Debian GNU/Linux to Declare GNU GFDL non-Free?

Syntaxis writes "There's some considerable argy-bargy in progress over whether or not GNU's own GFDL is a Free documentation license at all. At issue are "invariant sections" which cannot be removed from derivative works. Check out the thread culminating in the proposed motion to take action. The current consensus on Debian-legal does indeed appear to be that one of the FSF's own licenses is non-Free under the terms of the Debian Free Software Guidelines! Well, documentation for GPLed projects countermanding the very freedoms embodied in the GPL certainly seems insane to me."

how?

how did this happen? a slip in the legal department?

Re:how?

The GFDL's Preamble states: We have designed this License in order to use it for manuals for free software, because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does.

But the reality seems to be that Freedom [gnu.org] to the FSF only really matters when it comes to software. A quick look at the FSF's audio section [gnu.org] shows that their interpretation of Freedom doesn't extend very far in other areas. Would software released under a license that allows "verbatim copying and distribution" be considered FSF free?

Debian takes a broader view [debian.org] that everything in the distro should be "Free". It may sound a bit anal to expect that manuals, audio and graphics should be covered by the same rights to modification, but the sad fact is that it's not just an academic point. Quake2 may be GPL software, but the graphics, music, etc are not covered by the GPL. Since Debian groups software into Free and Non-Free sections, it's important that the distinction is pointed out...regardless of whether it's Quake 2 or GCC.

Re:how?

You'll see that the FSF is concerned with free documentation [fsf.org] as well. The problem here is that some people are misunderstanding the invariant section provision of the FDL. As stated in that link, not every piece of writing is the same thing as software. The FDL insists that all the technical instructions be freely modifiable so that someone who creates a derivative piece of software can also modify the manual to keep it accurate.

However, some parts of a manual might be literary or express an author's opinion. This might be a political opinion ("software should be free") or it could be a technical opinion ("monolithic kernels suck"). But whatever it is it doesn't make sense for the creator of a derivative manual to change those opinions--that would be lying about the original author's intent.

The FDL recognizes that an author may have the need to guard these sections (remember, they can't have anything to do with the instructions to use the program). It doesn't make the manual any less free.

Is this the part where we find out that

RMS and Darth Stallius are the same person after all?

I like this line

"This is the stuff of which nasty flamewars and misspelled Slashdot headlines are made"

Same with GPL

The GPL can't be modified, yet it is stuck into just about every free package in Debian. If it could be changed, then the software's license could be altered. Bill Gates whould just have to "embrace and extend" the GPL to gain whatever control he wished. We NEED non-free pieces to protect the FREEness of the software. Ironic no?

Also the book "Steal this book" should be banned for false advertising.

Re:Same with GPL

Either you're trolling, or you don't understand the issue. Of course you can't change the license; that would make no sense. The issue has to do with the fact that the GFDL allows portions of the licensed document to be marked "invariant", meaning you can't change those parts. This is logically equivalent to what you would have if the GPL allowed authors to mark parts of their source code as unmodifiable. It essentially means that the document is not entirely free, but only mostly free. The debate within Debian has been about whether "mostly free" is good enough.

Re:Same with GPL

In order to avoid this, invariant sections must not be about the subject matter of the documentation (e.g., in the mathematics text, they could not be about mathematics). As to why the GFDL provides extra protection for content which does not actually belong in the document, I have no idea. Somewhat ironically, the FSF has a diatribe which makes the same point you've made, which is an invariant section in the gdb manual.

The FSF has an unfortunate habit of trying to make a soap box out of any piece of text they touch. In fact, the GPL itself states that programs under the GPL should, in addition to anything else they might do, display sections of the GPL (which, according to the license of the GPL itself, would mean that they would have to show the whole thing, since you can't modify the GPL), and it includes a substantial amount of text which is not actually part of the license and which the author of the software may not agree with.

Debian has some weird licencing rules.

I think Debian take all this licencing stuff far too seriously. They have some insane clause about the kind of software they can include in their distros. Personally, I think they should include anything that they are legally allowed to.

Re:Debian has some weird licencing rules.

You obviously don't understand what the Debian project is all about. It isn't just a Linux distro; it is specifically a free software distro. They don't want anything that isn't free-as-in-liberty. That's why they have the licensing rules that they do.

Re:Debian has some weird licencing rules.

I'm sure [Debian is] technically quite excellent, but all I need is to get my code written and my research done. From my perspective, as long as I can burn a RedHat CD and install it on as many computers as I want, it's no less "free" than Debian, regardless of whether or not it includes Netscape 4.

Which is perfectly fine, I'm sure, with most Debian developers and users. Use RedHat, by all means. It is good enough for personal use, maybe even as a server if you don't have to run tens of boxes (and maybe even for that, nowadays, I've heard good things about their network upgrade system, "RedHat Network" or what it's called).

As for myself, I use Debian mostly because I like its quality and stability, its reliable maintenance infrastructure (having to maintain a lot of servers makes oneself very partial to that apt-get, dpkg-reconfigure thing), and the overall sense of order that I draw from their packaging process.

Having said that, I would be more careful than you before dismissing the ethical underpinnings of the Debian organisation, or even stating them as a reason for not using Debian. That is the point in your reasoning that I'm having most trouble dealing with, akin to not taking free Unicef mugs and postcards because of all that human-rights agenda and stuff behind them.

Once upon a time, I was a rabid Amiga enthusiast. I learned a lot of things on that cute little box. Particularly, I learned to work the Video Toaster and the included Lightwave 3D. I learned it so well that I started making a living off it, doing video processing for a small publicity agency and an industrial design department at my U. Then the Amiga died, and having all that skills (well, most of them) suddenly became as useful as being the world's greatest kazoo player: who gives a shit? I had to learn something new to stay in business, because the company that made the tools I used stopped making them (or, at least, I became unable to buy them, which amounts to basically the same thing).

The same thing happened a couple more times with other commercial products (say Borland's C++ builder, Cisco's Netsys software, maybe Sun's Java in the not-too-distant future). So now I'm predictably more cautious when choosing what tools to spend time learning and using, for both work and play. Open source stuff doesn't die --not unless it really needs to die because it is replaceable with something undisputably better, and also free. And even so, nothing really dies until there remains absolutely no one still using it (I also can pull from memory several first-person examples, such as Dumpleton's OSE library, the GNU Pascal Compiler... hell, the wonderful Nethack).

Back to Debian: I appreciate greatly that the dudes putting toghether this wonderful distro are so picky regarding the license of the software (or, in this case, documentation). It saves me the work of doing that myself. I know that, for everything I apt-get install, I can spend my time learning every detail without worriying about that knowledge becoming useless and obsolete anytime soon.

So, wrapping it up: I'm also the kind of guy who looks for the tool that works. And I also place highest on my priorities to "get things done", rather than some abstract philosophical issues. But I deeply thank Debian for being so strict, even pedant, about legal issues. There is no other computer system that makes me feel as comfortable about spending any amount of time with, than a Debian system.

If you want true open source on anything

Stick to public domain. GPL is no more free than Microsoft, just each end an extreme. Microsoft, no source, no tampering, no nothing. GPL, Always source, no matter what. Sure it works for some, I have a few projects I own that are GPL, incidentally because I originally unknowingly GPL'd them, and a few now that are public domain. I prefer public domain, there is little to no worries at all on it. I'm even more free in the choice because I could one day take works I put into public domain and use them in a closed source application, such as for consulting work. People will benefit from the source I had originally made, and I benefit in the fact I can use the work in closed environments.

Re:If you want true open source on anything

In case you didn't realise, any works completely of your creation can be used in any way you like, as you're the sole copyright holder.

So the only case where you wouldn't be able to take a work you've released under GPL and include it in a closed source application is where you've either (a) originally taken source that was under GPL or similar and added to it or (b) applied patches from people where those patches were supplied in the understanding that the resulting app would be released under GPL or similar license.

In other words, your comment about releasing your own works into the public domain because it gives you more freedom are wrong.

NO, sorry

I'm not a GPL zealot by any means.. however..

TO start with, both Microsoft's code, and, say, RMS, are protected by copyright laws. In that, sure, they are the same.

Microsoft, however, makes you agree to a bunch of additional terms above and beyond the protections it would be provided under just copyright law. Stuff like "no reverse engineering" "No benchmarks" "not transferrable to another system" etcetera. You get the idea.

RMS code, released under the GPL, does NOT require you to accept ANY license at all. The GPL is NOT a use license.
You are free to do anything with the code that standard copyright laws allowed.
IN ADDITION to that, you can choose to accept the terms of the GPL, which grants you additional rights ABOVE and BEYOND what copyright alone allows you to do. You are still free to contact the copyright holder and request other licensing as well.

So it all really depends on what you mean by freedom. I agree, real freedom would be simply releasing it into the public domain, where anyone can do anything at all with it. The GPL is just pushing an agenda.

Just plain wrong.

Stick to public domain. GPL is no more free than Microsoft, just each end an extreme. Microsoft, no source, no tampering, no nothing. GPL, Always source, no matter what. Sure it works for some, I have a few projects I own that are GPL, incidentally because I originally unknowingly GPL'd them, and a few now that are public domain. I prefer public domain, there is little to no worries at all on it. I'm even more free in the choice because I could one day take works I put into public domain and use them in a closed source application, such as for consulting work. People will benefit from the source I had originally made, and I benefit in the fact I can use the work in closed environments.

The problem here isn't the GPL. You just don't understand it right.

If you put a project under the GPL, you can still use your own code in a closed project. Since you own the copyright, you can release the code under multiple copyrights. Releasing it under the GPL is actually better because it means no one else can use your code in a closed product.

Stick to public domain. GPL is no more free than Microsoft, just each end an extreme.

This really makes me wonder if you're trolling. This statement is just silly. With MS software you have the rights granted to you by copyright law, but they are restriced by a license. With GPL'ed software you get the rights granted to you by copyright law, plus additional rights are granted to you is you agree to the GPL. The GPL does not attempt to remove any of the rights granted by copyright, it actually gives you more. MS licences try to remove rights granted to you by copyright while not giving you any more.(BTW, this makes the GPL perhaps the most legally binding of software licences.) This is like saying that $1,000,000,000 is no better than $.01 because they're different ends of an extreme. It's nonsense. It's like saying facism is no different than democracy because they're both forms of government.

I like his sig:

G. Branden Robinson "To be is to do" -- Plato
Debian GNU/Linux "To do is to be" -- Aristotle
[e-mail] "Do be do be do" -- Sinatra

Lack of pragmatism

People who choose the FDL usually know little about its specifics -- they just want their documentation to be available under a copyleft license. Thus, virtually all FDL documents are free of invariant sections. Instead of condemning the license outright, a pragmatic approach would therefore be to define a threshold of invariant content (say, 20%), after which a document is no longer considered free. The FDL is the standard license for free, open content (see also Wikipedia [wikipedia.org], perhaps the largest open content project), and ostracizing it entirely will get us nowhere.

Upon reading the post, however, what I see is a bean counter mentality that can really be dangerous to open source projects as a whole. I shudder at the thought of hundreds of package maintainers being contacted to deal with this "license issue", which is really a non-issue to anyone with some common sense. This time would better be spent working on real problems -- it's not like Debian has none of those ...

Re:Lack of pragmatism

a non-issue to anyone with some common sense
Yep! As the author of a couple [lightandmatter.com] of [lightandmatter.com] GFDL'd books, this whole thing makes me cringe.

The idea of invariant sections is a very reasonable one. For instance, if I write a book with a dedication to Martin Luther King, I don't want someone else to come along and release a version where it appears that I've dedicated it to Adolf Hitler. Duh!

We don't live in a free-information utopia, and we don't even know what such a utopia would be like (although I'm pretty sure that in my utopia people won't be able to pull the King-to-Hitler switcheroo). So let's deal with reality. Maybe some of the people engaging in this silly debate should spend some time writing some documentation instead of arguing over licensing. This kind of over-zealous ideological navel-gazing is really pathetic.

Allowing removal would be enough.

Imagine this scenario: someone takes your book which is dedicated to Martin Luther King, then adds a lot of useful stuff to it, and then attaches a rant explaining how Adolf Hitler is much cooler than King and the original author was a wuss for choosing the wrong person to dedicate the book to. And then makes that rant an Invariant Section.

Now you can't use any of that new documentation in the next version of your book, unless you're willing to take the pro-Hitler rant as well. But the other person is free to take your new material, as long as he is willing to take any new Invariant Sections you add. This is the opposite of what copyleft is supposed to achieve.

None of this would be an issue if the GFDL allowed removal of Invariant Sections.

Good for dcumentation bad for programs

There is a diffrence between documentation and software, a good documentation license need not
be good for software, and vice versa.

So I think the GNU documentation license is ok for
documentation, and should be allowed for
documentation, but NOT for software.

Clarification

Debian isn't about to remove all documentation licensed under GFDL, only the documentation that takes advantage of the invariant sections (or some other non-modifiable features of GFDL). Unfortunately this includes most of the GNU project documentation since the GNU project has marked the usual GNU propaganda blurbs invariant.

What's strange is that according to GFDL the invariant sections must not be about the actual subject of the documentation. Instead must be "secondary sections", as described in the GFDL:

A "Secondary Section" is a named appendix or a front-matter section of the Document that deals exclusively with the relationship of the publishers or authors of the Document to the Document's overall subject (or to related matters) and contains nothing that could fall directly within that overall subject. (Thus, if the Document is in part a textbook of mathematics, a Secondary Section may not explain any mathematics.) The relationship could be a matter of historical connection with the subject or with related matters, or of legal, commercial, philosophical, ethical or political position regarding them.

The "Invariant Sections" are certain Secondary Sections whose titles are designated, as being those of Invariant Sections, in the notice that says that the Document is released under this License. If a section does not fit the above definition of Secondary then it is not allowed to be designated as Invariant. The Document may contain zero Invariant Sections. If the Document does not identify any Invariant Sections then there are none.

Frankly, it seems to me that the GNU project would have added the invariant sections only force their political statements to be carried everywhere along the documentation. Many people have pondered that if they request the operating system to be called GNU/Linux, why don't they add a clause in their license to demand that. Well, maybe they have started moving towards that direction.

Correct me if I'm wrong

As I understood it, the license allowed for invariant sections so that you can include "Originally written by..." at the top and then prevent anyone from changing it; I've always intended to distribute laboratory manuals under this license, and I can think of some other things I might want to make invariant, primary data for instance. In source, you might conceivably use such a clause to require someone to include a Trojan (or, gasp, an unfree supplementary component!), but since it is a document we're talking about, I do not see the problem. If I decide to make false or misleading text invariant, why use my document (or fork of a document) at all?

If I understand correctly, absolutely nothing prevents you from adding entire additional sections to the document - including, if necesarry, screaming tirades against sections you were forced to include.

Let me put it another way - I release the documentation for my software under this license. What invariant text could I possibly add that is genuinely going to interfere with someone's free speech?

Oh, the Irony.

Isn't funny that RSM keeps hammering the difference between free (as in beer) software versus free (as in speech ) software; claiming how important and critical the latter is...

And yet the GFDL is the perfect illustration of that analogy; upside down! That documentation is very free (beer), but specificaly prohibits modification of the invariant sections which is limiting speech.

Sorry, Richard, can't have the cake and eat it too. You figure it's evil for someone to publish code not wanting someone else to fiddle in it, but you figure that as long as you feel it's important to you, documentation can be so protected?

-- MG

No problem

Sure, you can't modify the text of an invariant section, but with a little ingenuity you can completely change its meaning.

For instance: after a glance at the license, it doesn't say you can't move the invariant sections. Think of the possibilities!

Also, say paragraph 1.2 is an invariant section. Just add section 1.1:

1.1 The content of section 1.2 is an obsolete project policy, reproduced here for historical purposes only.

1.2 <invariant section...something about making source available or some such...dunno>

Stallman doesn't believe in total freenes

I just went to a talk given by RMS here in Austin at SxSW [sxsw.com] where RMS spoke on copyright.

For at least half of the talk, he spoke regarding the history of copyright and was absolutely boring at all hell (perhaps it's because I only have Lessig's Free Culture [lessig.org] talk to compare to).

For the second half of the talk, he began to outline how he thinks the copyright office should work (he admits this isn't a perfect system, but he thinks this is how it should be). Essentially, he narrowed down all intellectual works into three catagories:

1. Functional
   These are works that serve some sort of functional use within society. This includes text books, manuals, and software. These works should be free as in speech.
   
   
2. Biographical
   These are works that are compliations of a particular authors opinions. RMS thought these could go either way. Maybe they could have a limited period of monopolistic power (of course no longer than 2 years).
   
   
3. Aesthetic
   These are works that only have aesthetic value (in other words, they are the shiny things of the world). Stallman stated that a copyright system should allow a 2-3 year monopoly on such works (this means the RIAA could still do all it does but that you'd be allow to trade songs that were 3+ years old).

Now, I believe there are some major holes in this, but I brought up the point that software licenses surely are functional works within society and therefore the GPL license itself (the actual document that you include with your software) should be free as in speech (it currently disallows derivative works).

Stallman had no answer for this and instead spent 15 minutes explaining to me why using the term "Intellectual Property" meant that I couldn't even begin to understand the issues at hand.

I've always been a defender of Stallman but I lost an awful lot of respect for him that night. I fully support Debian in this matter.

Is Debian the bad guy?

If I read the GFDL correctly an Invariant Section simply concerns it self with secondary sections. Secondary sections being sections which contains information about authors, publishers and so on. They are not allowed to contain content regarding the overall subject.

To me this seems fair. Invariant sections ensures that all people involved in the creation of this document is properly credited. When writing free documents all you have is the credit and the Debian people want to take that away from you. I must have misunderstood something, this can't be right. If it is then Debian just became the bad guys. When we're developing free software / doucuments / whatever the only thing we have is our name. We don't expect to get paid, but most of us would like the credit.

I'm sure I misread the GFDL, if not I didn't I'm really disappointed with Debian.

It's about time...

As I see it, the greatest thing about the Debian project is the fact that they don't subscribe to the typical herd mentality so often seen in the open-source community.

I've seen many, many Debian developers using "GNU/Linux" to describe the operating system, which does give credit where credit is due.

However, the GNU project's goals often frighten me (inasmuch as I give a shit), and it's nice to see that someone in the community is willing to point out their mistakes.

Many have pointed out that you could put the content of an entire work in an "invariant" section of a GFDL-licensed document. I believe there may be certain rules regarding the proportion of invariant sections to non-invariant sections, but defeating this is akin to defeating the Slashdot lameness filters: a definite time-waster, but not impossible.

The GNU Project is shady. Make no mistake about it: The GPL restricts choice as much as an NDA would.

I often wonder how many successful works the GNU Project could claim if it weren't for the restrictions inherent in the GPL. One oft-cited (but quite relevant) example is GCC: stagnation left many unsatisfied, so EGCS was started, blah, blah, blah. Basically GNU took (with permission) the work of those who had made EGCS a much better compiler, and renamed it GCC.

To contribute to GCC, in fact, it is not enough that you GPL your code and give a license to the GNU Project. No, you have to ASSIGN COPYRIGHT of the code to GNU, basically saying that the code is no longer yours, and that you would no longer have the right to take code from an existing work (such as a commercial compiler which you wrote) and contribute it to GCC, because you would no longer own the original code due to copyright violation.

Does this remind anyone of recording companies requiring artists to hand over their original works?

Everything done in a GNU project benefits the FSF (at the very least, with added prestige) -- they can claim that they, and they alone, own the code. This includes the right to, if they chose, hire coders to develop the HURD into a useable OS kernel (refer to my sig here), and release it under a closed-source license. Or, to make major improvements to GCC and sell it commercially under a non-GPL license.

If Walter Bright decided to allow the FSF to use major portions of his C++ compiler, which he sells commercially (and includes, I believe, much better support for C++ templates than GCC), he would have to assign copyright of his code to the FSF, therefore preventing him from using it in releases of his commercial compiler in the future.

The FSF is brain-dead, folks, and kudos to Debian folks for having the cojones to point out one of the more obviously stupid flaws in a GNU license.

(Many may note the fact that I focus a bit on compiler issues here. I have followed, to some extent, the GCC development lists, and from what I have seen, it can be a pain in the ass to contribute to GCC. Apple has many improvements to the compiler in their internal tree, and I often wonder if more of those improvements would have been rolled back into GCC by now if not for the hoops they have to jump through in order to get those changes submitted.

I've seen people make feature suggestions on the list which the Apple guys say they've already done and tested internally. The response is often, "We've done this, but we weren't sure if anyone else would find it useful. We'll look into getting permission to release it." It seems obvious that getting permission to hand over copyright would make that process a little harder.

Why do I focus on compiler issues so much? Various reasons... quality of generated code on Intel vs. other architectures, KDE slowness due to C++ linkages, blah blah blah. The compiler is key to getting code to run quickly on modern CPUs, as anyone pushing a non-Intel architecture would do well to remember.)

Don't trust the FSF. Appreciate their work, but don't hand over your firstborn. They can do whatever they w

Re:It's about time...

As I see it, the greatest thing about the Debian project is the fact that they don't subscribe to the typical herd mentality so often seen in the open-source community.

What Debian has is a set of clear guarantees that it promises to maintain to the community: the Debian Social Contract [debian.org]. Because of this, it cannot be beholden to political alliances, such as allegiance to one desktop project (vide Red Hat's closeness to GNOME) or even to its own ideological ancestor, the FSF. It has to operate by its principles, not by the opinions of whoever happens to be its leader at the moment.

One of the principles of classical liberal politics is to be ruled by laws rather than by men. In monarchies and oligarchies, the organizing principle of society is the leadership of a special individual or group: the king, the aristocracy, the ruling party, or what-have-you. Allegiance is to this leader, and alliances with other polities are founded on amity among leaders: hence the marriages of political alliance in medieval Europe. In liberal societies, the organizing principle is not the leader principle, but rather the basic law, or constitution.

In this regard, the FSF is in many ways illiberal: yet Debian, in so many ways the FSF's descendant, is thoroughly liberal. Debian is organized by rules, rather than by adherence to a leader.

The GNU Project is shady. Make no mistake about it: The GPL restricts choice as much as an NDA would.

This, however, is going too far [slashdot.org]. (For one, I think you mean EULA rather than NDA. NDAs aren't even related to copyright licensure; they're just contracts.)

Both EULAs and the GPL are founded upon the base of copyright, but from that base they build in opposite directions. EU