OSNews

We've been through this over nine thousand times. You might not agree with the FSF's definition of freedom, but it's internally consistent even if _you_ don't accept it.

Also, in before anti-GPL shitstorm.

edit: typo

Edited 2007-06-01 15:42

Well said. Sadly, these days too many are taking advantage of the freedoms that the FSF has worked long and hard for, and they don't give one iota about the FSF. Ungrateful bastards.

We're seeing a strong push from the BSD users, and from corporations to try and invalidate the GPL, or circumvent it, in order to destroy the ideals of the FSF. This is very sad in my eyes, the FSF is one of the few bodies that pushes hard for [software] freedoms.

The software patent system is both horrendously broken, and also unneeded. There is another statistical evidence to show that software patents destroy competition, promote monopolies and also destroy the little guy from being able to invent (and protect his software invention). If software patents are so valuable, and so necessary, how has the EU survived for so long without them?

Can you imagine if the greats like Sir Isaac Newton, Albert Einstein, Johannes Kepler, Nicolas Copernicus, Ptolemy and so on and on patented their ideas? We'd have gotten nowhere, and society from a scientific point of view would have stalled.

The FSF and GPL promotes sharing software, for everyone, for the betterment of everyone. If you want to modify code and distribute it for a profit, that's fine (disproving all of thsse who call the GPL/FSF communist), just simply redistribute your improvements back to the community that helped spawn you software.

As archiesteel has said, if you don't like the GPL, then don't friggen use it. Simple. Piss off, use your BSD/proprietary code, whatever, don't reserve the right to bash the GPL/FSF because you disagree with its software principles.

I am very surprised at how many people are anti GPL/FSF these days, but I guess it has to do with the increasing numbers of Windows refugees, who come to GNU/Linux because it's free (as in cost) and fail to realise the other fundamental freedoms that the GPL/FSF provides you with. Talk about being stupid and ungrateful.

Dave

Although I saw two or three other things in your post, here's the biggest thing I have to correct in your comment:

The FSF and GPL promotes sharing software, for everyone, for the betterment of everyone.

I propose to prove that what the GPL promotes isn't sharing, and that there is no-one that it benefits.

First, as to sharing.

If I create Program X with Code Y, and Code Y is GPL v2 code, I'm allowed to use it for myself. However, I cannot distribute Program X (even for free*) because I would be forced to give away the code as well (i.e., Code Z, something that's part of X, but which is neither X nor Y). And that's not sharing.

Sharing is based on giving unto others what you want of your own free will, as people helping other people. Giving up what you are forced to give up by some external force, with the people being subordinate to the Glorious Movement -- that's fascism.

Second, as to who it benefits...

How do I, the would-be user, benefit by using code licensed under the GPL? How do I benefit by not being allowed to create a work on my own terms? Why must I (and thousands, maybe tens of thousands of others) re-invent the wheel if I want to be able to control my project? Why must I be forced by a copyleft clause to have every would-be-king take my work, tweak to make it slightly incompatible, so they can work against with my own work? Why can't I have the benefit of the doubt, and why must I be presupposed by a license to have evil intentions? Why is having a closed central point of control so wrong? Why is making money from my work, if I can get it, such a bad thing?

How do you, the would-be distributor, benefit from authoritarian "free" software licenses like the GPL? Will requiring others to give up their work help your project spread? Will mandating the distribution of a large, obfuscated license with anything created with your work make it any more likely that it will be obeyed? Will it make it easier to get private interest (and thus extensive bug-testing, adoption, and most importantly, a user community)? Will it make it easier to get corporate interest (and thus funding for full-time researchers, debuggers, coders, and the broad marketing and adoption that will come with their using your code)? Is artificial moral high ground you've built up really worth the loss of support, quality, and adoption of your work?

If you're going to give it away, at least have the curtosy of releasing it into the public domain (or as close as possible in current international copyright law -- a few countries won't let you do that, I'm told). If you're not going to do that, then don't call it 'free', and don't cover yourself in the drapery of 'freedom.' Call it open source if you like -- because the source is open, true, in the same sence that Microsoft's shared source initiative makes the code open -- but it's not by any means acually free or promoting acual freedom.



* I mean free as in actually free, not free as in any of the 'free as in...' metaphors for things that aren't freedom that you hear being batted around so much lately.

Edited 2007-06-03 07:11

I don't think you understand what sharing means: "to allow someone to use or enjoy something that one possesses."

The GPL is, in fact, sharing with others. You are not forced to share your work, because a) you are not forced into using GPLed software, and b) you don't have to redistribute it even if you do.

So, in fact, if you do not want to share freely with others, then you simply do not use GPLed code to build your application. It's as simple as that.



That is complete and utter BS. You are not forced to do anything. I'll repeat it, since your quasi-religious opposition to Free Software seems to cloud your judgement: you do *not* have to use GPLed software to build your own, and if you do you do *not* have to redistribute it.

It's very simple: if someone chooses - of their own free will - to release their software under the GPL, then you are only allowed to redistribute it if you abide by the conditions set by the copyright holder. If you're not ready to do that, then don't use the license.

The most hypocritical part of it, of course, is that you decry the GPL for not allowing you to use someone else's code to produce software which will give *no* rights to those who will use it. In other words, you criticize the GPL for not giving enough freedom to others, even though the software you say you would like to distribute would not give *any* freedom to others.



Ah, now I see the root of your misunderstanding. Licensing software under the GPL *isn't* giving it away at all. Rather, it is *licensing* the software to third parties, and giving them *more* rights (or freedom, if you will) than they would normally have under copyright law. However, the creator of the work still retains copyright over his code (even if you use GPLed code, you still retain copyright over *your* code, incidentally). You seem to imply that coders shouldn't have the right to choose the GPL for their software, which basically means that you do not respect their IP rights.

We already know that you personally don't want to freely give away your code, or release it under a license that gives users the freedom to study, modify and redistribute the code (otherwise you wouldn't whine so much about being bound by the terms of the GPL, despite the fact that you do *not* have to use GPLed code in the first place). We get it. Now can you stop polluting every GPL-related thread with your fanatical drivel? Thanks.

If you want to give away the code of program X, then you have no problem, because you can just release it under the GPL.

If you don't want to give away the code of program X (actually the part of program X you have labeled "Z") then don't give it away, keep it for yourself.

If you want to be able to sell program X as a binary only, then you have two choices: (1) write all of program X yourself, and don't sponge the Y part off of others, or (2) get in contact with the authors of Y and buy a commercial license for that code from them, and then you can use Y in your commercial product program X.

The only thing you may not do is use (gratis) other people's work (Y) for sale in your for-profit program X.

Please note: you may not use any of Microsoft's "Y" either, so this is no different in the case of GPL'd "Y".



You are not forced to do that. If you want to make a commercial X, and it must use Y, then either write your own equivalent of "Y" as part of X, or get a license from the authors of Y to use Y commercially as part of your X. You are not licensed under the GPL to do the latter, so this commercial license must be a separate arrangement between you and the authors of Y.

Edited 2007-06-03 11:49

What you're describing is a BSD style freedom, something that smacks of false freedom to me. Where is it morally right that person A can take my code, improve it, fork it, keep the src code private and resell it to make a profit without returning one iota of a thank you to me, the original coder who may have developed the VAST majority of his code? mmm? Answer that one for me please.

I don't care about business, I care abou the people. The BSD license does NOT care about the people, and that's why it's very popular with business. They know they can take, not give back to the community, and get away with it.

Tell me, if you designed a computer system, and sold it, to make a profit. I came along, copied 95% of it, made a few tiny improvements, kept those improvements from you (and the rest of the competition) and it became a best seller and outsold your computer and you went bust, would you be happy? You might call that capitalism, and it is (at its worst I might add - I'm not a fan of capitalism, it promotes human greed, one of our worst traits as a species imho), you might call it fair business. You might even argue that my example is a bad one, because patents would protect you. I'm ignoring patents for in my example for a deliberate reason, to make a similar enough analogy to the FSF vs BSD licenses and software appropriation.

Dave

I am very surprised at how many people are anti GPL/FSF these days, but I guess it has to do with the increasing numbers of Windows refugees, who come to GNU/Linux because it's free (as in cost) and fail to realise the other fundamental freedoms that the GPL/FSF provides you with. Talk about being stupid and ungrateful.

I lose patience over it. However, as friend much smarter than me reminds me, so many problems come from ignorance, which can be remedied with education.

I find it helpful to think of two ideas of the free software movement. First, software should be free. Second, free software must include its philosophy. So much collective vested interest is threatened by one or both of these ideas and so unwilling are these opponents to attack the ideas directly, that instead we see all manner of fallacious, elaborate, deceptive, ridiculous hate directed against the FSF, the GPL, RMS, etc.

Maybe the most important group of opponents comes from those with the most to lose in the face of their own stubborness: developers, lawyers, journalists, bloggers, salespeople, academics and many others active in the Microsoft development ecosystem. In this most sophisticated group will be sincere people but also all manner of shills, turfers, trolls, you name it.

A related group belongs to, as you say, the Windows refugees (refugees that are fortunately on the increase). It's good to remember that many of them may have spent their formative late teens and twenties steeped in a culture of no knowledge or experience of the four freedoms, and especially that these freedoms matter. This group will be the noisiest by sheer numbers.

Some other groups are likely more established but dropping off. They may even oppose Microsoft, who they may view as more an embarrassment than a criminal. Nonetheless, they very much have a stake in non-free software. Mini-Microsoft-wannabes, in a sense?

Another group might be the rebels who rebel for rebellion's sake. They are "independent thinkers" who are suspicious of any advice, even if they are unable to find any holes in it.

Another, rather pathetic, group are sentimentalists, who are just stuck in the past. It's easy for any of us to fall prey to sentimentalism, of course.

A more interesting group which often intersects with the other groups are those who confuse the open source campaign and the free software movement. Clearly, the two camps are on the same side, but there are two goals that are mixed up: the open source goal of sheer numbers and the free software goal of people working only with free software. The free software goal simply takes more committment than the open source goal, so it's no wonder that the other goal is more popular, and popularity breeds popularity.

In short, I find it helpful to realize that "inertia" explains much of the hate. Moreover, it is encouraging that the vast bulk of the planet will be far more predisposed to the ideas of free software.

Edited 2007-06-03 15:54

I gather it is freedom from one point of view or within a specific context rather than a notion that the term freedom is a catch all. This difference in the perception of freedom is generally what sparks the Linux/BSD flame-fests around here.

As an example, I'm free to pay lots of money for Vista Ultimate.

Being free to pay for Vista is 'choice', not fundamental freedom.

edit. I should add that I'm not disagreeing with your first point.

Edited 2007-06-01 15:54

Being free to pay for Vista is 'choice', not fundamental freedom.

You have a very strange definition of freedom.

Freedom is Choice residing with the Individual. They are not bound to purchase the decisions other Individuals choose to make.

I'm still in the mind that the FSF trying to step above the patent system, is just as bad as other companies using it. Some patents exist for good reasons, and good intentions.

If the FSF has issues with patents, it should be lobbying for change. At the end of the day, the written law is above the GPL3 "contract", and the terms of the GPL3 can be thrown out as not within the bounds of law, the same as any EULA.

edit: two typos

Edited 2007-06-01 15:58

I'm still in the mind that the FSF trying to step above the patent system, is just as bad as other companies using it. Some patents exist for good reasons, and good intentions.

If the FSF has issues with patents, it should be lobbying for change. At the end of the day, the written law is above the GPL3 "contact", and the terms of the GPL3 can been thrown out as not within the bounds of law, the same as any EULA.

Excellent comments. Too often, people solely think of patent holders as big corporations. But there are plenty of small-time inventors who deserve the protections of the patent system against the big corporations. GPL3 doesn't seem to acknowledge this need at all.

Those are a very small minority. The vast majority of software patents are held either by big ISVs, or worse yet, by companies that do nothing but amass patents for litigation purposes.



This is not new in GPL3, it was there in GPL2. It is perfectly consistent with the FSF's well-articulated opposition to software patents.

BTW, the FSF isn't alone in thinking software patents are bad. I'm sure most ISVs agree (including Microsoft, who has had to pay hundreds of millions in patent-related lawsuits). The reasons these ISVs build patent portfolio is to protect themselves from litigation, not to protect their R&D investments.

Software patents should be scrapped, pure and simple. Copyright law is quite sufficient to protect IP as far as software is concerned.

Actually, no. The GPL is an extension of copyright law, i.e. the copyright holder gives you *extra* rights, as a recipient of the code (in other words, he gives you rights you normally wouldn't have). There are, however, conditions on those extra rights, one of them covering patents. If you don't respect those conditions, then the copyright extension no longer applies, and therefore you lose those extra rights. That is *completely* within bounds of the law, and does not contradict patent law at all.

Some patents may exist for good reasons, but it is difficult to make any good argument in favor of *software* patents. The FSF believes they should be scrapped, and therefore the GPL (version 2 *and* 3) have anti-patents provisions. Don't like 'em? Don't use the GPL or GPLed code.

"Don't use the GPL or GPLed code."

You're a large business utilizing open source components to drive your product, or have an open source product.

Some small third party decides to upgrade their code to GPL3. This code is directly used in your own project, and you do not wish to use the GPL3, or are unable to do so - you will have to re-engineer your code to use the third party code as a library, if possible at all.

The GPL is viral. At some point, maybe now, maybe three or five years down the line, any company currently utilizing GPLd code (and that's a lot) are going to run into the GPL3 and have difficulties or conflicts trying to manage their code.

It is never as simple as just "not using" the GPL. Thousands of companies have a huge investment in GPL2 code. The GPL3 only adds to the burden of companies trying to diligently keep within tricky licensing requirements.

Kroc wrote:
"Some small third party decides to upgrade their code to GPL3. This code is directly used in your own project, and you do not wish to use the GPL3, or are unable to do so - you will have to re-engineer your code to use the third party code as a library, if possible at all. "

and what would your reasons for not being able to use GPLv3 be? employing restrictive drm in your project? or including patented code which you'd later like to sue others for using? if so, the very reason this small third party decided to upgrade to gplv3 would likely be so that their code could NOT be used in your project and the likes of it.

I think they should have that right over THEIR code.

Some small third party decides to upgrade their code to GPL3. This code is directly used in your own project, and you do not wish to use the GPL3, or are unable to do so - you will have to re-engineer your code to use the third party code as a library, if possible at all.

No, all they have to do is "fork" the last version that is gpl2, and stick with that one. They have the choice of upgrading to the gpl3 version or not. That simple.

They also have the option of not using it at all if they think getting it for free under the gpl3 is not worth the hassle.

No, you don't. The original code you used is still licensed under the GPLv2. Even if it is relicensed later on, the copy *you* got was under the GPLv2, therefore that is the only license that applies.

You cannot, for example, release code under the BSDL, then when someone uses it in their own software (under the BSDL or, say, a proprietary license) change it to the GPL to force them to change their own license. That's simply not how it works.



No, it doesn't, unless you want to use new code that was never licenses under the GPLv2. Your argument is based on a false assumption, and therefore is false as well.

Eeh... using the code as a library is linking so the GPLv3 kicks in.

Besides that a license chance won't hurt you. You can just keep using the code released under the former license. Or rewrite your code. You don't have to re-engineer anything, since you have source access to your own product.

Add to that that you are using the property of somebody else. If you can't accept the terms, cease to use it. Or accept the terms. If it truly is a small third party thing it is easy to replace. If you don't want to do that you are simply to lazy.

It is never as simple as just "not using" the GPL.

Ah, you've hit the nail on the head! The GPL is a copyright license crafted to illustrate the paradox inherent in the superposition of patent and copyright law in the context of the software industry. The rhetoric of freedom and morality is an overlay that binds a social movement to a political statement against exclusive patent agreements: licensing, cross-licensing, and now covenants.

The very existence of the GPL is a threat to any distributor that relies on exclusive agreements to distribute their products. The great success of the GPL software has turned a hypothetical threat into a troubling reality for many vendors. The GPL requires vendors to do what many simply can't--to grant their customers any rights required to redistribute the software. Any third-party upstream code they use can become GPL at any time, it extends to any linked code, and it propagates through distribution.

The GPL intentionally polarizes the patent landscape in the software industry. Vendors can depend entirely on exclusive agreements, or they can reject them altogether. There is no middle ground. Is this fair to vendors that already have numerous exclusive agreements? No. It really puts them in a tight spot. Maybe they should think about lobbying for patent reform?

The intent of the GPL is to use copyright to stop patents from destroying the software industry. It's not fair to the beneficiaries of the present system, but they really don't have any choice in the matter. These giants thought their money and clout could protect their interests, but they're not going to lobby to drastically weaken copyright in order to protect their interests in patents. As much as their businesses rely on patents, they rely on copyright a whole lot more.

Note that the GPL is not against software patents in principle, it's against the way they're used in practice. If a software vendor wants to patent an innovation and protect their limited monopoly to exercise that innovation, then that's justifiable. But software vendors don't typically pursue patents to establish limited monopolies. They seek patents to defend themselves against the patent system. They amass patents as political capital and leverage in the back-door negotiations that control the balance of power in the industry.

The Microsoft/Novell deal was about Microsoft trying in vain to introduce the free software community to how industry giants use the patent system. Free software is getting too powerful and influential to ignore in the industry. Microsoft is trying to get free software vendors to sit down, drink brandy, smoke cigars, and negotiate terms for how we're going to coexist as new members of the industry elite. We've been invited to the secret club.

Novell was kicked out of this club some years ago and wants back in. To the rest of the community, this pomp and drama is strangely foreign. Why not compete on the merits of your offerings and let the market decide the balance of power? Because then the establishment can't compete with agile startups and their disruptive technologies. These corporations believe that their participation in the formative era of the modern software industry entitles them to extra power today. Software patents serve as a legal basis for this power, providing the fuel for the fear tactics that hamstring the industry and its customers.

It really doesn't matter whether or not you believe that software patents are hurting the software industry and its customers. The very existence of the GPL threatens vendors that participate in this patent licensing charade, and its success puts the future of these vendors in jeopardy. It's ironic that, in a patent landscape that begs for cold war metaphors, the real nuclear option turns out to be a copyright license.

Edited 2007-06-02 01:54

My heart bleeds for them. NOT.

I don't hear you bitching about changes to proprietary software EULAs which continously remove end user rights. Poor big business. It's so hard...I mean use someone elses code, abuse it, make profit and then bitch when you have to give something back to the community. wah wah wah.

Dave

100% correct in my eyes.

Dave

It is the other way around: GPL got patent troubles? Then kill the software.

I strongly believe the FSF should rethink its patent provisions.

Software freedom more important than software patents in GPL (Restate without Freudian slip )

I strongly believe the FSF should rethink its patent provisions.


One of the most frequently overlooked yet obvious facts is that the FSF is representing an entire community whose priority is the four freedoms and so the FSF must consider other issues such as software patents within the constraints of that priority. With the GPL3, the FSF has already bent over backwards to satisfy those who want to profit from damage to those freedoms (e.g., Tivoization allowed in business and organizational contexts).

The obvious question then is, in what ways could the FSF rethink the patent provisions with no further compromises of the four freedoms? Unless you have an answer for this, I suspect that it is you who needs to rethink some things.

Edited 2007-06-04 00:23

Why? The patent provisions of the GPL amount to these: (1) if you hold a patent that pertains to software you are releasing under the GPL, then you agree to give all recipients of the software the right to use that patent, and (2) if you are "conveying" GPL software to other parties, then you agree to give everyone the same rights to the code, not just to the people you actually supplied the software to.

If you don't agree to those terms, then don't distribute any code under the GPL.

Why should the FSF re-think these provisions in your view?

They simply spell out the same provisions as are implied (but not explicitly stated) in the current version of the GPL.

What exactly is wrong with these provisions, in your view, that requires the FSF to re-think them?

Edited 2007-06-04 01:15

It's odd that people single out the GPLv3 language on patents, ignoring the very similar language in other open source licenses, like those used by Apache or Mozilla.

I think you hit the key point there -- 'ignoring.' Who uses the Apache or Mozilla licenses, outside of their creators?

Code that builds on the Apache code uses that license. GPLv2 isn't compatible with Apache so code can't be shared, but GPLv3 is compatible.

I think you hit the key point there -- 'ignoring.' Who uses the Apache or Mozilla licenses, outside of their creators?

I use the MPL.

@Hiev

I personally don't support GPLv3 and I'll try to avoid it in the software I use or make.

You will not be able to avoid it because most, if not all, GNU packages will become GPLv3 ...

]{

Then I'll switch to something else.

Like what? All the BSD's use GNU packages, as does Mac OS X and even Solaris is GPLv3.

So that only leaves, you know, a Microsoft OS or one of the hobby OS'es.

Sort of a faustian bargain, don't you think?

]{

So be it.

I honestly prefer MS over the FSF any time.

Edited 2007-06-01 17:14

Good bye and good luck ---

GNU tools come with an exception clause, meaning that you do not have to make your code, that uses these libraries and tools, does not have to be GPL. It can be proprietary if you so wish.

The issue is in libraries or code for entire applications constructed by other users that you'll want to use and that may be GPL3 (or GPL2 for that matter). They may or may have the exception clause and you'll have to comply likewise.

Don't kid yourself. GNU projects were forked before, they can be forked again.

Just like the linux kernel, GNU has evolved from the point where it's a group of grassroots developers banging away on keyboards in the name of free software. There are companies like IBM, HP and Intel providing contributions, and not in the name of the four freedoms, but simply because it facilitates their own corporate objectives, with the benefit to the free software community being an unintentional benefit.

Not saying it will happen, just pointing out that it could. The v3 proponents inevitably point to the GNU toolchain and utilities as some sort of inalienable influence, when the fact of the matter is they could be forked overnight if the companies floating a considerable portion of the code as part of their investment in the multi-billion dollar Linux Inc. market decide they're uncomfortable with the uncertainty or even extremism surrounding v3.

And there lies the gamble for v3, which is the predominant reason that they chose a quasi-open process for drafting (I say quasi-open process because Moglen made a point of mentioning at the beginning that v3 would not be drafted by consensus, they were merely looking for input). If the industry rejects v3, then the free software community is SOL because those vast resources will be quickly diverted to v2 alternatives. Under the current licensing, the FLOSS, OSS and commercial segments have achieved a remarkable, even unprecedented, equilibrium that benefits all. Don't underestimate the significance of that, or assume that it will be a simple transition.

No way to know how things will shake out until the license is released and the dust will settle, so it's just food for thought.

True.

The thing I would note, however, is that you seem to be cheering for the losing side.

Microsoft has made patent noises about FOSS. GPL3 protects FOSS code against patent noises far better than GPL2 did, and GPL3 has potential to disarm the Microsoft/Novell attempted work-around.

I would guess that most FOSS projects would opt for moving to GPL3, and there is a strong chance parts of the kernel will also. Even Linus has wavered in his initial opposition to GPLv3, and the murmurings from Microsoft have got to be a strong incentive to consider a move to GPL3, even for the kernel.

All of GNU will certainly go to GPL3, as will critical projects like samba.

You can have a mix of GPL2 and GPL3 code in a given Linux distribution, there is no problem with that.

What you will not be able to do is hold to a policy of GPLv2 only, and expect to continue to compete as a GNU/Linux/FOSS distribution.

Not saying it will happen, just pointing out that it could. The v3 proponents inevitably point to the GNU toolchain and utilities as some sort of inalienable influence, when the fact of the matter is they could be forked overnight if the companies floating a considerable portion of the code as part of their investment in the multi-billion dollar Linux Inc. market decide they're uncomfortable with the uncertainty or even extremism surrounding v3.

It's nice to be reminded that human beings are (increasingly it seems) at the mercy of corporations.

That's a lie. The GPL does not kick in for such situations since such elements are trivial and the GPL does not kick in for trivial elements.



That's not true. You can choose any license for your code and even link to GPL. Only in regard to distribution of binaries does the GPL kick in. And it only limits your freedom to choose a license if you take the code belonging to somebody else. Nobody forces you to that. Ergo you are not forced.

Your claim is equal to this: "McDonalds demands I pay them in order to get a hamburger. Ergo McDonalds are limiting the freedom I have to take what I want."

OTOH, in a perfect society (Utopia) the copyright law would be identical to the MIT license (or the more complex revised BSD-license). But since humans are imperfect and the law reflects this the GPL is the best choice for most things. Since you don't code and don't make a living it's easy for you to choose the MIT license (my favourite in regard to simplicity - the GPL is getting more and more difficult to understand).

That's a lie. The GPL does not kick in for such situations since such elements are trivial and the GPL does not kick in for trivial elements.

I'd love to know a complete list of all the times you can use GPL code and not adhere to its license. You keep bringing them up.

Basic copyright law: you can't copyright a sentence or anything trivial. You can't copyright a name either (that's why we have trademark law).

The GPL is based on copyright, ergo it cannot negate any part of it, such as provisions for fair use, which would cover such things as a single line of code.

Thom's post was actually quite well-argued, and over took a very moderate approach. I agreed with most of it, but he *is* wrong in stating that a single line of GPL code would be enough to make a program GPLed.

Absolutely.

There is a website about GPL violations.
http://gpl-violations.org/

Anyone who does even the slightest research about GPL violations that have actually occurred knows that there are two typical outcomes:
(1) there is only a small piece of the proprietary program that was GPL code. That piece is replaced, and everyone is happy, or
(2) the "proprietary" program is in fact only a slightly modified version of GPL code ... and the vendor has not in fact really written the code to begin with. In this case, the vendors invariably agree to publish the source code and small modifications (and in the case of embedded firmware, continue selling the product) ... and everyone is happy.

You can use GPL anyway you like. The GPL kicks in the moment you distribute. No sooner, no later.

a) You have unrestricted use rights;
b) You have unrestricted modification rights;
c) You DO NOT have unrestricted distribution rights.
d) Accepting the license is not required in order to have use rights and modification rights.

Is that easy to understand?

According to the GPL the license only covers distribution. Therefore you cannot "have to abide by the license" in order to USE or MODIFY the code. You only have to abide by the license if you DISTRIBUTE. And you can only abide by the license in case of DISTRIBUTUION.

It really scare me how the "Advocates of freedom" make any excuse to justify the obvious misstakes, think by your selves at least once.

The original claim tried (and failed) to push a trivial case as resulting in a conversion to GPL. That was the claim, that the case could be trivial. What you are implying, linking to a GPLed library, is a nontrivial use of code, and all of the nitpicking in the world is not going to change that, dude.

Edited 2007-06-02 00:11