Why Should I Sign Copyrights To The FSF? 111
Honza Jirousek asks: "The issue of signing copyright for patches to GNU software or even whole new GPL-ed programs to the FSF came up several times in various dicussions, last time in recent cphack threads and in this Wired article on the same topic. Some indicate this may even be a requirement for accepting larger patches for some GNU programs, such as emacs. I never managed to find more specific information on that. Can someone explain this practice and point to more information? I understand the positive effects of this (FSF being better positioned to defend the copyleft, possibility to change the licence to newer versions of GPL etc), but I can also see an interesting side-effect - in some cases it effectively puts the FSF in the privileged position of sole copyright holder, who can re-license the code. This is similar to special provisions of 'original author' in licences such as NPL, often criticized exactly for this. Or am I getting it wrong? Contrast this to programs such as Linux kernel, where the copyright is so distributed, that re-licensing will never be possible." (Read on...)
Pay particular mention to the mention of section 17 U.S.C. 205e of the U.S. Copyright Laws. What does it actually do to the protections offered by the GPL and should we be worried?
Re:Sign if you want code undefended (Score:1)
You seem to be assuming that if the FSF has the right to defend the copyright, that they will do so. History has shown that signing over ownership just allows the FSF to sit on their asses as even more copyright violations of GPL and LGPL code takes place. I have heard of several cases of violation of the GPL and LGPL but I have yet to hear of a single case where the FSF has taken a company to court to get the problem resolved.
Take Tripwire Security [tripwire.com] for example. Since Janurary of this year, they have made a demostration [tripwire.com] of violating the Lessor General Public License (LGPL). Despite the fact that the LGPL has less demanding terms for redistribion of resulting binaries than the GPL does, Tripwire has refused to honor those terms and the FSF has failed to alter Tripwire's behavior.
Please consider the following issues:
LGPL requirements ..............| Tripwire Security
- -------- .....| No notice is provided .....| that the GNU libc is ....| included or that any ....| part of the Tripwire ...| package is covered by .......................| the LGPL
................................- -------- ......| None of the files in ..........................| the Tripwire package
................................ | contain the texts of
................................ | the LGPL
................................- -------- .......| Tripwire executiable .....| display a copyright ........| as part of a banner .........| but the Library copy-
................................ | right is excluded.
................................- -------- .....| Tripwire Security has ....| not provided the source ....| code, reference to the .....| source code or even the ...................| version number of GNU
................................ | libc used.
................................- -------- ...........| The Tripwire programs .....| are statically linked but ...| the object files or ...| source code is not ...| available even when the .....| company is requested in .......| writting to provide one ...........| or the other.
................................- -------- .....| Tripwire EULA states: ...| "You may not reverse ....| engineer, decompile, ........| translate, or disassemble .....| the Software..." ....| ............| ......| ............|
................................- -------- ........| Tripwire only supplies .......| the LGPL work as linked ........| with their programs. ....| The EULA states that it .....| may only be copied .......| for backup or archival ..........| purposes.
................................- ---------------
--------------------------------+----------------
Must give prominent notice
with each copy of the work
that the Library is used in
it and that the Library and
it's use are covered by this
License.
--------------------------------+----------------
Must supply a copy of the
LGPL.
--------------------------------+----------------
Must include the Library
copyright notice if during
execution any copyright
notices are displayed.
--------------------------------+----------------
The complete corresponding
source code for the Library
must accompany the program,
or a written offer for the
source code.
--------------------------------+----------------
If the program is an
executable linked with the
Library then complete object
code and/or source code must
be provided so that the user
can modify the Library and
then relink to produce a
modified executable.
--------------------------------+----------------
When a program is combined
or link with the Library and
distributed under different
terms, those terms must
permit the modification of
the work for the customer's
own use and reverse
engineering for debugging
such modifications.
--------------------------------+----------------
With each redistributed
copy of the Library, the
recipient automatically
recieves a license from the
original licensor to copy,
distribute, link with or
modify the Library...
--------------------------------+----------------
As I said before, each of these issues have been going uncorrected since Jan 2000. Since not a single issue has been addressed regarding this blantent demostration of disregarding LGPL redistribution requirements, I find it highly unlikely that the FSF has the resources to enforce their existing GPL packages (which has even more strict redistribution terms than the LGPL), and as such can not provide copyright defense for even more GPL packages signed over to them. Unfortantly, if you want your GPL terms defended, you will have to do it yourself!
Re:So they can sell it. Re:Sign if... (Score:1)
ChangeLogs (Score:1)
Re:Release of ownership NOT necessary. (Score:1)
I think you miss the point. FSF is working toward the goal of making all software free. Releasing free software and allowing it to be snapped up into a non-free project is... less pro-active... than writing a license that ensures its future freedom. The GPL is contagious by design.
IMHO, it's a pretty smooth move.
(Disclaimer: I've nothing against no-strings code, BSD-style licenses, etc. I'm just saying I can appreciate the slickness of the GPL, in light of RMS's philosophy.)
The most important reason of all (Score:1)
With such a significant value add, how can you not assign copyright to the FSF?
--
Re:The GPL is meaningless otherwise (Score:1)
A question worthy of discussion (Score:1)
Chas - The one, the only.
THANK GOD!!!
Re:Distributed Copyright? I don't think so. (Score:1)
Distributed Copyright? I don't think so. (Score:1)
I don't know where you got the idea that the kernel's copyright "was distributed". Linus holds that copyright. If you steal his code, he can come after you. Just because you GPL a program does not mean that you give up your rights to that program. If you send patches to the owner of that package, you DO give up your rights to that code.
Linus can say poof, the kernel is closed and any new releases will be under the Microsoft EULA. The only thing we could do is take the last GPL'd kernel and work on it from there. I doubt that will happen, because no one would use the new closed kernel.
Re:Gosh! (Score:1)
Re:Gosh! (Score:1)
I find the whole notion of "we know whats best for you, so let us make the decisions and give us rights" very limiting to my freedom, but I'm very interested in hearing why it may still be a good idea, like having their attorney's defend it for me, etc. All in all, its a very foreign concept to me of handing over control of my creation, but I suppose that lack of control in inherent in code thats public, no matter what the liscence - which may mean that handing over the copyright to FSF shouldn't be such a big deal to me. Either way I think its worthy of a discussion, rather than just reading one sided arguments from once source.
Broad misunderstandings (Score:1)
Assigning code to the FSF ensures that your software becomes part of the GNU system a modest but very capable suite of programs for running your computer.
No one is required to write for GNU, indeed people are encouraged to use software written for GNU on non-GNU systems.
Read the rest of this post (Score:1)
Assigning copyright to FSF means that the GNU system is legally defensible as a unit. It's the GNU system, copyright (c) Free Software Foundation. It gives cohesion to the entire system. And it provides a legal framework to protect the system under the terms of its release
I'm sure that all of us would welcome the situation that existed when software was shared as source without much thought, but that situation does not exist at the moment. Right now, writing software is subject to many laws. FSF should and does operate in the current legal context, as unpleasent as that may be.
[disclaimer] I work for GNU, I do not necessarily speak for GNU.
Re:Release of ownership NOT necessary. (Score:1)
The FSF would, presumably, do everything in their power to hasten the day when they become completely redundant.
Like Lenin's position that the tyranny of the proletariat was necessary to hasten the withering away of the state and the flowering of a state-less communist society?
xemacs protection (Score:1)
Re:Distributed Copyright? I don't think so. (Score:1)
He can, of course, issue those parts under a new licence if he so pleases, he can't do the same with the parts that belong to other people.
This could have helped Gnutella (Score:1)
Formal Logic vs. Law (Score:1)
The trouble is, you'd have to translate all the precedents (laws, judicial decisions, and common law) that it depends on or refers to as well. I'm pretty confident that the entire body of law is not logically self-consistent (in a formal, absolute sense).
Re:Feed Stallman's ego (Score:1)
I guess over the course of the last year, I have become one of the "suits". You wouldn't know this by looking at me, but it is the case.
I have taken research and development responsibilities for a fairly well funded company in need of technical direction. While not all of our projects are open sourced, some of our projects are based on open-source technology.
The older members of the executive staff are always looking for "patentable" material. Whether is be direct patents, or process patents. To date, and with my track record with the company, in perpetuity, at least my company should be able to do 2 things.
Abide by GPL and LGPL licensing.
Make money.
I have even started a program to contact the maintainers of any software we use, and provide whatever assistance we can to help ensure the continued development of the software.
In my limited experience with law, the "intent" of the document would be carefully examined. In the case of the GPL, this may be a powerful weapon against a company like Motorola.
In the case where they find methods to invalidate the license, there is always the possibility that they will unknowingly open a whole that could invalidate all the licenses currently held by Motorola themselves....
While I believe this is an issue worth watching, I'm not sure that it will be the end-all.
And, even as a suit, and with the problems that pertain with balancing my moral standards and fiduciary responsibilities, I always remember that our country was founded through revolution. And in the case that the extreme evil comes knocking, we will always have that recourse...
Currently, I like the musings of Jon Katz, and others of his ilk. I do strive to become like them, and hope that others like me will work to bring the "suits" and the "hackers" together in a working ground.
I've already had some luck in these types of endeavors...
Re:The most important reason of all (Score:1)
Hold your horses - nobody is calling the *kernel* GNU/Linux. That name refers to distros, built upon glibc and a whole load of other GNU software, e.g. Debian.
Re:xemacs protection (Score:1)
Re:Lame but perhaps common query? (Score:1)
Re:Sign if you want code undefended (Score:1)
Didn't they do that for some derivative of Emacs?
They certainly have persuaded many violators to back down without taking legal action. Their standing in the world means that their threats carry more weight than, say, threats made by me. And they have a lawyer doing stuff for them, which is more than I can say.
Re:Release of ownership NOT necessary. (Score:1)
Yep, or Mao's position, or even Senator McCarthy's. Quite right - the FSF have to be judged by their track record too.
Re:Two reasons (that I know of) to sign: (Score:1)
The GPL doesn't have to have a meeting of the minds in order to work. Refusal to accept the GPL is not a valid defense against infringement.
The reason why not is because the GPL grants conditional permission to do what would clearly be illegal otherwise. The GPL (and any other license based in copyright law (most commercial software, shrink-wrap licenses aren't)) is not a tool for the plaintiff in a infringement suit, but a tool for the defendant. It allows the defense to say "The plaintiff granted me permission, in the form of this license, which I complied with". If the plaintiff can show that the copyright was infringed, and the defense disclaims the validity of the license...it doesn't look good for the defense. The FSF or any other plaintiff doesn't even have to bring up the GPL if the defense doesn't mention it.
One more time: (Score:1)
Amen.
Obdurate RMS may amuse--but his word is on par with the laws of physics.
Re:Not only that... (Score:1)
treke
Can we get a laywer in here? (Score:1)
One good reason (Score:1)
Re:The GPL is meaningless otherwise (Score:1)
1) Microserf, in it's own spare time, goes out and writes, say, a device driver for the Linux kernel, GPLs it, and releases it into the wild.
2) Microsoft decides since it's employee created it, THEY own the rights to the software, and file suit against RedHat, Caldera, Linus, etc.
-or-
1) Microserf, in it's own spare time writes, say, a device driver for the Linux kernel.
2) Microserf assigns the copyright for said code to the FSF
3) Microsoft decides THEY own this code, but the FSF says "Sorry, it's ours, here's our piece of paper!" ?!?!
Your basic premise seems to be along the lines of, "Well, I don't have the right to let you use it, but I DO have the right to give it to you free and clear."
This is non-sense. Either MS owns the code, in which case both the GPL AND the copyright assignment are invalid, OR MS does NOT own the code, in which case both the GPL AND the copyright assignment ARE valid.
This does not seem to be a situation where you can mix and match. If I'm misunderstanding your point, please tell me. This is bugging he hell out of me now, and I'd like to understand your reasoning.
Re:The GPL is meaningless otherwise (Score:1)
On that note, I'll cede the argument to you. I should have known better than to try to inject logic into a conversation about law, and I truly and humbly apologize.
Re:The GPL is meaningless otherwise (Score:1)
The Linux way leaves us open to blackmail or destruction by ANY malicious company whose employee didn't double check his contract and wrote kernel code. The FSF way guarantees that the code will always be free
I really have to take issue with the above statement. In the first paragraph, you claim that Linux (GPL, with no copyright assigned to FSF) is vulnerable to manipulation if some large company whose employees have written code for the kernel decide to take issue with that fact and claim ownership. Your second paragraph goes on to claim that assigning the code to the FSF keeps this from happening.
Pardon me while I say "huh?" If someone is claiming something is being illegally distributed, it does not MATTER that the copyright is assigned to FSF. By the same logic you apply in paragraph one, the programmer who had no right to distribute it into the linux kernel HAS NO RIGHT TO ASSIGN COPYRIGHT TO FSF.
Re:Gosh! (Score:1)
--SolidGold
Re:Motorola (Score:1)
In Moto's case...they bought the rights.
Re:Motorola (Score:1)
Then obviously you have never had to make sure these beasts keep running and current.
>rumors, rather than verificable data.
Considering I've SEEN this 'data', why don't YOU verify it? Go buy a phone switch.
Re:Release of ownership NOT necessary. (Score:1)
The problem is -- how do they do that for software with which they have no legal connection? Party C can't just march into court and demand that Party A stop hassling Party B. For illustration, use A = Juan Miguel, B = Elian, C = Marisleysis. It's up to A & B to work things out. Therefore, the FSF can help more effectively if they are (in whole or in part) Party B.
Re:The GPL is meaningless otherwise (Score:1)
You seem to be confusing binary logic with reality -- "either MS owns the code...or MS does NOT own the code".
In reality, whether MS asserts ownership is orthagonal to whether the FSF acknowledges that ownership, which is in turn orthagonal to how the courts would rule in the end, which is in turn orthagonal to whether the courts would issue an injunction against further distribution by the FSF prior to the trial (say, in response to a motion for preliminary injunction by MS).
If that statement doesn't make sense to you, and if my lawnmower analogy doesn't help (see one of my messages from yesterday), then it's unlikely I'll be able to educate you via /. posts. Take some law courses, consult a good IP lawyer who can teach these basic concepts, etc.
But I will try one more tack: the idea of a "universal truth", which is the context in which you've written your objections to my logic, is one that is less and less acknowledged, IMO, in today's US judicial system (as well as its politics and populace)...
In reconciling the different views, or interpretations, of that one "universal truth" (say, about the legality of the FSF's continued use of the Microserf-authored code fragment), the courts would treat a signed release such as the FSF demands as being worth more than its weight in gold, when it comes to protecting the FSF's ability to continue distributing that code fragment, avoid punitive damages, and even avoid a judgement against it.
Re:The GPL is meaningless otherwise (Score:1)
Re:xemacs protection (Score:1)
I thought there was also some sort of problem with the original redisplay code, but I can't find a reference to that.
Not only that... (Score:1)
This is that old communist thing, you don't need that fancy car, so give it to me, so that it does not get stolen, we will take care of it.
Erm... (Score:1)
I also think that distributing the copyright and releasing it under the GPL for keeping it as sourcecode are 2 different things. Perhaps some people feel the urge to distribute the copyright of their material to the FSF, but I personally can't see any reason for that: it doesn't matter if you do that or not if you use the GPL for your sourcecode. It only has affect on your rights on the sourcecode YOU wrote if you want to take it OUT of the community and do something else with it. Distributing the copyright to other people makes you unable to do that. (No, that's not evil, think! you can come up with some situations wherein this can be useful).
Freedom, that is not equal to signing away your own property to others.
--
Re:Release of ownership NOT necessary. (Score:2)
I guess the "Free Software Song" just didn't prove as seductive as he'd thought it would be, so he had to switch tactics.
Details? (Score:2)
Re:Copyright assignment considered harmful (Score:2)
The assignment isa paper form signed by both parties.
Re:Motorola (Score:2)
> > switches? And why should "the compiler and
> > tools" be part of such a switch?
> Then obviously you have never had to make sure > these beasts keep running and current.
You misunderstood my "should". I didn't ask why source for the cell phone switch would be *nice*, but what obligation Motorola have to provide it. Be doesn't provide source for BeOS, even though they include gcc.
> Considering I've SEEN this 'data',
> why don't YOU verify it? Go buy a phone switch.
Why should I? I have no need for a phone switch, and you have presented zero evidence that there is a problem. Would you spend *any* money verifying *any* claim from an anonymous
Don't sign if you are a sue-happy (Score:2)
> Emacs?
Nope. The FSF has always managed to resolve the violations without going to court. Not always as fast as some of the sue-happy
Motorola (Score:2)
Re:Motorola (Score:2)
It doesn't even make sense as stated. Why should there be source for the cell phone switches? And why should "the compiler and tools" be part of such a switch?
The actual problem is
Two reasons (that I know of) to sign: (Score:2)
The other reason I've heard is that there must be something in writing in order for the license to have any legal standing. What that must be, I don't know, but presumably signing the copyright over under a particular license would do the trick.
My understanding (IANAL, of course) is that the problem occurs because the GPL isn't an agreement with a particular person. Because of this there's no document that represents the license agreement, and so there's no legal standing for the license. Hopefuly, a court would realize that all parties (well, most all) are well aware of the license, and that there is a meeting of the minds. But who knows whether that's even relavent in the twisted mind of the court.
Since I'm not really sure if there is a reason behind the legal fact (or if it's just a historical artifact), I don't know whether something like having a copy of the license signed & dated by a notary public would help.
Re:FSF Assignment How's and Why's (Score:2)
Actually, while it is ironic, yes, it makes a lot of sense. GNU is I guess somewhat against copyrights, hence copyleft. It is mentioned in numerous places on www.gnu.org that since copyright seems to be the rules of the game right now with respect to IP, why not use copyright to protect your work? In fact, that's probably the only feasible option.
It is ironic, but it isn't hypocritical - if they didn't do it that way, it'd be pretty much impossible to make the GPL stick.
Assigning allows for re-licensing the code... (Score:2)
As multitude of other people have pointed out, by assigning your patch/code copyright to the FSF, the FSF retains full ownership of the software.
I'm not going to go into the aspects of using the FSF as a big stick to hit GPL infringers with. Instead, I'd like to point out something that people seem to miss w/r/t this whole thing:
By assigning your copyright to the FSF, you allow them to relicense the code anyway they want, with no input from you. Now, I'm not screaming blue bloody murder here, it's just something that people should be aware of.
The FSF sometimes will relicense code to companies which would like to use the code, but can't take the GPL (for one reason or another). A historical example was Motorola. They licensed gcc from the FSF under some closed-source terms, and paid big bucks to the FSF for it (it was close to $2million, IIRC). Those terms included the stipulation that improvements from Motorola were to remain inside Motorola. Essentially, the FSF simply sells copies of the code base. By retaining full copyright on programs, the FSF has the power to do this sort of thing whenever it wants.
Now, I don't believe that the FSF will ever abandon the GPL, but you need to be aware that your code may be sold in a non-GPL format to XYZ company. If this is fine with you, well, it's a nice way to support the FSF (monetarily-speaking). If you have problems with this, then don't assign your copyright.
This has been an information-only post.
-Erik
Re:FSF Assignment How's and Why's (Score:2)
--
Changing licenses (Score:2)
Having a single owner for anything (including copylefted code) has both advantages and disadvantages. The advantage is that the owner, being a single entity, can decide to adapt the license when appropriate. This allows flexibility, although possibly causing a code fork. But this advantage is also a disadvantage, depending on your point of view.
For example, consider the issue of selling proprietary services using free source code. That is, you take some useful open source thing, say, slashcode, modify it, and then set up a website to sell the results. Well, that may or may not be what Taco would like, but under GPL right now it is fair game. (Assuming slash is under GPL; if it is not now, just assume Taco assigns the copyright to them.)
OK, one thing you might do to prevent it, is come up with some modified licence, the GPL++, which also specifies that if you use the code in *any* public way (not just to compile public binaries), then you have to provide source. That may or may not be a good idea, well specified, etc. etc. -- but let's assume it is.
OK, well then some projects are going to have an awful hard time to moving to that license, because they have many copyright holders. Sure, it is at least theoretically possible to round up 100 people who have contributed, just not that easy. Even if you can get hold of all of them, you have to get them all to agree that the changed licence is a good idea; and knowing the wars over GPL vs LGPL and the Berkeley licence, and Artistic, etc, this is a problem.
For the FSF, if they did decide to do it, they could rather easily. And they could do it for all of their offerings in sync, thereby getting network effects.
Again, whether this is good or bad depends on
Re:Where would the disgruntled submitter come from (Score:2)
So, why does the FSF feel they need extra protection? Why isn't the same mechanism that everybody who uses FSF software needs to trust sufficient for the FSF itself? Anybody who uses software from the FSF has to make do with the "COPYING" notices that are included on-line; if you ask the FSF for a separate written license for GNU software that put the GPL in writing, they refuse (or at least have done so in the past).
I'm not attacking the FSF over this. They have contributed a lot. But I still see some kind of contradiction here: it seems to me either we can trust electronically affixed copyright notices and licenses or we can't. If there is some argument why the FSF needs written notices and copyright assignment while the rest of us are safe with on-line notices and the FSF holding copyright, I'd like to understand it.
Re:FSF Assignment How's and Why's (Score:2)
I'm sure they feel that way, too, but the reality is that the courts and everyone else not aligned with the FSF feels that intellectual property does indeed have owners, so thus must GPL'd software in order for it to be defended.
It's a little like going to a Community Party meeting and be hit on for a donation. Sure, they're against private property, but the reality is they exist in a system that respects private property, so they have to have some to further their goals.
Re:Gosh! (Score:2)
It is a good idea to read the FSF's copyright assignment form [gnu.org]. Among other things, there is the following paragraph:
So even after you have assigned copyright, you can ask for an licence to use the work (possibly selling your code as part of a proprietary application). And the FSF can't refuse this request, as that would break the contract, causing ownership of the code to revert back to you.
Misconception (Score:2)
All they are saying is, if you want to be part of GNU EMACS, distributed by the FSF, then you have to assign your patches to the FSF. If you want to fork, and do your own, you are more than welcome to.
Gosh! (Score:2)
Oh well, I guess since there isn't such a thing we'll have to endure miles and miles of flames and incorrect information (not to mention Portman's Complaint).
--
Have Exchange users? Want to run Linux? Can't afford OpenMail?
Re:If the specs aren't open, it's not really for s (Score:2)
And this would be different from BMW's standard practice in what way? There are people lining up to pay big bucks to get treated like that. Go figure.
Note: This is the primary reason I'll never own a BMW - heck, even their tools are proprietary - you can't even work on one of the things without a few kilobucks worth of special tools only available through BMW!
Re:Release of ownership NOT necessary. (Score:2)
Public domain does not ensure that somebody can't exploit everybody's hard work without also putting the added/modified code back into public domain.
Hmm (Score:2)
-JD
Re:Why assign? (Score:2)
Assuming that you are originally licensed something under version 2 of the GPL, if a later version is released, then you may license the software under that version instead. This newer license presumably allows you to relicense under version 3, or any later version.
This way, if a loophole is found, it will not apply to any future GPL'd software. On the other hand, if you do license your software to version 2 or any later version, then you trust the FSF to not set the terms to something you disagree with. This is, however, safe IMO.
Re:Gosh! (Score:2)
Obviously GNU has a website, and the prominent person alluded to is Richard Stallman, who basically runs GNU.
Re:Why assign? (Score:2)
I wasn't sure the earliest GPL revs allowed you to choose the later license and obselete the current. (And I have seen software explicitly under one and one only version of the GPL).
All we're left with is the question of trust. There are certainly benefits to be had from assignment, but there is always the danger, percieved or real, of the FSF doing something 'wacky'. Myself, I'd trust the FSF to hold my wallet, but others might find 'the spokesman' a little too, um, radical to be trustworthy. (clears throat).
What limited assignments could be made to mitigate or eliminate the risk? Could we do something along the line of 'I assign thee, the FSF, copyright to this version of said software, but this version only . All later versions, including those derived from this version of said software, will be copyright original author
Re:Equipped to fight vs. willing to fight (Score:2)
Enforcing the GPL isn't the FSF's problem in cases where copyright isn't theirs. In fact, they can't do a damn thing beyond 'We think Z company stole M authors code, which was GPL.' (Remember boys and girls, libel and slander suits can kill if you are a disinterested party.) Their take was probably 'Why is this person calling us?'
Has copyright been assigned to the FSF? Have the authors been contacted? Was the commercial vendor asked about dispensation from the author? No suit has been filed, and we don't have Bruce Perens screaming bloody murder, so I can say with certainty that none of the above happened, if indeed any theft of GPL material transpired.
Oh, yeah.. One author failing to enforce his licence does not make that license flawed. It could have been any of the source-included licenses, BSD, GPL, SCL, NCL, AL, etc..
If you are serious, I'd advise you to come forward with further detail.
If the specs aren't open, it's not really for sale (Score:2)
But without the code, you have to use the kernel version that the binary module was compiled against. So for future kernel upgrades, you are dependant on Lucent seeing fit to recompile their driver. It is as if we don't really own that modem, we are just leasing it while we ue this kernel. We don't have the same freedoms as if we had bought a piece of hardware that was *really* being sold.
Would you buy a car for which the company kept secrete the method of doing some key maintainance, so that they could shaft you at dealership in a couple of years ?
If nVidia was really selling hardware like a bunch of true-blue capitalists, then they would just release everything you could ever want to know about that card, thus making it more valuable to you. But instead, they have to hold something back. This limits what I can do with the card now and in the future.
Not that nVidia isn't far from lonely in their little effeminate neuroses. The basic pattern is this: a company does work and produces something of value (software, hardware, whatever) and then they want to sell it and get money. But then they get all breathy and twitchy because they don't trust themselves to be able to continue to do good work, so they try for some manipulative contorted way of selling it but somehow still owning it -- not give the source code to keep you on the leash, not tell you how the card works so they stop releasing drivers and force you to buy the newest one, etc.
That's the problem with capitalism these days -- nobody has any fucking rocks. Why do I want to buy a video card from a bunch of Marys who don't even trust themselves to be able to stay in the black without some strings attaching me to them so they can pull on me in the future ?
I could see it if it was some ultra-low margin type of business where you needed to eke out every cent to just stay in the black. But this is the computer industry, for God's sake. It's the biggest boom ever, and their screwing around with this stuff, hiding piddly little things that would cost them nothing to release so they try to scratch another few bucks out later on -- or worse, some of these companies have no clear idea how keeping the secrete helps the bottom line, they do it out of an instinctive lack of confidence.
I support capitalism. When some real capitalists have a nice card for sale, we might do business.
Regardless of nVidia's current support of Linux, they are still purposely circumcribing the utility of the card to keep me tied to them. Why should I do business with a company that sees fit to damage the goods as part of some plan to manipulate me and my wallet ? Like Intel secretely marking down the speed on some CPUs when they thought they needed more CPUs in the low end of the market, rather then just marking down the over supplied high end.
I don't want nVidia's "support" (well, that would be nice, I guess), I just want the information which makes me not dependant on that support. nVidia can't sell me an unfettered piece of equipment. Fuck 'em. So they licensed some technology from other companies that won't let them do it ? So what ?
I'm going to wait until someone has something for sale, and buy that.
Re:If the specs aren't open, it's not really for s (Score:2)
I realize that releasing the drivers is not trivial for nVidia. As you say, and I believe, "There is significant stuff in those drivers which nVidia does better than everyone else and which could be used in drivers for other cards. If they released it, they would lose a significant competitive advantage that they currently have."
But that stuff is not really available to me until they release the source. So they are in a have-your-cake-or-eat-it situation: do they want to sell their technology, or just keep it ? If they aren't selling it, why am I giving them money ?
They could release it under a license that would force a competitor to also release all specs to use it on a card. (It might be hard to enforce.) Then they could at least provide to *me* the ability to exercise their product to the fullest and continue to use it as long as it works. And nVidia would do just fine as long as their stuff worked better.
I don't expect nVidia to forfit a wad of cash to give me a gift. I just wish they would forfit a useful piece of computer equipment in exchange for *me* giving *them* a wad of cash.
If not, they can at least be more clear in the advertising. I didn't see "This video card's performance is limited by secret software you can't change or extend" written on the box. I didn't see "This video card might work with the next version of your OS if we still give a damn about you" written on the box. They want to pretend like they are selling something in the way that one sells a hammer or a toaster. But they are really selling something closer to the way that IBM and SGI used to sell mainframes -- tied up with a rat's nest of service dependancies, so it took a spreadsheet to figure out how much it really cost. nVidia wants to extend the Windows-world type of business model to linux: they are reserving the ability to improve certain things to themselves. The fact that their card works at all while many other cards don't is irrelevant to the principle of the matter. They see as integral to their business model a limitation on my abilities.
We should fight the expansion of that type of business. It is far too pervasive. If nVidia is so awesome, why can't they survive without keeping secretes about how to use their cards ?
Let's re-visit that quotation: "There is significant stuff in those drivers which nVidia does better than everyone else and which could be used in drivers for other cards. If they released it, they would lose a significant competitive advantage that they currently have."
Can I interpret this to mean that a large part of an nVidia card's performance comes from the software in the driver ? In that case, maybe nVidia's *hardware* isn't so awesome compared to other video cards, but the edge comes from *software* ? But wait, aren't they supposed to be a *hardware* company ?
If a major part nVidia's value that they are selling is in the software, if they are largely a software driver company, why should we treat them any differently from MicroSoft or SGI or anyother company that sells closed software, with all the limitations inherently associated with it ?
If nVidia did release their complete specs, then other companies probably would learn from them, and nVidia would actually have to make better hardware. Immagine that, a hardware company having to make good hardware to stay in business. Now that sounds like capitalism.
Until then, I'm going to stick to this ancient S3 with 2 MGs of video RAM.
So they can sell it. Re:Sign if... (Score:2)
Assigning the rights to the FSF allows them to sell these rights for closed sourced versions of the code.
Example: GCC sold to Motorola
Re:So they can sell it. Re:Sign if... (Score:2)
If it is because you thought the GPL ment no closed source forks, guess what...you were misinformed. Now you ARE properly informed.
Don't like the FSF selling GPLed code in closed source versions? Create the GTG that has the HQM licence (fsf and gpl and next letters) and say you are exactly like the FSF managed GPL, but you don't sell rights to make a closed source version. If you views are better, you will have people stop making GPLed code and making HQM licenced code instead.
(If you feel better the 1st time I posted this idea about a closed GPL fork, it got pushed to -1 flamebait, then moderated back up when others said 'yes this did happen'. Others have been pissed over it, because they saw the GPL as some kind of religious thing. If we want a REAL flame-fest with religious ovetones, lets get Brett Glass to junp in on this thread. Mr. Glass baits GPLers quite well.)
Re:If the specs aren't open, it's not really for s (Score:2)
Can I interpret this to mean that a large part of an nVidia card's performance comes from the software in the driver?
No. In fact, nVidia does have the best and fastest x86-based graphics hardware available today. The thing is, 3D graphics is extremely broad. There are an almost infinite number of features that could be implemented in hardware. The specification for OpenGL is several hundred pages thick, despite being almost too brief to understand. No hardware ever made implements every OpenGL call on-board. However, the hardware vendor must supply a complete OpenGL implementation. Otherwise, the application programmer would have to check every feature to make sure it is supported, and implement it themselves if not.
Of 3dfx, ATI, Matrox, and nVidia, nVidia is the ONLY company that has written a complete implementation of OpenGL for their cards (I don't know about S3, but I'd guess not). Their hardware accellerates more OpenGL calls than anyone else's, but it is still far from doing everything, and they have to support their older hardware, which does alot less. Thus, they have to write software to do the rest. Obviously, this software would be very useful to other graphics card companies wanting to write drivers for their cards.
On Linux, OpenGL is much better. 3dfx, Matrox, and ATI (I think) have based their Linux drivers off of Mesa, which is a more or less complete OpenGL implementation. However, I am programming a 3D game engine (GPL'd), and I see a distinct difference between running it on Mesa-based drivers and running it on nVidia's new drivers. The nVidia drivers look much better, especially with lighting. I would *like* to see them release their souce code so that everyone would get these superior drivers, but I don't expect them to do it because then customers of other companies would get the advantage that only nVidia customers have now (this is how capitolism works, BTW).
You may be wondering why I am so eager to support nVidia. Well, like I said, I am writing a 3d game engine. nVidia has traditionally gone to great lengths to support new features in their cards even when games don't actually use them yet. Developers love them for this, because nVidia is rapidly expanding the amount of cool stuff that can be done in games. But if people go out and buy, say, a 3dfx Voodoo 5, it is like taking a step back. If too many people end up using V5's, then I won't be able to make games that use these advanced features, because no one with a V5's will be able to use them, and they will all blame ME for their buying mistake.
So, the choice seems to be between nVidia's high image quality and high driver quality and everyone else's medium image quality, medium to mediocre driver quality, and open source code. (John Carmack himself said that he believes nVidia's drivers are better than anything he and UtahGLX have come up with.)
Go ahead, take the mediocre solution with the open drivers. But don't complain to ME when the software I write doesn't work on it.
------
Re:If the specs aren't open, it's not really for s (Score:2)
Apparently, you weren't listening.
Releasing the drivers is NOT trivial. There is significant stuff in those drivers which nVidia does better than everyone else and which could be used in drivers for other cards. If they released it, they would lose a significant competitive advantage that they currently have.
Again, I understand the advantages of open source drivers. I have written around 100,000 lines of GPL'd code, and I'm proud of it. I just hate the fact that everyone on Slashdot expects nVidia to basically forfiet a big stinking wad of cash so that they can give us a gift. And then you go and say "Fuck 'em" when they don't do it. And you call them greedy?
BTW, the nVidia kernel module is not tied to any specific kernel version. It comes with a source code layer which you compile for your particular machine.
------
Code reuse. (Score:2)
Re:The GPL is meaningless otherwise (Score:2)
Please talk to a lawyer before you make statements like that in the future, especially in public forms. IANAL, but IMO you don't know what you're talking about.
Let me put it more simply. You decide to spend a few hours mowing your lawn, but your lawnmower does not work. So you go to your neighbor and ask to borrow his. He says "fine, go ahead", you bring it home, and start mowing.
His wife comes home, yells at him for not mowing their yard, and, on the spur of the moment, he says "but our lawnmower was stolen".
She says "hey, our neighbor [you] was using a lawnmower just like ours when I drove by, I'm going to call the police because it's obvious he's the one who stole it -- his own stopped working, and obviously he didn't go out and buy one that's exactly like our 2-year-old model!"
So she calls the police while her husband tries to escape her wrath by staying very, very quiet.
As the police handcuff you and haul you off to jail, based on the clear evidence and the testimony of the wife (as well as the silence of her husband), ask yourself this one question:
The issue is basically the same: whether party A is considered to be criminally liable for using something that party B claims belongs to it depends to a substantial degree on whether party C, which party A could reasonably be considered to have thought owned the disputed property, made what appeared to be a legal transfer of that property to party A in the first place.
If you had gotten that signed agreement in the above scenario, the police would likely take it into evidence, maybe consider whether you'd forged it, but mostly focus on the dispute among (and possible false report filed by) your neighbors, leaving you alone to mow your yard. (If they take the lawnmower, at least you remain free.)
Similarly, if Microsoft claimed both Linux and, say, GCC contain substantial amounts of code belonging to it, and the FSF has a copyright assignment on file for that code while nobody has any such assignment on file for Linux, it's less likely the courts would order that distribution of GCC cease pending trial than Linux. Similarly, the threat of a finding against distributors of GCC would be lower than of Linux, especially the threat of punitive judgements.
In the GCC case, the legal issue would mostly revolve around whether the Microsoft employee had behaved properly in signing over that code; the FSF would be considered to have acted improperly only if something like knowingly accepting an improper assignment could be proved.
In the Linux case, the legal issue would start by asking whether Linux contains code Microsoft can show it had a copyright for. Once that's answered "yes", Linux's ability to be freely distributed is seriously damaged, and while the courts would certainly look carefully at how this situation came about, the burden would more squarely rest on vendors of Linux to show they didn't know the code belonged to Microsoft than it would on vendors of GCC.
IMO, from a legal as well as social standpoint, the FSF's approach to handling the code-copyright issue is on significantly surer footing than the approached used by the Linux kernel.
(That isn't surprising, since the FSF, unlike probably any of the significant contributors to the Linux code base, hired actual intellectual-property lawyers to recommend the best way to protect GNU against all sorts of potential attacks. That doesn't mean they indeed came up with the best answer, but it does mean that if those lawyers believed the Linux model was better, the FSF would likely have loved to choose that over the tedious form-filling method it currently uses for core components of GNU. The main remaining question, of course, is whether the FSF itself can be trusted to have good, up-front motives -- again, best to ask a lawyer to what extent a US non-profit organization that seeks contributions, even in the form of source code, can turn against its previously stated aims and goals.)
Re:Where would the disgruntled submitter come from (Score:2)
Well, think it through! Use deductive logic, for example. Consider the alternatives, and see how viable they are compared to the current model.
Public domain: there's no single organization that holds records of contributions to the public domain, even just for single projects, that I'm aware of. So if MacroEvil Corp decides to try to stop PING (PING Is Not GNU), the PD clone of GNU, from being distributed by filing lawsuits against distributors claiming ownership of major chunks of code, there's no easy way their PD status can be confirmed or even spoken for in court.
Collective copyright ownership (the Linux-kernel model): if MacroEvil acquires, say, 20% of the copyrighted code, it can successfully stop distribution for a substantial period of time. Even if it claims to own that much, it could stop distribution for quite a while.
Corporate ownership: if RedHat is bought up by MacroEvil, they can turn RPM (for example) into a proprietary app overnight, and can use legal tactics to hassle anyone who continues to use its GPL'ed predecessors if they're truly cynical about the state of copyright law (and who knows what laws might be passed in the future to make this scenario feasible).
Church ownership: if Scientology bought up 20% of the individual copyrights to Linux, how successful do you think they'd be in stopping its continued distribution should they decide Linux posed a significant threat to them?
Government ownership: if they don't want you distributing something, they can warn you. If you don't listen, they can send armed agents to your home, knock down your doors, and take all your computers. A massive government/media campaign to, say, encourage The People to favor "returning The People's code to The People", i.e. take it out of your hands, precedes this action, causing some 70% of the populace to not only support the use of force, but reject any attempts by Congress to investigate its propriety.
Non-profit organization ownership: The FSF somehow gets "taken over" by proprietary-software zealots and tries to sue people who continue to distribute GNU under the GPL. Here, not only do all the defendants have clear rights under copyright law and the GPL itself since the FSF made all those assignment-form-based agreements, the US government, as far as I've been able to determine, takes a good hard look at whether the FSF loses its NPO status (among other possibilities) due to turning its back on its own mission.
So, if you don't, as an individual, think you should "have to trust the FSF" in a way the FSF doesn't quite trust its contributors, extend the quoted phrase -- "have to trust the FSF more than...who or what?" You're receiving code from someone (you didn't write it, obviously) under some terms (let's assume you mean Open Source, aka free software, or maybe just GPL'ed software). Who do you trust more to let you continue using and redistributing the software than the FSF, or a similar non-profit organization? What keeps that person or organization from "changing its tune" down the road? Would more of those of us who actually contribute code trust the entity you appoint more than we already trust the FSF (or some other non-profit, since your question doesn't really pertain to the FSF as it does to the FSF's requirements for assignments)?
In particular, perhaps you should consider that the FSF's special efforts (and tedious ones at that) to secure clear, documented rights to distribute its core GPL components to you in the first place constitute a substantial reason to trust what it distributes to you in the second place?
(Socioculturamathematically speaking: trust generally erodes for each pair of hands in a transaction; therefore, the FSF must substantially raise the trust bar to ensure the level of trust it wants users of its GNU system to be able to place in it. Your claim that "either we can trust it or we can't" tilts at the windmill of black-and-white logic: in fact, you cannot trust electronic copyright assignment, period. You can, however, assign a trust value to it generally, and significantly narrow that interval in specific cases. I'm arguing that the FSF's distributions of GNU have among the most trustworthy copyright assignments you're likely to encounter in source code, GPL'ed, Open-Source'd, or otherwise, because of its mission, its nature as a non-profit, and the large population of people who entered into signed agreements with it to maintain that very trust.)
IANAL, but I have yet to learn of any option quite as good, given all the risks, tactics, etc. I've seen or been warned about to undermine the GPL, as signing all pertinent copyright over to a well-run NPO (and I do hope the FSF is sufficiently well-run in this regard).
Disclaimer: IANAL, and, in particular, I haven't enough expertise in non-US and International law to even think about evaluating the relative feasibility of using "offshore" methods of protecting GPL'ed code (and other free software) from legal and other forms of attack on continued free redistribution.
Re:The GPL is meaningless otherwise (Score:2)
The problem is not the injection of logic into a conversation about law. In fact, my statements are entirely grounded in logic; to the extent they are not, they are unsupportable.
A mistake so many people -- perhaps "computer types", especially -- make when it comes to discussing legal issues is that they assume there is some "objective truth" that everyone usefully agrees to, leaving the "law" (the courts, judges, legislation, etc.) to merely determine how that truth is to be implemented in human (and corporate) lives.
Whether there is an "objective truth" is not the issue per se -- it's the assumption that one can base the legal system on the belief that everyone recognizes and agrees to it.
So the mistake people make is not so much saying, e.g., "either MS owns the code or it does not", because that may well be determinable in an objective-truth sense.
No, the mistake is when they go on to say "and because that is determinable a priori, there is no need for parties involved in a dispute over said ownership to fully document the relevant transactions", for example.
The proper framing of these issues may well start with "either/or", but for most any useful conclusions to be reached, the very next steps must include "forking" that either/or into a variety of statements such as "Party A claims Fact M at Time T", "Party A claims Fact M' at Time T'", "Party B claims Fact N at Time S", and so on.
And each of these statements is, in turn, a claim, i.e. "Party B claims Party A claimed Fact M at Time T".
And since we're discussing whether Party B should take some action at Time P to head off potential problems at Time P+N, we must, logically, recognize the probability of a variety of claims made by various parties.
Only after we've accounted both for all the distinct claims and the potential future claims can we begin to reasonably assess legal strategy.
The FSF has done this, both well and wisely, not only IMO, but in view of the history it has had to date with various challenges to its dealings.
It's not unlike quantum mechanics vs. Newtonian physics. (Not that I'm an expert on either, but....)
I.e. in Newtonian physics you can say, definitively, that Particle P is at Location X,Y,Z moving at Velocity V in direction A,B, or some such thing. Given a set of particles with such precise information, it's easy to conclude the likelihood of collisions, etc.
In "reality", aka quantum mechanics, we can't say these things about any particle. Now, sure, it's tempting to assume that, at some level, there is such information encoded somewhere (outside our universe, i.e. outside our realm of perception), but we can't observe that level of detail for any particle.
But what we can do is make educated guesses about probabilities of the future behavior of particles based on past observation.
The processing required to assess the likelihood of collision among a given set of particles becomes not only harder, but intractable (impossible to determine yea or nay) in certain circumstances (at least, that's the impression I get).
So, yes, certainty is a wonderful thing, and to some extent all our logic is based on it, but to assume we can establish the certainty of any particle -- including a concept like "MS owns the code or it does not" -- before it ever becomes an issue is to make a potentially fatal mistake.
Re:Sign if you want code undefended (Score:2)
The situation that shows best what kind of threat legal action against GPL violations could be is what NeXT tried to do with the GCC objective C extentions [gnu.org]
Its not as definative as a win in court, but probably far less costly. Its still a feather in their cap and can be used in negotiations against future infringers. ("You really think you can win if we take this to court? NeXT didn't think so, so they folded".)
Re:So they can sell it. Re:Sign if... (Score:2)
Anomalous: inconsistent with or deviating from what is usual, normal, or expected
Sign if... (Score:2)
Anomalous: inconsistent with or deviating from what is usual, normal, or expected
Re:The GPL is meaningless otherwise (Score:2)
The GPL is not meaningless without the FSF. Any organization could serve as the clearinghouse for copyright assignment for an Open SOurce project. I'm still not convinced that one is necessary.
A Class Action suit by those who contributed code to the Kernel could be formed with Linus as the lead complaintant. The problem of perhaps not being able to identify the individuals is thereby avoided. Anyone claiming to be in the group could demonstrate it by providing copies of the patches they provided.
In the case that someone contributed code that they didn't have the authority to contribute, the FSF would be in just as much trouble as the Linux Kernel group. But, frankly, this is not a real problem. The code in question could be pulled out and replaced with functionally equivalent and non-infringing code fairly quickly. Furthermore, the complaint would have to disclose exactly what code was infringing.
Anomalous: inconsistent with or deviating from what is usual, normal, or expected
When to fight (Score:2)
If FSF fought every violation, then they would not longer be well equipt to fight other battles.
In my case, I had spoken to the ACLU, not to have them fight for me, but just to make them aware (Mattel using litigation and threats of litigation to intimidate people into surrendering their rights). I provided the ACLU with information to make more clear the misrepresentations that Mattel / MSI made in their briefs.Mp>
Lame but perhaps common query? (Score:2)
I took a GPL (not LGPL) CGI-C library to get some upload-file-to-webserver stuff; modified it extensively to do what I want it to do, and use it on a webserver which is about to go live on the internet.
So I've forked some GPL; no problem. GPL allows that AFAIK. But am I obliged to distribute my source? I am running (on MY machine) but not distributing the binaries
Since FSF don't seem to be pursuing Mattel, I can't see that I'm in any legal danger for a site restricted to 12 users (may grow as the company grows), but from a legal point of view, what are my responsibilities?
Slashdot Poll Suggestion (Score:2)
Why should I assign my copyrights to the FSF?
Disclaimer: the above is meant as a parody on the typical Slashdot Response. Don't read if you do not have a sense of humour.
What about authors' credits? (Score:2)
Re:xemacs protection (Score:3)
Nice theory; too bad it has nothing to do with reality.
All of the work that went into Lucid Emacs had copyright assigned to the FSF. Let me say that again, since people seem to selectively forget it: every version of Lucid Emacs, from 19.0 through 19.10, had FSF copyright notices, and had all the appropriate paperwork signed and delivered to RMS. That did not prevent the fork. The fork happened for reasons that will hopefully be made clear to anyone who chooses to read the archive of the debate at the time [jwz.org] rather than relying on rumor and half-remembered fragments.
It may be that the fact that today's XEmacs has many different copyrights in it makes a merge with FSFmacs be less likely, by virtue of the fact that RMS absolutely demands copyright assignment. However, the copyrights had nothing to do with the Lemacs/FSFmacs split, and the reasons for that split still exist as well: so the copyright issue is surely lost in the noise. Even if the XEmacs folks assigned RMS their copyrights tomorrow, a merger would still never happen, for all the same reasons as were true in 1992.
Allowing FSF to Take Infringers To Court (Score:3)
Note that contrary to fairly common paranoid fantasies to the contrary, the author can always retain a non-exclusive copyright, as per the assignment agreement: [gnu.org]
Re:Gosh! (Score:3)
This is what you *will* give up: if someone infringes your license, then you won't have the right to sue them. The FSF will have that right, and it's possible that they will not pursue a case while you might have done so. On the other hand, pursuing a case takes money and good legal counsel, which the FSF has or can probably raise quickly if GPL case was going to court.
The other thing you will give up is your right to distribute your source code under a different license. You could distribute a program called "Duke of URL's Hello World" under the GPL, but then one day decide to sell someone else a copy of "Duke of URL's Hello World" under a restrictive shrink wrap license without source code. If you were the copyright holder, you could do that. If you assign the FSF all copyrights, then you're just any other schmoe out there, with GPL rights and that's it.
I am not a lawyer.
Copyright assignment considered harmful (Score:3)
Transferring ("assigning") copyright is DANGEROUS and I'm not sure the GPL or any other open-source software should be attempting to do this.
The problem is that the legal requirements to actually effect a transfer are complicated and difficult: the law wants to make sure that when you give up your copyrights, you really really meant to do that. I am not a lawyer, but my understanding is it requires an actual signature on an actual piece of paper--some unsigned statement posted to the internet or sent via email just won't do.
So all these copyright "assignments" that have supposedly given the FSF rights to a variety of software are probably unenforceable. This may be no big deal, but if people wander around relying on the idea that they've effected a copyright transfer when in fact they HAVE NOT, there could be bad consequences.
It is much easier to make a NON-EXCLUSIVE (and likely irrevocable) copyright grant, permitting the original author to anything at all they like with the software. That doesn't require a signature on paper--any statement in any written media (including email, web pages, etc.) will likely suffice. The contributor retains copyright ownership, but the original author gets to use the code in any way at all.
Both the contributor and the original author would then be able to go to bat in court for the software, as both would be copyright holders.
Of course, I'm not a lawyer... but people shouldn't get the idea that they understand all of the issues here. Copyright ASSIGNMENT (transferring ownership) is a difficult and thorny thing.
assignment raises troubling questions (Score:3)
I understand the motivation for the FSF to cross all their t's and dot all their i's. That was probably crucially important when they were still the only open source organization around. But today, lots of companies and universities release lots of open source software. Most don't bother assigning anything to nonprofits. And I don't know of any legal case where an assignment would have helped.
The FSF asks users of free software to trust the GPL. The impression that the requirement for assignment creates is that they themselves feel that the GPL has some risks. If that is the FSF's position, I would like to see this clarified more and understand what it means for users of GPL'ed and FSF software.
Some questions (Score:3)
First of all, I am certainly incredibly naïve in my not-a-lawyerness, but why is it that the copyright must be granted exclusively to the FSF? Why can't I write something like "Copyright (C) 2000 by [me], the Free Software Foundation [and perhaps also a few other people I might trust in the matter]"? Doesn't this give them the right to (1)sue anyone found guilty of a license infringement, and (2)relicense the code if needed?
In other words, if I understand correctly, the FSF is interested in rights (1) and (2) above. Why must the original author relinquish these rights so the FSF can have them? Why cannot they both have them? I see no reason why two groups of people's rights to the same code should come in conflict; in fact, there are many such cases.
As a matter of fact, this question of copyright transfer is dubious: it works in those countries that have the "copyright" approach to copy protection, but it won't work in those countries which have "intellectual property" instead. Although both are unified under the Berne convention, the important difference is that intellectual property is not transferable.
The whole question is pretty murky: can't someone make the effort of translating it all (copyright laws and various licenses) in formal logic so we can be sure, once and for all, what it all means?
Not exactly the same as NPL (Score:3)
Re:Release of ownership NOT necessary. (Score:3)
That's like saying that Jesus must be hypocritical because he couldn't have come down and saved everyone from their sins if nobody sinned. It is ironic, maybe, but not hypocritical. The FSF would, presumably, do everything in their power to hasten the day when they become completely redundant. (No arguments about the validity of Christianity, please, it was an illustration).
A Written Instrument cannot be Extinguished (Score:3)
This law says that if a company tries to coopt GPL software by buying exclusive rights from the authors, the GPL can be destroyed unless there was a WRITTEN INSTRUMENT SIGNED BY THE OWNER of the rights to the software transferring non-exclusive ownership to someone like the FSF.
If it weren't for this law, and a programmer's failure to follow the suggestion of the FSF, the hack of Cyber Patrol could have been an opportunity to test the perpetual freeness open source software.
The Cyber Patrol Case:
Some people reverse engineered Cyber Patrol and found out the access code for disabling the program. Reverse engineering is a violation of US copyright law, so Mattel (owner of C.P.) sued.
The two people who worked together on the hack, Matt and Eddy, settled the lawsuit with Mattel, and granted to Mattel all rights to the software and the accompanying essay.
After that, someone noticed (and told slashdot) that the source code of the hack contained the following line in a header file: "CPHack v0.1.0 by Eddy L O Jansson / Released under the GPL". After the GPL line was discovered, people wondered: Maybe the programmers couldn't sign over all of the rights to Mattel, because they had granted non-exclusive licenses to everyone who already downloaded the source code. That is, if it is released under the GPL, nobody should be able to take it out of the public domain. This proposition is something the Open Source movement has been waiting to test in the courts once and for all.
Unfortunately, there are a few things that made this a very bad test case for the GPL. In fact, there is no chance that this will be a test case. 17 U.S.C. 205e, of the United States copyright law states that the transfer of exclusive rights to someone includes the elimination of all nonexclusive rights (such as GPL rights) that may have been given to other people, UNLESS there is "a written instrument signed by the owner of the rights licensed" proving there was a transfer of nonexclusive rights. Matt and Eddy never signed over non-exclusive copyrights to the FSF in a "written instrument," so the opensource-ness of the code could be destroyed by Mattel. That is, the programmers' transfer of all rights to Mattel included the elimination of the nonexclusive rights that the public had already been granted by the line in the source code. As Professor Moglen stated in a Wired article: "This is one of the reasons why the Free Software Foundation strongly urges authors of free software to assign their rights to FSF. It does them no harm and it provides us with precisely the signed instrument."
The other reason this would have been a bad test case was Eddy's own explanation of his intention that he posted to slashdot:
Here is Matt's comment from his web page [islandnet.com]:Thus it seems the GPL assigment was accidental and incomplete. If the GPL assignment had been intended, though, there might have been an interesting loophole in the copyright law that allows for a solid release into the public domain, even without the signed instrument. According to a quote by Moglen in that Wired article, "New works made pursuant to the license at the time before Mattel [acquired rights to cphack] present Mattel with other difficulties." That is, if you altered and released a modified version under the GPL before you knew that Mattel had the rights to the program, you might have subverted Mattel's attempt to fully suppress the program, because you would have been a co-owner of rights to the new program, and you would have held onto what Matt and Eddy tried to extinguish.
The moral of the story: always obtain a written signed instrument that assigns non-exclusive rights to somebody other than the author, so the software remains open source forever.
Re:The GPL is meaningless otherwise (Score:3)
I am getting really sick of everyone attacking nVidia for supporting Linux.
Figure 1: Level of OpenGL support provided by various Windows hardware vendors
As you can see, nVidia has put a LOT more into their drivers than anyone else. I am working on a 3D game engine, and I have ported it to Windows, but it ONLY works on nVidia cards simply becaule no other company supports OpenGL lighting (yes, if it is not done in hardware, you are still supposed to have a software implementation. nVidia has both). If they released that open-source, they would lose an edge over the competition. YES, a significant portion of their drivers could be used in drivers for competing cards.
Figure 2: Register level specs for nVidia RIVA chipsets (128, TNT, TNT2)
If you don't believe me, go look. They also have un-obfusicated source code implementations there, and "driver development kits" for various OS's. Unfortunately, because they licensed some technology from other companies, they were not allowed to put up complete specs. The missing peices are rather important, so nVidia decided to write drivers for us so we could use them.
Figure 3: nVidia's options
I would like to see the drivers released open source as well, and I am not convinced that it won't happen. However, I talked to one of the engineers, and I can tell you one thing for sure: The zealot attitude of most Slashdotters is only making the situation worse. Anger will get us nowhere.
Now I am going to go install my GeForce 2, which came today. Joy.
------
Re:The GPL is meaningless otherwise (Score:4)
>mean, IANAL, but my understanding is that the
>copyright owner has the authority to license
>their work for redistribution. However, Linux
>doesn't appear to have a copyright owner. As a
>result, if the GPL collapsed legally and needed
>to be rewritten, then Linux is dead. Why? Because
>NOBODY has the authority to redistribute the
>kernel.
The GPL covers this eventuality, because it's version 2 "or future versions". The FSF could come out with a new version of the GPL that addressed whatever problem the license might have, and there you go.
Strangely enough, if the FSF got taken over it could come out with a new GPL that allows proprietary use, and kill it anyway, so really what are people losing by signing over the copyright? It's not like they're giving them more power than they already have: they can't.
That said, from an administrative standpoint requiring signing over the contract cripples GPL programs because the spur of the moment aspect of a lot of development is outright KILLED by it. If you have to fill out the equivalent of tax forms to get your contribution accepted, screw it.
This is why GNU languished for years and Linux took off immediately. Idealism gives us our frame of reference, but pragmatism is what works.
Rob
Why assign? (Score:4)
The second, which applies especially well to huge projects with hundreds of authors, is that the lead maintainers/authors don't need to do the whole 'who wrote what' game everytime there is a question of copyright and/or license exception. If I, as the PM of something as huge as Xfree or emacs, would not want to have to deal with even one disgruntled submitter of code six months after the fact. One of the lead assumptions in the second is that the FSF is merely a 'puppet' for the authors, and would abide by their decisions in the matters of copyright.
Well, I lied. I can think of a third. Orphan programs. Stuff abandoned by the copyright holder but still maintained by others, that needs an updated license (eg, a hole has been found in the 1.0 GPL it was released under). If I can't find all of the authors, I can't say 'This, and all later revs, are GPL v3.2'. I can't make an exception for the greater good and grant a conditionless license. (remember Corel?)
FSF Assignment How's and Why's (Score:5)
The rationale seems to be two fold:
1. Following these rigid procedures ensures that there is "clear title" to the code, and that it is properly licensed under the GPL. Who knows what hidden licensing bombs might lurk in the Linux kernel code.
2. If the FSF owns the code, they will have standing in court to pursue violators of the GPL. If the code infringed were owned by someone else, the FSF might not be able to go after the bad guys.
I've always been very down on copyright assignment because I think it is antithetical to the GNU believe that "software should not have owners". It seems highly ironic to me that the FSF is very demanding that they do in fact own the code.
OTOH, recently I've come across two practical cases where assignment has helped:
1. The case of someone trying to rescind the GPL as we have all read about here. The retroactive element is an interesting one, but not really that important. With FSF copyright we know that all future releases will be free software.
2. It simplifies relicensing. One of my projects recently merged with another one and the FSF agreed to switch licensing terms from the LGPL to the libstdc++ license (GPL + exceptions). This would have been very difficult to do with individual copyrights held by contributors.
Why I'm considering assigning copyright to the FSF (Score:5)
I was told that if I chose, I could assign the copyright for the entire program to GNU, or I could retain the copyright myself. Not knowing what I was doing at the time, I chose to keep the copyright since I felt that would be safer.
As it turns out, I think the main difference between assigning the copyright to GNU and not doing it is if there is a dispute, or if company X is using your GPL'd source code in some way that is contrary to what the GPL says, if the code's copyright is held by the FSF, they'll stand up for it and litigate, whereas if the individual keeps the copyright, GNU may not have a stake enough to through their weight against the company.
Frankly, if there's anybody on earth that I TRUST to hold the copyright and make sure that the GPL is enforced, it's Stallman and the FSF.
As for signing over patches, I think it's possible that the FSF has gotten burned in the past on submissions being written by people at their jobs, incorporating the patches, and then finding out later that the employee didn't have the write to submit that code under the GPL since it was written on the company's time and hence property of the company.
Signing over patches and explicitly authorizing them for the FSF is a procedural thing to make sure that the free software we use remains pure of other types of non-free code, and presents as small of a profile as possible to litigious lawyers who might want to take someone to court over possible misappropriations of source code.
I think it makes sense to even require this of people. Why else would you be writing software for the FSF if it wasn't going to be GPL'd? If that's the case, why wouldn't you want someone larger than yourself to be able to back and protect the code? The only problem with this I could see would be if you didn't trust the FSF. If that's the case, I'd wonder what your reasons for that are, but I won't disagree with your personal feelings about the FSF.
Release of ownership NOT necessary. (Score:5)
There are some good arguments for FSF owning copyright to works they destribute. (Although I too am struck by the irony and potential hypocracy-- thats aother point, see below.)
BUT you do not need to give up your ownership of Copyrights to give Copyrights to another. There is a concept under Copyrigtht law calls "joint copyright." This leaves your rights intact escept for the fact that it gives the joint tennent equal rights to modify,license, destribute, etc. I can't see a good argument why this would not qork for FSFs stated aims. Can anyone suggest one?
Also, its worth nothing that book publishers can and DO proetct their books, even though they don't own the copyrights outright. The contracts they have with the authors take care of granting them the right to pursue infringers. Although I dopn't know the details of those arrangments, they can and do exist.
On that second subject I promised to touch on. It is ironic, and perhapse hypocritical to some degree, that the FSF could not exist as it is without Copyrights. If FSF REALLY believed in 100% free, no restrictions software, they would not license but rather put the software into the public domain.
Instead, Stahlman developed the GPL as a way of attempting to bribe/cooerce others into the same release rules fro their software as hee wanted for his own.
Is requiring subscription to a poltiical agenda really any less intrusive then requiring cash payment?
Something to think about.
The GPL is meaningless otherwise (Score:5)
However, should a third party begin using the Linux kernel illegally, it would be interesting to see how it could be enforced. First of all, the author of the specific code would need to file the suit. Now, this might be a random programmer somewhere, without the legal staff of the FSF.
Also, it is questionable if the GPL would work. I mean, IANAL, but my understanding is that the copyright owner has the authority to license their work for redistribution. However, Linux doesn't appear to have a copyright owner. As a result, if the GPL collapsed legally and needed to be rewritten, then Linux is dead. Why? Because NOBODY has the authority to redistribute the kernel.
Without a single owner, it couldn't be placed under a new license that could be upheld. Furthermore, a single GPL license for Linux may not be reasonable. Why? Each section is owned by a different individual. It would appear, to me, that they would all have to have their copy of the GPL included, and you should, theoretically, need to agree to dozens of separate GPLs with different parties to redistribute the GPL.
This is a potential nightmare. Again, it all comes down to, do we do things the "Right Way" FSF style whoses body of lawyers and Ph.Ds have done their best to work within the legal regime. Of the "hacker way" of daying "damn the man" even if it means that the GPL'd code base is worthless because we haven't done it correctly.
For example, say there is code in the Linux kernel written by someone under the employ of Microsoft. They didn't aquire a release from Microsoft to license their code under the GPL. Microsoft wants to shut down Linux, and files a lawsuit against Redhat, Caldera, S.U.S.E., the owners of various FTP sites with Linux, etc. Why can they do this? Because they own some code within the kernel that is being illegal distributed.
The Linux way leaves us open to blackmail or destruction by ANY malicious company whose employee didn't double check his contract and wrote kernel code. The FSF way guarantees that the code will always be free, although if the FSF went renegade, we would need to fork it under the GPL. If the FSF has the authority to revoke the license (we don't know what the courts would say) and they chose to do so, there is nothing that we can do, as FSF code is everywhere, so we'd be screwed. Worrying about the FSF turning on free software is absurd, because if they did so, we're dead regardless of if we turn the code over or not.
Now, I can think of two recent slashdot articles demonstrating the pros and cons.
BeOS, a well liked company, took code accidentally into a closed module. The code wasn't really used, but it was there. Bruce Perens, as the copyright owner, was able to issue them a writ allowing it. If the FSF owned it, they would have probably demanded that BeOS open that module... what would be better? I don't know. The former built up good will, the latter would result in more open code.
The other example is the nVidia mess. Much like the example of a company building a derived work from GNU Readline (I think I have the right name), the FSF might have been able to demand that they open the module or face a lawsuit. Instead, the individual developer accepted the answer of: "we'll fix it in a few weeks by removing your code" instead of demanding that they open the module.
I would suggest that in nVidia's case, the latter would have been better. The reason is that I am fundamentally opposed to groups writing closed code to interface with the Linux kernel. They are using a technicality in Linus's interpretation of the GPL to avoid contributing back while using the GPL code to sell hardware.
Why is this different? Be is writing a closed source OS that tries to be friendly to the free software community that made a mistake, and is moving the code out of all CDs. nVidia is trying to profit off the free software community by using Linux to sell hardware without giving back.
I WISH that the FSF owned the code in question.
Alex
Why I assign to the FSF. (Score:5)
That said, I support doing copyright assignments to the FSF because it allows me to do the job I want to do. When I program for GNU, I don't want to deal with lawsuits, liability insurrance, and license agreements. Canada has a screwed up enough legal system without me trying to understand what happens in the USA (Much less the rest of the world). When the FSF takes over ownership of my program, I no longer need to worry about this. I also like the fact that the FSF can choose to change/modify the licensing as appropriate.
I'm also not without protection. Specifically two things: The GPL ensures that if the FSF does get corrupted, I am free to fork and continue (taking on the legal burden myself, or with whatever organization with which I choose to associate). I am also protected by specific provisions in the assignment form:
(d) FSF agrees to grant back to Developer, and does hereby grant, non-exclusive, royalty-free and non-cancellable rights to use the Works (i.e., Developer's changes and/or enhancements, not the Program that they enhance), as Developer sees fit; this grant back does not limit FSF's rights and public rights acquired through this agreement.
4. FSF agrees that all distribution of the Works, or of any work "based on the Works", or the Program as enhanced by the Works, that takes place under the control of FSF or its agents or successors, shall be on terms that explicitly and perpetually permit anyone possessing a copy of the work to which the terms apply, and possessing accurate notice of these terms, to redistribute copies of the work to anyone on the same terms. These terms shall not restrict which members of the public copies may be distributed to. These terms shall not require a member of the public to pay any royalty to FSF or to anyone else for any permitted use of the work they apply to, or to communicate with FSF or its agents or assignees in any way either when redistribution is performed or on any other occasion.
5. FSF agrees that any program "based on the Works" offered to the public by FSF or its agents or assignees shall be offered in the form of machine-readable source code, in addition to any other forms of FSF's choosing. However, FSF is free to choose at its convenience the media of distribution for machine-readable source code and may charge a fee of its choosing for copies.
The full text if you're interested can be found at the GCC site's [gnu.org] contributing section. If there are cases where there are "extraordinary circumstances", the FSF is sometimes willing to do up special assignment forms to help. RMS normally just deals with those, however.
If you're interested in seeing the information given to maintainers of GNU packages, take a look at Information for Maintainers of GNU software [gnu.org] on the GNU site, specifically the section on Copyrights [gnu.org].
(disclaimer: I program for GNU, I don't represent, work for, etc - My opinion only...)