulences in the XFree86-Lager grew silent, seem a further controversy from the fence to to break. As David Dawes of the XFree86-Projekt communicated, the XFree86-Projekt changes its license on a new version 1.1. A change of the license represents no point at issue in the reason still, became nevertheless already in the past restrictive licenses of liberals. This time the situation seems to be however more problematic, because XFree86 becomes more restrictive.
As license can be taken further very much from the liberal, can be changed, driven out and applied all programs under the "XFree86 License 1,1" without publication of the source code. Again was added however a clause, which means that each distribution and each product, which contain XFree86 must attach a note either in the documentation or in the application on XFree86 ("This product includes software developed by The XFree86 Project, Inc. (http://www.xfree86.org/) and its contributors"). That is problematic, as the Free software Foundation already meant in another case.
The organization had not GPL compatibly classified the first version of the BSD license in the past as and had expressed substantial doubts against a linking of GPL- and BSD applications. The famous "BSD advertising clause" does not make the license unfree, cause however practical problems, including an incompatibility with the GNU GPL, so the Foundation. In the past the Free software Foundation guessed/advised to use the straight XFree86-Lizenz because it was to a large extent with the BSD license compatible and the notorious clause does not contain. Thus conclusion might probably be, because a determination of the BSD clause as "GPL incompatible" makes automatically also the new XFree86-Lizenz for GPL applications "problematic".
Which follows from the earlier declaration of the BSD license, could extensive consequences both for the Distributoren as well as for other manufacturers have. Thus GPL applications may be linked against an GPL incompatible library, this require however a note in the source code - a condition, which will fulfill hardly an application. If this note is not contained, linking is not permitted.
Thus either if XFree86 should not change their license or the Free software Foundation their declaration, then the current version of XFree86 will be probably also last release delivered by the Distributoren. Because it is questionable whether Distributoren get involved in a "problematic" use. Options during a non--change might be either freedesktop.org, Xouvert or a new Fork. Not completely averse would be also the developers. As pro Linux from KDE Entwicklerkreisen experienced, also they are not with XFree86 content and think ever more frequently about one transferred to freedesktop.org. (thanks at Rene.)
"some worries" would require a trip here to read it in (semi-)English?
Here's the original BSD clause:
And here's the new XFree86 clause:
The problem with the original clause 3 of the BSD license is that it could lead to massive lists of acknowledgements tacked on to an advertisement meant to be fairly compact (e.g. a leaflette, banner ad, sign, billboard, whatever). This isn't the case with the new XFree86 license clause 3, where it only requires acknowledgement in the documentation or the software itself. While keeping track of those acknowledgements might prove difficult at times, it has nowhere NEAR the practical problems that the original BSD license had.
A GPL inkompatibel XMoose86 once bit my sister... No realli!...Mynd you, XMoose86 bites Kan be pretty nasti...
The new XFree86 license requires a statement in end user documentation, which is completely different. You can't really argue that adding a bunch of disclosures about where the modules you're using to your documentation is a huge burden. It doesn't add a substantial cost to your documentation, even if it's distributed in a printed form, unlike the cost of adding a page of disclosures to an ad.
Don't blame me; I'm never given mod points.
If I had a quarter for every time I said that, I'd ...I'd ...I wouldn't have any quarters. And I still have laundry to do.
This is the major problem with open-source software. All the project decide to make their own license so they can set special terms. This will only cause problems. If they would all agree on one license (GPL or not) then it would be much easier to get excited about using open-source software because I wouldn't have to read all the licenses to make sure they are compatible with each other. I also wouldn't have to worry about using a particular piece of software in an illegal manner because I would learn from Project A that I can't do certain things with it and then I would know that the same will be true for Porject B.
"But I'm still right here, giving blood and keeping faith. And I'm still right here."
This seems to be a big difference in practice - even hundreds of such lines in the docs would be manageable, while it would be pretty much impossible to include them all in, say, a banner ad.
Of course, this doesn't mean that it's not GPL-incompatible, because it still is a requirement the GPL doesn't have. But frankly, who cares?
Programming can be fun again. Film at 11.
If the GPL is unwilling to be compatable with anyone else, why should anyone be too worried about being compatable with the GPL.
... unless you are someone who feels threatened by free software in general, or people who differ from your vision of free software in particular, and therefor prefer fragmentation over cooperation.
The GPL has been THE reference license since probably before you were born (tongue in cheek).
BSD and GPL are the two original free software licenses. The BSD folks have made an effort to insure that the BSD license is compatible with the GPL not because they share the GNU philosophy (they don't), but to avoid fragmenting the free software world through stupid licensing incompatibilities. FreeBSD changed their license to make it GPL compatible, and GPL v. 2 was changed likewise to be compatible with a wider range of interests (including commercial interests that are shared with the BSD community).
The GPL is the only license many enterprises will consider releasing their erstwhile proprietary code under, as it protects them from having competitors snatch up their code and incorporate it into a competing proprietary product (in their view, competing GPLed products are not an issue, as they can reincorporate the best improvement into their GPLed product). Many of us who write code will not consider a BSD style license because we do not want our code used by freeloaders who incorporate it into non-free, proprietary products.
There are enough (perhaps a majority, even) free software and open source developers who feel this way that the GPL is, if not the majority license, a sufficiently large piece of the OSS / FSS pie that being incompatible with it means losing a huge portion of the community's input and integration.
FreeBSD, as vehement as their disagreement with the GPL is, chose to deliberately modify their license to make it compatible with the GPL for exactly these reasons: because there is room in the community for both views, but no reason whatsoever to fragment the community over those views.
After all, if one licenses under a *BSD style license, and if therefor one doesn't mind having their code placed into a proprietary product, why should one mind having it incorporated into a GPLed product (unless one's goal is simply to fragment the free software world and undermine the cooperation that makes it so effective).
Which makes one wonder about the motives of someone who would post such an inane comment actively encouraging such small minded thinking ("we don't use their license, we don't like them, so why should we cooperate!")
The Future of Human Evolution: Autonomy
If they go with a more proprietary license, they should probably also change their name to Ex-Free86. :P
Any good copy left license will require sublicensing by that license as a requirement for compatibility. It is too hard to capture the concept of copy left in any other way.
... but lots of people like that aspect, and there is plenty of worthwhile GPL software out there some of which even the XFree developers use. It is the choice to suddenly change to an incompatible which needs the justification IMO.
GPL keeps source open, that might not be everyone's definition of open source
"If XFree86 doesn't back out, it could spell doom for the project as a whole,..."
Hmmm...do I smell Xouvert? or perhaps freedesktop ?
Regards,
Steve
Can somebody explain why this new 1.1 license is necessarily incompatible with GPL2 / LGPL? True, it is an annoying licence change as the FSF article explains, and may not be a smart move for the project. But annoying doesn't make it incompatible. And no one even said (that I can determine) that original flawed BSD license was in fact incompatible; just undesirable.
In fact, this seems to be less restrictive than the GNU FDL license for documentation. It's not the same as past famous GPL-incompatible licenses, such as an old version of the Python license.
That being said, one would hope that the continued work on the next generation of GPL will consider whether or not "advertising clauses" will really result in GPL incompatibility. The mission of the FSF and the GPL is to make sure that the code can be freely used and reused. It's unclear how requiring positive attribution would interfere with that. Aren't there options for that sort of thing in FSF's Free Documentation License?
Admittedly, it's a slippery slope -- imagine a license with a clause requiring binaries be accompanied by a message advocating a particular political position. Or a particular sexual position, for that matter...
The Busy Coder's Guide to Android Development
Contributors: You said the code would be left in the CVS under our supervision!
XFree86: We are altering the license. Pray we don't alter it any further.
Imagine if a software system required 75 different sentences, each one naming a different author or group of authors. To advertise that, you would need a full-page ad.
Christ. And I was worried about Iraq, gun control and third-world starvation for a moment there.
Bow, nigger. h
Read the openssl license sometimes.
The author of that license seems to hold a deep grudge against the GPL, and specifically coded his license to make it incompatible (explicitly!!).
Anecdotally, it actually seems very common for BSD advocates to hate on the GPL. GPL users have no problem incorporating most BSD stuff with compatible licenses, because the product is GPL'd its not going to bother them.
The BSD crowd dont seem to be afraid that proprietary interests will advance the code such that the free version atrophies, but they do seem concerned that a rebadged GPL version could do just that: become the new "official" version. That would preclude any more commercial forking they had planned.
Personally, I thought the commercial fork pipe-dream was last used successfully by bill joy. I dont know why it still has so many adherents- proprietary is clearly not the wave of the future.
Why doesn't everyone use the BSD license?
Because it includes "the freedom to abuse". Companies say the GPL is anti-commercial, but the free OS with the most commercial interest is GNU/Linux.
Would Sun contribute to GNOME if the GNOME license allowed IBM to take Suns work, modify it and not give back? GPL makes a level playing field, everyone has to play by the rules, and history has proved that companies prefer that situation to the BSD situation.
In an ideal world, yes we would all use the BSD license, but while were in this world, copyleft seems to be preferable.
Expert in software patents or patent law? Contribute to the ESP wiki!
Here is a human-translated version of the german article:
The upcoming version of XFree86 is, due to a change in the licensing agreement, incompatible with the well-known GPL, and linking GPL applications with XFree86 will present itself as "problematic."
The troubles in the XFree-86 camp have just died down, but a further problem is beating at the gates. As David Dawes from the XFree-86 Project shared, the XFree-86 Project is changing its license to a new version 1.1. A change in the license does not, in principle, present a reason for causing problems, restrictive licensces have become more liberal in the past. The situation is, however, this time more problematic - the XFree86 license is getting more restrictive.
The license can still be seen as very liberal - all programs under the "XFree86 License 1.1" can be used, published, and advertised without the publication of the source code. A new addition, however, is a clause which states that every distribution and every product that contain XFree86 must affix a remark, either in the documentation or in the application ("This product includes software developed by The XFree86 Project, Inc (http://www.xfree86.org/) and its contributors"). This is problematic, as the Free Software Foundation has remarked in another case.
The organization has classified the first version of the BSD license as not GPL-compatible in the past, and expressed doubts reguarding the linking of GPL and BST applications. The famous "BSD Advertising Clause" doesn't make the license non-free, but causes ither problems, including an incompatability with the GNU GPL, and for that reason the foundation [GNU] has asked the Free Software Foundation in the past to use the XFree86 license, because it is largely compatible with the BSD license and does not include the feared advertising clause. And with that may be the end, because assigning the BSD-clause as "GPL incompatible" makes the new XFree86 License for GPL applications automatically problematic.
The resolts of the earlier BSD License's declaration could have far-reaching effects for the distributors as well as for other manufacturers. GPL applications are allowed to be linked to a GPL incompatible library, with the need for a remark in the source code - a condition that will be met by very few applications. When the remark is not present, the linking is not allowed.
If XFree86 doesn't change its license, and the Free Software Foundation doesn't change its declaration, then the latest version of XFree86 will also be the last release rolled out by the distributors, because it is questionable whether distributors will get involved with a "problematic" use of the software. When the license is not changed, the options will be either freedesktop.org, Xouvert, or a new fork. Developers could be not entirely dissatisfied. As Pro-Linux from KDE-developer circles experienced, the developers are also not entirely satisfied with XFree86 and are still thinking about changing to freedesktop.org (thanks to rene.)
If the new license is generally seen to be GPL incompatible, this could make things more difficult for GPL projects such as fbcon, DirectFB or even mplayer's output drivers that are based on drivers from XFree86.
...Write your own code.
Ive heard that said so many times to people who have expressed a desire to use GPL code but dislike the GPL license. Why doesnt that apply here? It doesnt have to be GPL compatable, and if anyone dislikes that, they are free to extend the GPL compatable version, or write their own implementation.
I don't see any moral difference between RMS insisting that you call the operating system "GNU/Linux" and the XFree86 people insisting they get credit for their work. (Technically I see a difference, as there's nothing forcing you to call it GNU/Linux. But morally it's the same thing.)
Actually the GPL section 1 requires you to: publish on each copy an approriate copyright notice
This applies to both source and binary distribution. While this is not a real a advertising clause it does require you to acknowledge the original author of the program. So even with the GPL you have the problem of many copyright sentences in combined programs.
This isn't the case with the new XFree86 license clause 3, where it only requires acknowledgement in the documentation or the software itself. While keeping track of those acknowledgements might prove difficult at times, it has nowhere NEAR the practical problems that the original BSD license had.
You are right, it isn't as immediately bad as the original BSD clause, but it does appear nevertheless to be incompatible with the GPL, and therefor with a huge volume of free and open source software development. This in and of itself is a problem (c.f. my other post on this subject regarding fragmentation of the open source / free software community).
But, more to the point, this is a practical problem.
Consider: one of the long term unbeatable strengths of free software (and the real reason entities like Microsoft are so desperate to destroy it at any cost) is that free software can be built upon and improved indefinitely.
No need to reinvent the wheel every five, ten, fifteen, thirty, fifty, or one hundred years. Incremental evolution can and largely has replaced gutting and starting over from scratch.
Which means we could be using an unrecognizably (but incrementally improved and altered) version of GNU/Linux, XFree, or what have you a century from now. And, lest you think this is far fetched, consider that the concepts in UNIX (and Linux) date back over thirty years, and that COBOL code written even longer before that is still deployed in many legacy installations. Incremental improvement and modification is almost always preferred over revolutionary "rip out everything and start over" (indeed, the best reason for doing the latter is to get out from under the Yoke of proprietary software lockin and move toward a model -- free software -- that allows incremental, long term gradual improvement instead of triannual reinvention of the wheel a la Microsoft and, to a lesser degree, Apple).
Think of how many attributions your documentation (or your bloated code) is going to have to contain after ten years. Twenty or thirty years. Fifty.
We're talking hundreds of names at least, and by the end of half a century, probably thousands. All of that bloats the code, and, over an indefinite time frame, will probably mean a seperate volume of documentation just for the attributions!
It is, long term, as unworkable as the BSD clause. The only real difference is that scalability isn't as immediate an issue. The curve of the function is the same shape, it is merely the coeffecient that is different, and while the bloat may coming in years instead of months, it is coming nevertheless.
The Future of Human Evolution: Autonomy
Trying not to bite on what might be flamebait, but the GPL does not restrict how you licence your code. It only restricts how others licence it.
Using the GPL allows me to say "here you go, use the code however you like but don't ever stop others from doing the same". If you make a change to my code then you are welcome to keep it to yourself or, more usefully, to redistribute it but you can never change the conditions under which I originally released my code. If you don't like that condition then go away and recode it yourself, duplicate the effort, and miss out on the community effort but don't think you can use my code in whatever way you want.
In my book that isn't virulent.
This will only affect contributors to XFree itself. XFree isn't the only X implementation and therefore stuff that links against X libraries can't be considered a derived work of XFree any more than it can be considered a derived work of a commercial X server or of XOuvert or fd.o's X server.
"'I pass the test,' she said. 'I will diminish, and go into the West, and remain Galadriel.'"
- JRR Tolkien.
And, frankly, I can understand why some people are a little pissed of, even if I don't share their feelings. From the point of view of a developer using a BSD-style, permissive license, GPLed code is just as impossible to integrate as proprietary code is, so there already is a schism in "the community". Cooperation between GPL and BSD (or rather, copyleft and permissive) projects is effectively a one-way street.
I, frankly, do not understand why the BSD-License zealots (which are a tiny fraction of the BSD folks) get so pissed off.
They have made a conscious choice in the working of their license to allow their code to be arbitrarilly relicensed, even with draconian, proprietary restrictions and have taken the stance that by doing so, they have maximized the freedom of the downstream (derivative) developer. Which, by definition, must also include other free software developers whose specific views a freedom differ slightly, or even a great deal, as well as those who do not believe in freedom at all (ie. proprietary developers). Then, with the next breath, the zealously anti-GPL crowd would add "but not another FREE license we happen to disagree with."
It is hypocracy in the extreme to make a claim to freedom, then with the next breath to decry those who practice freedom differently than oneself while claiming it is perfectly okay to remove that freedom completely. Thankfully, that is a form of hypocracy the vast majority of the FreeBSD folks, including the leadership, do not engage in at all (and in fact, have purposly avoided by making their license GPL compatible).
It is not a hypocracy certain BSD-License zealots have avoided at all, or those who enter the community as agent provocatuers seeking to stir up conflict where none really exists, quite probably at the behest (and paid for) by certain interests who feel threatened by free software of whatever variety.
The Future of Human Evolution: Autonomy
That's a good reason for the GPL. That's the place the GPL makes sense.
But, what if your first priority is widest possible influence? For instance, you are trying to propagate a new protocol far and wide. In that case, I believe, that you would be wise to BSD the reference implementation.
Absolutely! The Ogg-Vorbis folks did this very thing.
Perhaps I didn't make it as clear as I intended. Both licenses have their place, both are good, and fragmenting the community through incompatabilities because one doesn't like the GPL would be a disservice to both the GPL and *BSD communities (as both do cross-polinate one another, with ideas and code).
Dual licensing is appropriate in some cases. BSD licensing is appropriate in some cases, and GPL is apporpriate in some cases.
What isn't appropriate is to advocate allowing folks to make free software proprietary, and with the next breath decrying folks who wish to take the same software and relicense it with vastly less draconian restrictions, but nevertheless more restrictions than it had originally (i.e. the GPL).
Choice is important, and the best way to maximize people's choices is to keep our free licenses as compatible as possible, and compatability withh the GPL, as one of the two fundamental reference licenses of the free software community (FreeBSD being the other), and as the license under which a large portion of the free software in the world is licensed under, is a very important part of that.
The FreeBSD folks, much to their credit, recognized that a long time ago. Alas, some of the more zealos folks in their ranks (along with some of the more zealous folks in the GPL ranks, and certainly the numerous agents provocateurs folks like Microsoft have seeded our ranks with), will probably never recognize (or at least never admit) as much.
The Future of Human Evolution: Autonomy
To define freedom:
freedom ( P ) Pronunciation Key (frdm) n.
1. The condition of being free of restraints.
To define liberation:
liberation ( P ) Pronunciation Key (lb-rshn) n.
1. The act of liberating or the state of being liberated.
Now that the terms have been defined, it is quite clear that freedom says absolutely *nothing* about "maintaining freedom," though liberation *clearly* reflects what GNU is trying to achieve with the GPL.
I'm not certain why potential lack of "GPL compatability" is of concern to anyone. I don't see anyone criticizing the GPL for "lack of BSD, MIT, and Public Domain license compatability," which I find to be a serious problem.
I find it apalling that slashdot has the audacity to state that "The FSF has a good page about the problems with BSD-style advertising clauses." So, let me get this straight, a site/post that highlights the problems with copyleft in the GPL is a troll, and a site/post that highlights the problems with attribution in the BSDL is insightful? Definitely /. bias here.
On a practical level, the advertising clause shouldn't be an issue, (it never has been, with Internet Explorer, Mozilla, or Mac OS X) as far as practical use is concerned. Also, advertising clauses are not anti-copyleft. It is simply a matter of the GPL deciding to break compatability with such licenses. GPL-like licenses such as the MPL I believe have no such restriction.
Simply put, the only real problem with advertising clauses is that RMS doesn't tolerate them.
hint to modders, this comment: "-1 True".