XFree86 4.4: List of Rejecting Distributors Grows

Bootsy Collins writes "Yesterday, we discussed Mandrake's decision to revert their release-in-development from XFree86 version 4.4 back to version 4.3 because of issues with the new XFree86 license. To update this, the list of OS distributors opting out of XF86 Version 4.4, and future releases, based on licensing concerns continues to grow. While Fedora seems to be "preparing to support multiple X11 implementations", Red Hat has explicitly stated that they have no plans to ship XFree86 v4.4 under its current license. Also add to the growing list list Debian, Gentoo, and OpenBSD."

Re:What is the issue?

[Llanfairpwllgwyngyll]

It is incompatible with GPL licensed code which links with it. It's actually the GPL'd bit which gets its licence broken.

They will have to revert or face a fork IMHO

Re:What is the issue?

[bartjan]

Believed to be not compatible with the GPL.

The license itself is probably Open Source, but distributors would break the license of every GPL'ed program that links with XFree86.

Re:The Question

[richie2000]

On its face, the new license seems both reasonable and fair -- however, it also seems to create a lot of questions regarding how it should be interpreted [xfree86.org] and this is causing all the noise. My guess (and sincere hope) is that a clarification from, and possible minor re-write of the license by the XFree Project, Inc will clear this all up.

Re:The Question

[spacecowboy420]

From what I understand - correct me if I am wrong - The license basically says "We're GPL, but some files included may not be - you better check those out for their specific licenses". Which means you have to deal with every single file associated with xf86 individually and there is a bunch. Most are saying that this is too big of a hassle and likely to be a great source of pain - and simply not worth it.

Re:NVIDIA?

[pyr0]

You are right, they are non-free. That's not the issue here. The issue is that NVIDIA cards are currently the best you can use for 3d acceleration in linux. Yeah, ATI has binary drivers now too, but from what I understand they tend to be unstable and games don't run so well. It would be a hard pill for me to swallow if I couldn't play some games in X anymore.

Doesn't seem as ugly as TeX's license

[niminimi]

IIRC, you're not allowed to
distribute modified versions of TeX,
only patches alongside "pure" TeX.

So how can this be worse?

Re:What is the issue?

[Anonymous Coward]

Among other things, it is not compatible with the GPL. That means that would be a GPL violation if a distribution chooses to supply GPL software linked with xlibs.

Lots of desktop apps are affected.

See the incompatibility with the GNU GPL [gnu.org] and more practical problems [gnu.org]

Re:The Question

[ogre57]

Short answer, yes, for binary distribution it is that bad. For more [slashdot.org] than you want to know ..

Re:The Question

[Asmodai]

There are already some forks.

The license was changed on some files owned by the XFree86 Project.

One of the major nits, as can be seen on the mailing list, is that code flows from XFree86 to say the Linux framebuffer project, which can incorporate the code, but it can never flow back due to it being GPL (viral nature and such).

The license only requests proper attribution in software and/or documentation like other third parties are getting.

Nothing spectacular if you ask me. Only thing that's getting people in a knot is that it is incompatible with the GPL's 6th clause (funny how people actually rate a 3 clause license with the only requirement proper attribution to be more restrictive than the god knows how many clauses GPL, but that's another discussion).

I fail to see what people are making a fuss about.

Text of License

[Bouncings]

Announcement: Modification to the base XFree86(TM) license.

After a thorough re-examination of the XFree86(TM) license and reviewing
how it fits in with the Project's long-stated licensing philosophy ("You
can do what you like with the code except claim that you wrote it."),
The XFree86 Project, Inc. has made some changes to its base license.
This license review was prompted by a desire to ensure that XFree86 and
its contributors are receiving due credit for their work. The text of
the modified license can be found at
http://www.xfree86.org/legal/licenses.html.

The purpose of these changes is to strengthen the "except claim you
wrote it" clause of the Project's licensing philosophy regarding binary
distributions of XFree86. While the original license covered this
adequately for source code redistribution, it has always been lacking
where binary redistribution was concerned.

This modified license falls easily within the long-standing XFree86
licensing policy, and so there has been no change to the classes of
licenses acceptable for code contributed to XFree86. In fact, some
contributions to XFree86 were covered by a similar license already.
Contributors to XFree86 remain free to retain copyright on the code they
contribute, and can also choose the license for their code within the
long-standing XFree86 licensing policy.

The license change applies to the base XFree86 license, and to source
files that explicitly carry a copyright notice in the name of The XFree86
Project, Inc. Copyrights and licenses in the names of others will not
be affected by this change. Furthermore, only a subset of such files
with an explicit copyright notice in the Project's name will initially
carry the modified license, which is the core XFree86 components, and
the source files where there is no explicit author list. The license
in the remaining files with an XFree86 copyright will only be changed
with permission from the listed authors.

The license change will be fully effective as of the 4.4.0 release.
The initial draft of the changes will be included in 4.4.0 RC3
(4.3.99.903). A source diff showing the initial draft of the changes
is being made available for review with this announcement, and can be
found at . All XFree86
contributors are invited to review the changes, and notify us of errors
and omissions so that they can be corrected before the 4.4.0 release.
Such notifications, as well as comments about the licensing changes
should be directed to the [e-mail] list. XFree86 contributors
are also encouraged to review the license change, and let us know if
they wish to make similar changes to licenses in their name.

* XFree86 is a trademark of The XFree86 Project, Inc., and is pending
registration.

Re:Will we be stuck behind the times?

[thelasttemptation]

except if you notice that openbsd and gentoo, both source only compiles won't include it. The problem is the licence is simlar to bsd's but requires extra credit to be placed all over the distro. Anything that links to the xfree code that is gpled or uses a bsd licence is breaking the gpl/bsd licence because of the extra stuff they need to do in order to link to the libarys. In order to make a distro with 4.4 included, you would need to rewrite every app that links to X to the X licence, else you are breaking the gpl and as a distro maintainer, you'd be responcable for the breaks.

Re:What is the issue?

[Vaakku]

Here [slashdot.org] is a thread where you can read about why distro makers think that new XFree license is evil.

Re:The Question

[Asmodai]

The license says none of that kind. Did you even *read* the new license?

Labeling this as informative shows just how careful people read what they comment about.

Re:Why do they have a problem?

[evilviper]

so what exactly is the problem the distros have with the new license?

For one thing, any additional restrictions make a license incompatible with the GPL... That's enough to stop most everyone right there.

Re:Use Xouvert or FreeDesktop

[Anonymous Coward]

Given that Xouvert [xouvert.org] is practically dead, this really leaves FreeDesktop [freedesktop.org] as the only other alternative.

If you think this is flamebait, visit the projects' websites and take a convince yourself of the facts.

Re:This is what I can't stand about the open sourc

[byolinux]

They're changing their license, even though XFree contains GPL code.

They might well be writing FREEWARE (don't confuse gratis with freedom, there) but I think they probably would care if their code wasn't being used by anywhere near as many people as it was previously.

GNU/Linux wouldn't be without a graphical desktop, there's a wealth of projects out there.

Please take a look at http://www.freedesktop.org/ [freedesktop.org]

It's the advertising clause stupid!

[bsdnazz]

The xfree86 V4.4 license adds

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.

vs.

http://www.gnu.org/philosophy/bsd.html

Re:Doesn't seem as ugly as TeX's license

[Satai]

You're thinking of qmail [cr.yp.to]. TeX was put into the public domain; it's even been included in many commercial products. See TUG [tug.org] for more information.

Re:What is the issue?

[nickos]

Taken from a usenet post by Paul Cannon from linux.debian.legal on 2004-01-30:

The new license has a reworded disclaimer, and a numbered list of terms instead of the terms simply being stated. It goes farther than the old one in specifying that the conditions apply to binary distributions as well as source.

The change that causes problems is the addition of the third condition:

"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."

Several posters on slashdot and elsewhere have mentioned the similarity between this and the old, obnoxious BSD "advertising clause":

"3. All advertising materials mentioning features or use of this software must display the following acknowledgement:
This product includes software developed by the University of California, Berkeley and its contributors."

The FSF is quite clear (see here [gnu.org] and here [gnu.org]) in that they do not consider licenses with the advertising clause to be compatible with the GPL. In addition, the same reasons they give appear to apply also to the clause added by the XFree86 folks. That is, one cannot distribute something under the GPL with added restrictions like the one above quoted.

Re:Use Xouvert or FreeDesktop

[IntergalacticWalrus]

"But the FreeDesktop libraries will be LGPL whish is much much worse, because then we will have ZERO comertial applications." No. GTK and GNOME's libs are all LGPL, yet we see plenty commertial (sic) apps. The problem is with GPL libs, ie. Qt. "Free software faggots probably wont mind, but for those of us who work for a living the FreeDesktop initiative is really worrysome." Plenty of people make a living on Free Software, you know.

Re:The Question

[Anonymous Coward]

This sort of thing has happened before. Emacs/XEmacs, for example, and OpenSSH branching from the last open software release of ssh.com's original code, and RedHat's use of the "egcs" instead of the released gcc compiler.

It seems wasteful for XFree86: it's hard enough for us to integrate new video chipset drivers into our displays without additional X display code forking to cope with.

Re:Every cloud has a silver lining

[kinnell]

According to the gentoo list linked in the article, there is an effort to write a compatability layer for XFree86 drivers underway, so this should at least make it useable, if a little clunky. In any case, all the necessary information for writing drivers should be in the XFree drivers, so porting them should be a lot easier than writing drivers from scratch for someone who understands the code. I think it's more a matter of mindshare than anything else.

Re:What is the issue?

[ookaze]

I do not think there is any GPLed code in XFree86, I thought there was one licence for all of the project. OpenSource code perhaps, but not GPLed.
I say this, because RMS warned several years ago, I think, that putting your code under XFree licence (so it must have been mandatory) was putting it at odds, and was very dangerous, in case the project behaves badly.
That's just what happened.

You say there is chunks of code that have been copied from the linux kernel. I think you are talking about DRM, and I would say that was the other way around : DRI/XFree code was copied into the kernel.
XFree is so important, even I am considering contributing the little time I have to XServer sooner than I thought, if it does not see a flood of developpers in the two coming months (if this fiasco is not canceled, I mean).
I say "even I", because I have so little time on my hand, not for any other reason.

Y: Window system

[hatrisc]

Why this might be a good time to start developing Y: the successor to the X window system [ic.ac.uk].

Re:What is the issue?

[Gadzinka]

No, the code in question was sumbitted by its authors to XFree86 under the XFree licence. Remember, the author owns the code and he can licence it as many times as he wishes, every time with different licence.

So it doesn't make XFree86 ``gpl derivative''.

But those people who sumbitted those patches oppose changing the XFree licence on their code to something GPL incompatible. At least I've heard that Alan opposes, but I don't believe they asked all contributors if they agree to licence change.

Robert

Re:How exactly...

[bdeclerc]

is the advertising clause incompatible with the GPL?

Answer : Yes

More specifically: it is incompatible with clause 6 of the GPL, part of which reads:

You may not impose any further restrictions on the recipients' exercise of the rights granted herein.

Since their advertising clause is "imposing further restrictions" it is incompatible with the GPL.

Even more, Clause 2 of the GPL itself forbids redistribution of GPL'd code as part of a "Program" which also contains parts with licenses incompatible with the GPL.

Since parts of the current XFree86 are actually licensed under the GPL, the Xfree86 guys have two options :
a) remove all GPL-licensed code from XFree86
b) get approval from all authors of GPL-licensed code in XFree86 for a re-licensing of their code.

Considering that Alan Cox has already clearly indicated he will not accept relicensing of his code under something other than the GPL, legally the FXree86 people are already obliged to remove all Alan Cox's code from their relicensed XFree86 before distributing it...

The GPL isn't viral, if they release something which contains parts that are GPL'd, and other parts that are incompatible, those incompatible parts don't become GPL, but they do have to either remove the GPL'd parts, or relicense the incompatible parts under a different, compatible license (which may or may not be the GPL itself).

If they don't, they open themselves up for a lawsuit from the copyright-holders of the GPL'd code (which will probably be the FSF in many cases).

And this is why the Distro's don't want to touch XFree86 4.4, as soon as they distribute it, they themselves are doing the same, illegal, thing that XFree86 itself would be doing, and they would be open to lawsuits...

Re:The Question

[Asmodai]

``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.''

Yes, if you have end-user documentation it asks you to have a tiny line about how there's code from the XFree86 Project.

So where you get the idea from that it should be spit out during boot is beyond me. Yes, I know it can also be done in software. But that's done where normal attributions are normally kept, say an about box, or -EEK- perhaps /COPYRIGHT. And note that the software requirement is a MAY, not MUST.

Let me phrase a question back at you lot: "What is against giving credit where credit is due?" Because it looks like some common courtesy seems farfetched with a lot of people at the moment.

And by the way, it is similar to zlib's license, which is not mandatory, granted, but how many of you have actually credited Mark Adler and/or Jean-loup Gailly for their work?

Re:NVIDIA?

[Anonymous Coward]

Really? [ati.com]

Re:What is the issue?

[plcurechax]

The really interesting bit is that there is a lot of GPL-ed code in XFree.

I take it you mean FreeType which is included under a dual-license [freetype.org] of GPL and BSD-like.

Chunks have been copied from the linux kernel, and people like Alan Cox submitted patches

Alan Cox submitted patches [xfree86.org] are not under the GPL, but he wished to remain compatible with GPL applications (by using the old XFree86 license). The transfer actually has been from XFree86 to the kernel (fbdev) [mail-archive.com].

Re:So let me get this straight...

[Sancho]

No, they're having a hissy fit because they can no longer distribute binaries of GPL'd software linked to xfree86. This is because the GPL and the new xfree86 license are incompatible. The GPL clearly stipulates that additional restrictions on the distribution of the software is not allowed. Requiring attribution, as minor as this is, is an additional restriction.

XFree86 has never been GPL

[ink]

It used the old XFree86 license, which has historically been lumped in the BSD-ish license category. You could re-distribute the code as long as you kept source attribution and the license in place.

Re:NVIDIA?

[pyr0]

I never said they would cut off support. What I was implying was that perhaps X 4.4 will add some more extensions that NVIDIA decides to support with their future driver versions, rendering older X servers useless with the newer drivers.

Besides, if they are so concerned about alternative operating systems whey isn't there a linux-ppc or linux-alpha version of their binary drivers? I know if there was linux-ppc support I'd really consider getting some ppc hardware to run linux on. What is the barrier? Why can't they just recompile the driver for the different arch and release it?

Re:Debian refuses to include it?

[GeekDork]

Debian has 4.3 in experimental and it's working quite well. I wonder why they officially refuse to include 4.4 though, it's not as if it's likely to becone an issue anytime soon.

Re:Forking hell?

[Anonymous Coward]

On Mon, 16 Feb 2004, Theo de Raadt wrote:
[...]
> This is final; if that license stands, there will
> be forking.
>
> And if you don't like that, don't bother telling
> me. Tell them.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:The Question

[spiritraveller]

I suppose that the question here is: why? Is the new licence really that bad? Is this reaction warranted?

It is warranted.

Whether it's "that bad" doesn't matter so much as "is it really incompatible with the GPL."

If distributors would violate the GPL by linking GPLed programs to XFree4.4, they could be liable for statutory damages under copyright law.

One would hope that no author (of a GPLed program) would sue Debian (or others) for linking to XFree4.4. But hope doesn't pay the bills. Distributors need to comply fully with the licenses of the software they distribute.

Other than Debian, the distributors that have made this decision are businesses. It is not that they are GPL nazis... they are just dotting their Is and crossing their Ts.

Is the GPL license a problem?

[n0dez]

I just don't get it... there are lots of GPL incompatible licenses and nobody says anything about it. In fact, most Linux distributions ship that software. It seems that the only problem is the GPL license... am I wrong?


GPL Incompatible Licenses:

The original BSD license, the OpenSSL license, the Apache License, the Mozilla Public License, the Netscape License, the PHP license, the Apple Public Source License, ...

GPL Incompatible Licenses [gnu.org]

Re:Isn't this the end of the story, then?

[johnnyb]

You can mingle GPL and non-GPL code, as long as the non-GPL code is GPL-compatible. Therefore, you can mix BSD and GPL code, since BSD code can always be changed to GPL if someone wanted (hence it is GPL-compatible).

Re:The Question

[Nothinman]

That's the opinion of a few people in the GNU project, the GPL says nothing about what projects containing GNU tools have to be called.

Re:So let me get this straight...

[infolib]

there's probably GPL-ed code in XFree86

That would be copyright infringement. You cannot release other peoples GPL'ed code even under the original X license. (In fact you can only release it under the GPL). People may have contributed their own code that they released under GPL elsewhere, but they have then also granted it to the Xfree86 project under their license.

Re:Oh great, here we go...

[bfree]

You forgot one:

5) Has contributors tearing their hair out wondering what is happening their work and how, when and if it will ever be applied

Basically XFree86 is a closed development (you can watch their cvs commits) though it was releasing DFSG-free code. Now it is not going to be releasing DFSG-free code why would Linux distributions for one stick with it unless the act of replacing it was too hard (but it isn't, just take 4.3.99?)? Added bonus is they also get to setup a new open development model.

Re:Why is libjpeg not a problem?

[ajs318]

Because when you distribute an executable under the GPL, you are bound by the GPL to make an offer to distribute the source code. In effect you are distributing the source; it's just as though the recipient -- if they accept just the binaries -- has said "I don't really need that right now; hang onto it though, just in case I need it later, can you?"

Re:What is the issue?

[Bootsy Collins]

Why is this obnoxious?

What is the big deal about a few lines of giving credit where credit is due? I'm guessing from your response that it goes beyond that?

My understanding (which could be wrong) is this: it's not that that clause is morally or ethically a bad thing. It's that that clause makes the license incompatible with the GPL, which explicitly rules out putting additional restrictions on the redistribution of the code beyond those already in the GPL. It doesn't matter whether you or I or anyone else thinks the additional restrictions are reasonable; additional restrictions make the license incompatible with the GPL.

A problem with the GPL, then? I don't think so. How do you write a license that, in advance, imagines every possible restriction on redistribution of code and takes care of allowing reasonable ones while forbidding unreasonable ones? The ostensible purpose of the GPL is to preserve freedom, so that's the side the FSF wanted to err upon; so "no more restrictions."

So what if the new XF86 license is incompatible with the GPL? Well, that means that the redistribution of any GPL'd software that links against XF86 software (such as xlibs) is a license violation, and therefore illegal. So the redistribution of e.g. GNOME, KDE, etc., under these circumstances would be illegal.

So people are not upset that the XF86 folks (or, specifically, David Dawes) are making an unreasonable demand for credit for their work. They're upset that he's created this unsolveable license conflict where, previously, up until January 30th, none existed . . .a license conflict because of a license change which seems to be provoked by nothing (who, exactly, wasn't giving XF86 credit for their work), and which will likely take a lot of time and work (developing a new X server under a different license) to solve.

That's my understanding. If it's wrong, I hope someone who knows more about this will chime in.

Its Larger Than That

[EXTomar]

The problem with "the contributor clause" applied at the license level is that it is very pervasive. After all, why use a license that isn't pervasive?

So XFree has to list contributors. Anyone who writes an extention to XFree now has to list these contributors plus any contributors to make there extention work. Depending on how another developer uses that extention, they might have to carry all of those contributor forward as well. But a month has gone by since the original extention was released and more contributions have been made. To keep up to date not only do you have to sync source but sync contributors.

As for the GPL, the reason for incompatibility is that to use the GPL you must not put any more restrictions on the code no matter how innocent or benign you think they are. Sacrificing the freedom of the code just to make sure someone's name is plastered in the right spots is selfish. There are proper places to site contributsion. Just not at the license level.

This is the reason why it was a good thing that BSD abandoned this thing. I'm all for giving credit where credit is due but not at the license level. It gets to be a meta-maintaince nightmare to adhere to the license.

Re:This is what I can't stand about the open sourc

[infolib]

-1, wrong:
They're changing their license, even though XFree contains GPL code.

Where? Find one snippet, I challenge you! Incorporating GPL'ed code into Xfree86 under the Xfree86 licenses. [xfree86.org] (both 1.0 and 1.1) would be illegal.
Thus speaks the GPL:

You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.

Re:What is the issue?

[mindstrm]

It's more than that.

If Alan writes a driver, and puts it in the kernel under GPL.. Alan is still free to put that driver into XFree under the X license, and Alan is still free to sell that driver to Microsoft under whatever license he wants... because Alan holds the copyright.

Re:I can understand but..

[bobv-pillars-net]

From xc/lib/GLw/README.html [freedesktop.org]:

THIS SECTION CONTAINS MY PERSONAL OPINIONS AND DOESN'T REPRESENT AN OFFICIAL POSITION OF THE XFree86 PROJECT.

The first incarnation of this version of libGLw used eight header files from LessTif [lesstif.org], four for each Motif version. LessTif is covered by the GNU Library General Public License (LGPL) whose terms are not compatible with the XFree86 licensing policy. Since the copyright holder of LessTif is the Free Software Foundation [fsf.org] (FSF), I asked Richard Stallman, president of FSF and so called "leader of the Free Software movement", permission to redistribute a copy of those eight headers under XFree86 terms, still maintaining the FSF copyright.

Observe that I was not asking him to change the license of LessTif as a whole, but only to allow me to distribute copies of some header files containing function prototypes, variable declarations and data type definitions. Even so, Stallman said no because the files contained "more than 6000 lines of code". Which code? The LessTif headers are mostly copies of the Motif ones and don't contain any original GNU "code"! I can't still imagine a reason for Stallman's negative answer except for paranoia. He seems to ignore what Motif is and that LessTif's API is simply a copy of Motif's one.

After spending some time, I made my own headers, that became much smaller than the previous ones because I included only a subset of the Motif API and merged everything into four files: 417 lines instead 6000. Humm, perhaps I should be grateful to Sallman too :-).

Re:What is the issue?

[mirabilos]

They didn't neglect it being equivalent to the
4-clause UCB licence. They just said the wording
is different, and "if they would wanted BSD
licencing/spirit, they should have taken the
original UCB licence" (or something along the
lines; dig for the email in the other /. story
of yesterday if you need the exact text).

Re:Why is libjpeg not a problem?

[ampersandTHORN]

Because this requirement is not more restrictive than the GPL, because of the phrase "If only executable code is distributed" .

What's the big deal? READ THIS

[Anonymous Coward]

For God's sake people, stop asking the same question over and over. Read this:

[forum] [XFree86] Announcement: Modification to the base XFree86(TM) license.
Sven Luther [e-mail]
Sun, 1 Feb 2004 17:48:57 +0100

* Previous message: [forum] Re: Announcement: Modification to the base XFree86(TM) license.
* Next message: [forum] [XFree86] Announcement: Modification to the base XFree86(TM) license.
* Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Thu, Jan 29, 2004 at 11:58:38AM -0500, David Dawes wrote:
> Announcement: Modification to the base XFree86(TM) license.

Hello,

As discussed with David, i am taking discussion concerning the
problematics aspects of this licence change here. I think i understand
somewhat the reasons behind the licence change, but i wonder if all the
consequences of it have been thought of before doing the change.

Also, there are some confusing wording in one of the clause, which i
believe would best be clarified as to what the interpretations of them
by the XFree86 project are.

Also, first notice that my position is actually quite inconfortable,
since i am here mentioning the concerns of wider community and criticize
the new xfree86 licencing, in other forums, i usually do the opposite,
and take xfree86 side on this, so please do not react badly, and let's
have a rationale conversation about this, so that things can all be
resolved to everyone's satisfaction.

1) Possible confusion.

The following clause is the most problematic of all the licence, and as
such it would be nice to clarify it before starting a polemic about it.

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.

Ok, what does this mean exactly ? If there is a end-user documentation,
but it contains no third-party acknowledgement part, do you still have
to put the acknowledgement or not ? Also, is the choice between putting
the acknowledgement in the end-user documentation or the software a
choice that is free to make, or is the second an alternative only if
there is no enduser documentation. And what do you mean by in the
software itself ? If this software is a linux distribution for example,
would a file on the CD which is copied to the disk be enough ?

2) GPL incompatibility.

This selfsame clause is also the one which clashes with the clasue 6) of
the GPL.

6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.

And in the 'you may not impose any further restrictions' part. Since the
GPL does not force you to add acknowledgement in the end-user
distribution, then the clause 3) of the 1.1 XFree86 licence is indeed a
further restriction, which cause an incompatibility with GPLed software.
Now this is again modulated with the exact interpretation that is given
in the above point.

3) Where is the derivative work boundary ?

The problem is further muddled by the place where the boundary for
something being considered a derivative work. The GPL, contrary to the
LGPL, considers that everything linked with a another binary is a
derivative work

App Licenses must be compatible with their libs

[FreeUser]

Red Hat distributes Apache, OpenSSL, xinetd, all with GPL-Incompatible, Free Software Licenses What is weird is Apache claims their license is compatible.

Because they aren't linked together into one application.

Every XFree applications either links to an X library, or links to a library that links to an X library (insert as many levels of indirection as you wish).

Now that the x library licenses are no longer GPL compatible, every GPL X application is no longer legal for use with XFree 4.4. Which is the death of XFree, as we aren't about to throw out Enlightenment, Gnome, Mozilla, etc.

Far easier and more sensible to start using Xouvert or FreeDesktop than to dump millions of man-hours of contributed work simply to appease the vanity and anti-GPL zealotry of a few, regardless of how great their contribution was in years past.

In other words: License compatability between independent apps isn't an issue (each app's license can be adhered to independent of the others). License compatability between apps and the libraries they link to is absolutely critical, and XFree 4.4 breaks this with most of the applications that link to it. Which means Sianara XFree 4.4.

Re:Forking hell?

[Haeleth]

What are the chances someone will take 4.3 and fork it, and carrying on development as free software?

XFree86 4.4 is still free software. Yes, that's Free as in GNU. The XFree86 4.4 license is less restrictive than the old-style BSD license with advertising clause, and that license is included in the Free Software Foundation's list of Free licenses.

The problem isn't that it's no longer free, it's that it appears to be incompatible with the GPL. Different thing, different problem.

Re:Distro Maintainers making my decisions for me n

[Bootsy Collins]

While I can see the point of some of this, surely it isn't up to distro maintainers to decide which licences I can and can't accept for me.

No, but it is up to the distro maintainers to not break the law, which they would do if they distributed GPL'd software linked against libraries which are under a license that explicitly conflicts with the GPL. As has been explained by several people, multiple times in this thread.

Re:What's the big deal? READ THIS

[Anonymous Coward]

And Dawe's reply

[forum] [XFree86] Announcement: Modification to the base XFree86(TM) license.
David Dawes [e-mail]
Sun, 1 Feb 2004 13:03:28 -0500

* Previous message: Fwd: Re: [forum] [XFree86] Announcement: Modification to the base XFree86(TM) license.
* Next message: [forum] [XFree86] Announcement: Modification to the base XFree86(TM) license.
* Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Sun, Feb 01, 2004 at 05:48:57PM +0100, Sven Luther wrote:
>On Thu, Jan 29, 2004 at 11:58:38AM -0500, David Dawes wrote:
>> Announcement: Modification to the base XFree86(TM) license.
>
>Hello,
>
>As discussed with David, i am taking discussion concerning the
>problematics aspects of this licence change here. I think i understand
>somewhat the reasons behind the licence change, but i wonder if all the
>consequences of it have been thought of before doing the change.
>
>Also, there are some confusing wording in one of the clause, which i
>believe would best be clarified as to what the interpretations of them
>by the XFree86 project are.
>
>Also, first notice that my position is actually quite inconfortable,
>since i am here mentioning the concerns of wider community and criticize
>the new xfree86 licencing, in other forums, i usually do the opposite,
>and take xfree86 side on this, so please do not react badly, and let's
>have a rationale conversation about this, so that things can all be
>resolved to everyone's satisfaction.
>
>1) Possible confusion.
>
>The following clause is the most problematic of all the licence, and as
>such it would be nice to clarify it before starting a polemic about it.
>
> 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.
>
>Ok, what does this mean exactly ? If there is a end-user documentation,
>but it contains no third-party acknowledgement part, do you still have
>to put the acknowledgement or not ? Also, is the choice between putting
>the acknowledgement in the end-user documentation or the software a
>choice that is free to make, or is the second an alternative only if
>there is no enduser documentation. And what do you mean by in the
>software itself ? If this software is a linux distribution for example,
>would a file on the CD which is copied to the disk be enough ?

My personal interpretation is that the "software" is the actual binaries
containing the licensed code. Some software includes third-party
acknowledgments in an "about" popup. Some in a banner message at startup,
etc. I think "Alternately" is self-explanatory.

Regardless of the interpretation of this condition, condition 2, to
which I have seen no objections, requires that the full text of the
license be reproduced in documentation and/or other materials accompanying
the redistribution of binaries. That has the side-effect of reproducing
the statement in condition 3. It seems to me that if a redistibution
has no other third-party acknowledgements, then you're done. If there
are other third-party acknowledgements, then why is it a problem to also
acknowledge XFree86 and its contributors?

>2) GPL incompatibility.
>
>This selfsame clause is also the one which clashes with the clasue 6) of
>the GPL.
>
> 6. Each time you redistribute the Program (or any work based on the
> Program), the recipient automa

Re:What is the issue?

[cgreuter]

Why is this obnoxious?

I think what's going on here is that there's some confusion about what's a derived work.

The new license applies to XFree86 and anything based on it. This sounds fair until you realize that if you write a program that just uses XFree86, you are, technically speaking, combining your code with theirs by linking to the xlibs and including the header files. This means you need to credit them. For Mandrake to use XFree86 4.0, they'd have to go through the documentation for every app and library that uses the xlibs and add the attribution.

It violates the GPL because it adds extra restrictions on what may be done with GPL'd code. The GPL allows you to distribute a GPL'd program without the XFree86 attribution but the new XFree86 license doesn't.

My take on this is that it's unintentional. I suspect that someone didn't think through the implications and that in a couple of days, the XFree86 team will amend the license to say that just connecting to the server or linking against the xlibs doesn't constitute a derived work. At least, that's my hope. Otherwise, it could take years for a new de-facto X implementation to emerge.

layman's version

[Theatetus]

IAAL(ayman)

From the new XFree license (emphasis added):

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.

The GPL pretty much says that the only restrictions you can place on redistribution are the restrictions in the GPL. You can't add any other restrictions, however harmless they seem -- I guess the argument is that it's the start of a slippery slope, and IIRC some of the old UNIX systems had to have 3 screens worth of notices like that one.

I guess I can see RMS's point: if you let XFree require this notice, maybe Apache will want a notice too. And maybe Wall would want a similar notice for Perl. OK, that's annoying but we could live with it, even if we end up back in the UNIX days of multi-screen credit notices.

But then suppose NVidia releases a driver and says that you have to include an advertisement (not just a credit) for NVidia if you distribute it with the kernel -- or rather, they specify a credit message that many people would consider an advertisement. Well, now we've clearly crossed a line most Linux developers don't want to cross, but it's not exactly clear where that line was in the scenario I just mentioned. NVidia would say "all the other developers got to come up with their own credit text, why can't we say what we want to? If you don't let us have our free speech the terrorists have already won!"

And they'd have a point: if you let some people dictate terms to the GPL you don't have much grounds to keep others from dictating terms, and however innocent the first terms may seem, somebody will find a way to screw it up.

So, you have to kick the camel in the nose while that's still all he has in your tent.

Re:The Question

[Anonymous Coward]

How is mandrake and RH distributing XFree86 with their build requiring others to include acknowledgements?

Some of their software is a derivative of the XFree libraries and GPLed software. The GPL requires them to license it under the GPL, and forbids them from adding more restrictions to redistribution. The XFree license requires them to add a restriction, the so-called "advertising clause". Mandrake cannot comply with both licenses at once for the same piece of software.

Re:I can understand but..

[Unregistered]

They legally can't disribute it, though. All gpl code that links against xfree would need to be made compatable with the new xfree liscense which would in turn violate the GPL. That's why this is such a big deal.

Money. Not simply recognition.

[runswithd6s]

The problem isn't about listing contributors or copyright holders in a reasonable manner. The problem is with the "advertisement clause". With the exception of Public Domain licenses, in which the copyright holder waives his or her rights, or legal transfers, copyright of a work does not transfer with the work. Therefore, even in GPL compatible licenses, the original copyright holder MUST be listed in the copyright notices for the software. It doesn't mean that every copyright holder must be mentioned in ADVERTISEMENT for the software.

This type of clause is not there for legal reasons, rather for ego boosting and free advertisement. "I wrote this, therefore you must include me in all public advertisement." It burdens any distributors to include the advertisement clause with any archive, CD image, or boxed sets they make available to the public. This advertisement will cost money somewhere along the chain of disbursement. Whether it is in the cost of publication or the number of bits downloaded on the Internet, money is the issue. So, XFree.org expects distributors to foot the bill for advertisement of XFree.org's product. Do you now understand why distributors are rejecting the software?

Re:What is the issue?

[nestler]

Yet we don't see Linux distributers refusing to include products with those licenses.

You seem to imply that the Linux distributors don't care about those licenses and their GPL interactions, which is not true. The problem is that GPL programs cannot be linked to code (like libraries) that has advertising clauses.

I know that at least Debian is very careful to respect this restriction (GPL Ethereal is not linked with OpenSSL even though it loses functionality).

The real issue why this is a problem is that a lot of XFree86 is in the form of libraries (xlib). So any app that needs to link xlib cannot be GPLed. This screws up Qt, KDE, and many other things (basically any GPL app with a GUI).

It isn't as much of a problem with the things you name for a few reasons. A lot of old BSD licensed code got a licensing change (from the Regents) that removed the advertising clause. More importantly, a lot of the things you name (Apache, Mozilla) are not libraries so you don't have as much of a problem there.

This is a big problem and David Dawes and company have just made themselves irrelevant. They will have to back down (revert the license) or their project will be ignored from here on out. Distros will adopt other projects or do their own work on XFree 4.3.

Re:Is the GPL license a problem?

[forlornhope]

It is only linking between software with these licenses that is the problem. With most of the licenses above there are ways of getting around it. The problem with the new xfree86 license is that you have gpled software such as gtk and qt that are going to be directly linking with xlib. These plus the various window managers and other applications that link with xlib and are gpled are probably 99% of the use of xlib. I havnt verified this, but Im willing to bet that only 1% of the software that currently links with xlib in debian could keep doing that with this new license. That is why all these distributions will not be shipping xfree86 4.4. They couldnt use it and it would be a waste of packaging.

Re:What is the issue?

[Znork]

"Yet we don't see Linux distributers refusing to include products with those licenses."

It's no problem as long as you arent shipping any GPL applications linking with the incompatible code.

There are a whole bunch of GPL licensed applications that link to the X library code. The desktop environments, for starters.

So the choice is between not shipping the new version of X or not shipping any GPL licensed applications that use X. So you can have new X but no Gnome or KDE, or you can have the older revision of X and KDE and Gnome. Which would you prefer?

It's not that Redhat and the others have to insert acknowledgements, it's that they're not allowed to distribute GPL licensed components linked to code that requires them to insert acknowledgements.

Of course, there are ways around the problem. One I've seen would be to distribute the X libraries of a forked X linked with the GPL applications, and distribute the new X server as a completely standalone X server (which is how you can distribute commercial X servers with GPL applications). That would require double installs of a lot of things but it would probably be legal.

Summary of the license situation

[falonaj]

To accept a less than 'free' desktop for the 'free' desktop movement would not make sense.

Richard Stallman regards both the new license and some of the old licenses used in the X Server as GPL incompatible but still free [xfree86.org]. Recently he successfully suggested [xfree86.org] a compromise with XFree86, which ensures that at least the the client side libraries stay GPL compatible [xfree86.org].

The new license is nevertheless problematic because it requires distributors to change the end user documentation and all "other materials provided with the distribution". These changes require a lot of time, which can be spend better by improving one of the forks.

Re:What is the issue?

[bartjan]

GPL, section 6:

" 6. Each time you redistribute the Program (or any work based on the

Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License."

The combined works of GPL'ed code plus code under this new X license has the further restriction of not being able to distribute it without those advertisements. So linking GPL code to XFree86 4.4 means you are breaking the GPL.

Re:What is the issue?

[Anonymous Coward]

DRM is the Direct Rendering Manager (which I think is in the kernel). DRI is the Direct Rendering Infrastructure, which is the whole thing. The DRM is part of the DRI.

Re:Is the GPL license a problem?

[dgatwood]

Umm... go read the FSF's website again. All of the licenses mentioned meet FSF's definition of free (early versions of the APSL notwithstanding). None of them are GPL-compatible, however.

There's a separate bucket for licenses that don't meet their defintion of free....

Re:layman's version

[j7953]

I guess I can see RMS's point: if you let XFree require this notice, maybe Apache will want a notice too.

Note that Apache does have a very similar requirement in the Apache License, Version 1.1 [apache.org]:

"The end-user documentation included with the redistribution, if any, must include the following acknowledgment: 'This product includes software developed by the Apache Software Foundation (http://www.apache.org/).' Alternately, this acknowledgment may appear in the software itself, if and wherever such third-party acknowledgments normally appear."

However with Apache, it isn't a huge problem because the requirement has existed since day one so all of the Apache-licensed software has always been incompatible with the GPL. So anyone who wanted to write software and publish it under the terms of the GPL (and cares about copyrights) simply wouldn't use any of Apache's code. Or the author would add an exception to their license, something like: "This software is licensed under the terms of the GPL, with the exception that if you link this software with Apache code, redistributions are subject to the additional restrictions found in the Apache license." This would make it legal for distributors to attach the additional restrictions to the code. (Please note, I am not a lawyer, so please ask the Apache folks if you actually intend to use such a license.)

XFree86, unlike Apache, has not always had those restrictions, and there is a huge number of GPLed works that require linking with XFree86 code. Those works (usually) don't have any exceptions to the GPL in their licensing statement, so it is illegal for distributors to ship the software with the additional restrictions that the new XFree86 license requires.

Also note that the Apache Software Foundation has weakened their requirement for acknowledgement considerably in their new version 2.0 license, and as far as I know it is now considered compatible with the GPL by the FSF except for the newly added patent-defense clause (which the next version of the GPL might also have, so the licenses would be compatible then). The XFree86 project is moving backwards with its new license.