ESR: 'We Don't Need the GPL Anymore' (O'ReillyNet)

Posted Jul 1, 2005 17:51 UTC (Fri) by Ross (subscriber, #4065) [Link]

There are some potential logic faults here. One of which is that not all
projects are actively maintained -- sometimes there's just not much of a
development community left. I've personally seen this and it's a problem
even if the GPL is used. If the copyright holder's can't be found, the
license terms can't be enforced. Another is the fact the GPL is so widely
used it is a standard. You can move code between GPLed projects with ease
and minimal worry. You don't have to worry about strange unbalanced rights
like created by many of the corporate GPL replacements. I know that I have
to really consider things much more carefully before modifying non-GPLed
software (and in some cases decided not to).

Posted Jul 1, 2005 18:13 UTC (Fri) by bajw (subscriber, #11712) [Link]

I know that as a user of software, I am now to the point where I try to run exclusively GPL'd software when I have a choice. Other Free Software licenses are what I consider "second best", though I will affirm that the BSD systems are attractive to me, and I do use and support them.
The only place I use unfree software is at work, on computers I don't own. (Unfortunately, this is one...)
I'm not a developer, so my opinions may differ a bit from the opinions of many developers, but those are my choices as a user.
Sad to say that ESR has sunk below the background noise level for me, and doesn't provide a signal I care to bother with any more.

Posted Jul 1, 2005 18:26 UTC (Fri) by jimi (guest, #6655) [Link]

Not every project has huge numbers of developers behind it. The projects I'm involved in have at most six developers, of which three at most continue to contribute. Nevertheless, we have a substantial user base. But we couldn't compete against even a small shop with dedicated developers. The GPL protects our hard work.

But surviving closed source competition isn't even the point. The GPL is for the citizenry at large, not just a few developers. It gives everyone rights and freedoms that they would otherwise not have. The software does not need the GPL, we do.

It's not the software that's free; it's you.
- billyskank on Groklaw

Posted Jul 1, 2005 18:31 UTC (Fri) by QuisUtDeus (guest, #14854) [Link]

What is missed in ESR's remark is that big companies may someday want to take old/abandoned code (or new code by a resource-strapped developer) released under GPL and make proprietary versions of them.

Maybe ESR thinks this a good thing (since the code gets developed somehow, if not as open source).

By the time a project is large enough to compete head-to-head with new forks, big or small, the choice of license has long before been made.

How many big, thriving open-source projects would have gotten that far without the GPL to protect their early efforts? One could argue that it is a sign of the fitness/goodness of the GPL that there are indeed many large, active projects released under the GPL.

Could it be that the failure of the BSD-related systems to enjoy growth comparable to Linux (even now, after the fame of Linux has brought more attention to them) is due in part to their not using the GPL? If a developer wants to write code and give it away, will he want other companies to derive profit from it without him?

Posted Jul 1, 2005 19:05 UTC (Fri) by Junior_Samples (guest, #26737) [Link]

The GPL excels because it seems to coincide with what mathematical game theory would predict as the best strategy to insure an acceptable licensing outcome while minimizing risk. It is doubtful that SGI would have released the XFS under any other license than the GPL.

The GPL recognizes human nature. Less restrictive licenses depend on a Utopian world view.

However ESR might have a point with regard to mega-projects such as Linux. With smaller or fledgling projects it is a different story. Even fairly large projects such as Wine have suffered unfortunate forks allowed through lax non-GPL licensing. Another classic example of licensing failure is the SPICE circuit simulator whose non-GPL licensing has led to a myriad of incompatible component engineering models.

Posted Jul 1, 2005 19:41 UTC (Fri) by proski (subscriber, #104) [Link]

Quoth ESR:

The pros and cons of "viral licensing" is something I've been thinking about a lot recently. As far back as 1998, I suspected that allegiance to the GPL is actually evidence that open source developers don't really believe their own story. That is, if we really believe that open source is a superior system of production, and therefore that it will drive out closed source in a free market, then why do we think we need infectious licensing? What do we think we gain by punishing defectors?

To follow that logic, forbidding bribery is evidence that democracy doesn't work. If we really believe that the people can elect the best, the most honest and most qualified representatives, why do we need to forbid them take gifts for their work? Who do you think we are punishing?

The thing is, outlawing bribery is a necessary part of democracy just like outlawing non-free forks is a necessary part of free software development. There are many things that would not be needed in the ideal world, but the world is far from being ideal.

Posted Jul 1, 2005 20:21 UTC (Fri) by BrucePerens (subscriber, #2510) [Link]

We don't need ESR any more :-)

Posted Jul 1, 2005 20:31 UTC (Fri) by stijn (subscriber, #570) [Link]

Is it not true that companies should *want* the GPL? Elsethread somebody mentions SGI and XFS; I remember the same thing being said about (by?) IBM. The GPL levels and widens the playing field, and blocks off embrace/extend/strangle routes. This seems *especially* true for Linux, and I am positively sure that companies benefit from its license. The GPL is the best thing since sliced bread. Then again, it is about freedom, and Raymond's quest for world domination and focus on market share, market penetration, marketing and whatnot are in that respect wide off the mark.

Posted Jul 1, 2005 21:47 UTC (Fri) by Max.Hyre (subscriber, #1054) [Link]

Quite the cage-rattling. It's sad to see ESR losing his grip to this extent. From the first answer (``People who do what the GPL tries to prevent (e.g., closed source forks of open source projects) wind up injuring only themselves. ''), ESR speaks as if there is 0.0% difference between ``open source'' and Free Software. The preceding argument is false, if you believe freedom is the requirement. Anyone creating such a non-Free fork is locking users into whatever mods the forkers made to it. Further, he spreads, quite freely, the ``GPL is viral'' canard.

To respond specifically to few items:

Some time ago there was a monetary offer to get a Linux snapshot under BSD license. Would you have accepted? I wouldn't have had the right to accept Jeff Merkey's offer; I'm not Linus.

Is it certain that Linus will adopt GPL v3? It depends. I know for a fact that he is concerned that GPL 3.0 will overreach.

The GPL includes a clause that automatically shifts the license terms to any new version of the license itself. Isn't this a Trojan horse? No, because the clause says ``at your option.'' That is, the person receiving GPL v2 software gets to choose whether v2 or the later version applies.

[....]

The user chooses only for copies of the software still under GPL 2.

A total misunderstanding of how changing terms to a later license can happen. Linus didn't (and doesn't) have any more right to accept Merkey's money than ESR. Linux isn't his work to dispose of as he sees fit. It just as impossible for Linus to convert the kernel to GPL v.3. because many chunks are licensed under v.2 only.

It's the licensor's decision to add ``or, at your option, any later version''; that's not in the GPL itself. If it's under GPL v.2 without that clause, the user cannot upgrade the license. That's going to burn the kernel badly if there turn out to be compelling reasons to move to v.3. It requires finding all the contributors and getting their permissions. I strongly suspect that for chunks of it, it's not even known who the contributor is, much less whether they've moved to Tibet for the life of contemplation, leaving no forwarding address, to say nothing about those (we hope, few) who've died. Tough to get permission.

I suspected that allegiance to the GPL is actually evidence that open source developers don't really believe their own story. That is, if we really believe that open source is a superior system of production, and therefore that it will drive out closed source in a free market, then why do we think we need infectious licensing?

Here again, he confuses open source with Free Software. GPL users aren't claiming it's a superior development method. It's about giving user freedom, and ensuring that they get to keep it.

Q: It seems that Red Hat is selling its GNU/Linux distribution under a sort of user license that limits the freedom No. 2 provided by the GPL. The short version of the story, as I was told, is that if I buy a CD/DVD with the last Red Hat version and I make an ISO from that and put that online, I'll get sued. [....]

[I]f you republish a RHEL CD in either form, you could get sued for illegal use of the embedded trademark. [....] As I understand it now, what Red Hat has done is legally blocked republication of its entire RHEL distribution even though any component part is still GPLed and therefore freely redistributable.

[....]

What you can't ``share'' is Red Hat's integration work and branding.

Actually, what you can't share is Red Hat's branding: if Red Hat wants to keep their trademark, such a restriction is required. To redistribute the ISO, all you need do is grep `Red Hat' and globally replace it with `Foonly '. Burn that, there's no problem, and you've got RH's integration . (If there's nothing but Free (not necessarily copylefted) software on the image.)

While in the middle portion of the interview he speaks cogently to the relationship between freedom and innovation, and the importance of keeping your data out of proprietary traps, the rest is very unsettling. I hope he's just suffering a concussion from which he'll recover soon. I'd hate to think he's gone over to the Dark Side.

Posted Jul 1, 2005 21:56 UTC (Fri) by dkite (guest, #4577) [Link]

There aren't many absolutes in any racket, but there is one when it comes
to software:

The code will be made secret.

The reason is simple. All the value and work is represented in the code,
and if someone wants to guard it's value, they hide it.

The GPL takes this situation and gives value to collaboration and
openness. My contribution becomes more valuable when you contribute, and
visa versa.

There may be other licenses that accomplish the same thing. To suggest
that the one who started it all is obsolete is strange.

Derek

Posted Jul 1, 2005 22:42 UTC (Fri) by rknop (guest, #66) [Link]

I'm speaking out of my mouth here, without any data or analysis to back it up, but here's an idea I think worth considering.

Linux would not be as big, or as prominent in education and business, as it is today WITHOUT the GPL.

Reason: it would have been embraced and extended. From the point of view, Linux would have bene synonmous with Red Hat. (Indeed, it was, even in our with-GPL world, and still is to many.) However, without the GPL, at some point somebody in Red Hat would probably have figured out that they could ensure customer loyalty by not pubishing their extentions as open source. There would have been proprietary extensions in the most used distribution, and that would have become the business world's Linux.

Then Microsoft would have had a target to buy, and that would have been that.

Sure, Linux would continue to exist, as *BSD exists now, but it would still be in the state that it was in the lat 90's -- used by geeks and hobbyists, and by people who really knew what they were doing, but not widely recognized as an operating system that was a player. Just as Solaris and other proprietary Unix derivatives were much more important players in the early to mid 90's than the BSD from whom they heavily borrowed.

The fact that Linux was GPL meant that no company could completely take it over. The best they could do was offer the best support and services possible, so that people would want to stay with them; but no company could count on the sort of lock-in to the same level as Microsoft. That meant that there was no single target whose going down would take down Linux, either by being bought by Microsoft or just by drifting out of existence (as, say, Commodore did to the Amiga).

The GPL has been a huge philosophical boon to the free software community. It has also been a huge practical boon.

-Rob

Posted Jul 1, 2005 23:26 UTC (Fri) by gallir (subscriber, #5735) [Link]

Oh my God.

It's a pleasure to read the informed and smart comments in
LWN, specially those related to the importance and impact of GPL and the
ethical proposition of the free software movement.

I think these kind of balanced analysis of "free software" vs. "open
source" --specially if RMS was mentioned even indirectly-- or BSD vs GPL
were very uncommon few years ago. The boring slogan was BSD is more
free...

Perhaps people start to realise that no social movement is sustained only
by technical reasons.

Posted Jul 2, 2005 0:04 UTC (Sat) by donbarry (guest, #10485) [Link]

The entire framing of the "contest" by ESR is false. The Linux part of
GNU/Linux was merely the first practically useable kernel -- and would have been pointless had not a decade of work, essentially all of it under the GPL license, been made possible by the FSF and Stallman -- and -- the valuation of freedom as a significant virtue, and its codification in the terms of the GPL.

What was a virtue in the 50's and 60's (share and share alike on computers) was one because the prevailing values were set by academics. As corporations began to dominate the supply of not only hardware but software, they attempted and to some extent succeeded in displacing academic ethics. Stallman's great and signal victory was to reinject academic ethics, indeed, essentially the Golden Rule, back into software production as an intrinsic value, not an afterthought.

ESR's public career, starting much later, was essentially as a parasite to this movement, alternately being a critic or cheerleader in whichever capacity would get him the most press, but at all times rejecting any virtue in freedom as a value. If he developed his ideology in any consistent manner, he'd find himself some sort of third-rate Kropotkin.
The ideological confusion he has sown together with his tendency to
self-aggrandize in his rewritings of history which sometimes border on the delusional have certainly made a place for him as a figure in the history of these times, though not perhaps as the historical figure he'd like to be known as.

Posted Jul 2, 2005 0:31 UTC (Sat) by neilbrown (subscriber, #359) [Link]

By chance I was reading http://www.paulgraham.com/raq.html recently and
particularly "I'm about to become a teacher. How can I be a good one?".

Point 1 is about respect. You need to set and maintain high standards. Making the students like you is not so important and making the students respect you, and you do that by having high principles and sticking to them.

I see a connection between this and ESR's comments.

He seems to be saying that the reason for dropping the GPL is that is scares off potential members of our community, and the justification is that we don't really need it any more.

This seems like trying to make people like us by lowering our standards.

Paul Graham, as I read him, would disagree with that approach (at least if we see ourselves as teachers). So would I.

NeilBrown

Posted Jul 2, 2005 3:25 UTC (Sat) by clump (subscriber, #27801) [Link]

If I understand ESR's GPL position correctly, in that Open Source is a vastly superior method of development irrespective of the protections of the GPL, then I am afraid he didn't cite a lot of evidence for this claim. Further he states that the protections offered by the GPL harm the community because they "scare" away potential contributors, which is also a claim that lacks evidence. I think the issue isn't that easy to frame.

Whether the success of Linux in particular is merely that of Linus Torvalds or because of the side effects of a software license seems to miss parts of a bigger picture. Firstly, we're in uncharted territory; Linux is currently very successful and doesn't have a lot of historical perspective to look at. The GPL is very much a part of this success. It would be disingenuous to try to diminish the GPL's contribution to a community it helped create. Secondly, there have been very large and visible Linux contributions under the GPL. This very much is evidence of the opposite of ESR's point that the GPL scares contributors away.

My view: ESR says that the racecar going 200Mph never needed 4 wheels. I believe ESR might disagree and say: the GPL to Free/Open Source is nothing like a wheel to an automobile.

I am afraid I am not convinced that the GPL is unneeded. Linux and Free/Open Source software are alive and well at present, thanks in part to the GPL. Saying in 2005 despite this success that one of its key ingredients is unneeded doesn't go far without evidence. There is a lot of evidence to show the GPL's role in support the community, but not much to condemn it.

Posted Jul 2, 2005 4:28 UTC (Sat) by bojan (subscriber, #14302) [Link]

ESR: "That is, if we really believe that open source is a superior system of production, and therefore that it will drive out closed source in a free market, then why do we think we need infectious licensing? What do we think we gain by punishing defectors?"

Well, do we _all_ believe it? I would think that FSF in particular (i.e. the creators of the GPL) don't think that (or don't find it important). They want guarantees that _users_ won't get hurt. Ever. They want - well - freedom.

Some developer groups don't share that view (or at least it is not in their policies), such as Apache Foundation, and they make good code too. Nothing wrong with either, of course. However, the "we don't need the GPL" argument is completely misguided as it doesn't take into account the basic premise - people that are "in charge" of the GPL don't share ESR's views at all. So, they sure do need the GPL and whoever believes freedom is important (not just the production model) still needs the GPL or the closest equivalent.

For a smart guy, ESR sometimes doesn't make sense at all.

Posted Jul 2, 2005 12:09 UTC (Sat) by cjcoats (guest, #9833) [Link]

I am the lead on an open source project of relatively narrow interest (I/O and utility libraries and related programs for environmental models -- see http://www.baronams.com/ioapi/), and for that scenario Eric's statement

...They trap themselves unto competing with a small in-house development group against the much larger one in the parent open source project, and failing.

is wrong.

I have twice had to go up against much larger organizations--US EPA and the University of North Carolina--for license violations. Both of these organizations had tried to distribute "hacked" distributions with the copyright and attributions stripped out. Without the protection of the license, and with their institutional lacks of architectural vision, these organizations would have made a mess of the library the way they have with their own in-house models (CMAQ and SMOKE, respectively) that use this library.

Posted Jul 2, 2005 12:51 UTC (Sat) by job (subscriber, #670) [Link]

Raymond wasted many dead trees on his writings on what defines hackerdom,
coincidentally they all match his own values, long before the Linux
phenomenon, and many people are still angry with what de did to the
Hacker's Dictionary. He was lucky to catch the Linux updrift and one of
his papers caught on.

Thank you for that paper. It was a short and very interesting read. The
people who managed to configure and use fetchmail also thank you for
that. Now will you please stop comparing everything you see with that
software and every other people with your own image of yourself? Pretty
please?

He has for many years some sort of campaign against Stallman. Perhaps
because they share vastly different political views personally (which
might help explain why it is so important for him to define hackerdom).

It is also interesting how quickly he buddied up with Torvalds, while the
connection may not be entirely symmetrical, as he must have ways to
connect with other people. One paper and a mail fetching script may not
be enough to get your voice heard.

As for the license in question, Stallman knows what he's doing. He has
laid the foundation for the GNU project and his enormous work on the GPL
is part of the reason why we can run a free software desktop and still be
accessing almost all of the same services the rest of society uses.

There is enough empirical evidence that its role is important, and when a
company gives code to the Linux project, they need to be able to explain
to the shareholders why their competitors can't embrace and extend it. It
is likely that license is the reason that the dot com boom didn't focus
on FreeBSD, which at the time was more stable. There are exceptions such
as the Apache project, but as a daemon it is limited in scope and without
the economic value of Linux.

Posted Jul 2, 2005 17:10 UTC (Sat) by sbergman27 (subscriber, #10767) [Link]

My main question is this. Why did Eric just slit his own throat? There have been a couple of suicides by throat slitting lately. Maureen O'Gara did it. OK, that was, perhaps, not terribly surprising. But I do find Eric's sudden "suicide" to be surprising. I'm an OSS guy more than a "Free Software" guy. But this, more than anything else that I have ever seen, validates RMS's concerns about OSS. Without the GPL, exactly what would encourage corporations, for whom self-interest is not just the primary, but the *only* concern, to cooperate with each other, and with the rest of the community, out of pure self-interest?

OK. The GPL is a bit of an obnoxious license to those who want to exploit others. It's even a bit obnoxious to those who don't. But it is the glue that binds us all into a real community. Corporations, individuals, nonprofits. It is in all of our best interests to cooperate because we know we are not going to give, give, give, and then get shafted.

When human beings are involved, the optimal social mechanisms tend to be rather less efficient than the theoretical maximum.

Posted Jul 2, 2005 20:36 UTC (Sat) by peace (guest, #10016) [Link]

Bruce Perens mad a quip above about maybe not needing ESR anymore. What this whole episode has made me wounder is if we need Open Source anymore. I have always been torn between the ease of communicating the concepts of Open Source and the more awkward conversation that generaly ensues around Free Software.

Observing the the rif raf that seems to be jumping on the Open Source bandwagon has made me question whether the Open Source "movement" has the central value structure that is needed to ward off evil in the face of temptation. I have no such doubts about the FSF.

I have also had the pleasure of seeing RMS speak a few times, the last time at the Boston Social Forum. He is dedicated and clear thinking and most importaintly entirely without showmanship and spin. Open Source lately seems to be getting cought up in the glitter and lime lights.

Ranting aside, Open Sourcers have done a ton of great things for Free software and the world at large. No doubt, major props. However, ESR's statements are indicative of a rot that could confuse the value of user Freedom with the concept of a development model. This would be a terrible thing.

Kind Regards

Posted Jul 4, 2005 1:22 UTC (Mon) by lilo (guest, #661) [Link]

I'm sympathetic to Eric's view, but I'm not convinced his view of things is good for FOSS. FOSS is not just about the code, but also about the communities that surround it. I've seen a lot of people who are into open source just to have the source code. Some of them stay with us, but many of them do not. People with a moral focus and an ethical rationale tend to stay around for the long term.

You don't have to use the GPL to contribute to the cohesiveness of the FOSS community, but the GPL, and the people who believe in it, have contributed a lot to that cohesiveness. It'd be unfortunate if Eric fostered the impression that cutting loose the GPL would be a good idea.

Posted Jul 4, 2005 3:22 UTC (Mon) by bignose (subscriber, #40) [Link]

> I've actually been making the argument that the GPL is rationally justified only if open source is an inferior system of production since 1998.

That may be because you misread the intentions of the GPL: to give freedom to software users and ensure they can keep it. Until there's no threat that any users of software could have their freedoms unfairly restricted, I don't see how any tool that works toward that goal can be obsolete.

The GPL's usefulness is to the users of the software, and only incidentally to the software author. It so happens that software authors often like to please their users, so licensing software under user-friendly terms has a benefit to the author as well. That benefit, though, is *not* the purpose of the GPL -- protecting the users' freedom is its purpose.

Regardless of whether or not free software is technically superior, it is *always* superior in terms of users' freedom. So long as the GPL works toward this end, it's entirely relevant.

Posted Jul 4, 2005 8:40 UTC (Mon) by Acrisius (guest, #22348) [Link]

Eric's thesis of why people write free software is summed up in one of his essays, I forget which. He says that it is simply a natural consequence of human creativity: wrap the Internet around a mind and software flows (by analogy to electromagnetism: wrap a magnetic field around a conductor and current flows). As though it were a natural law.

I think he's only got part of the picture. I think that many of us are motivated to create and contribute free software by the very ideals embodied in the GPL. The GPL may be the most widely used license in the noosphere by historical accident, but I suspect not. Most of us know in our gut that, in a large enough crowd, there will always be some arsehole who acts selfishly and spoils it for everyone. Maybe that arsehole will be me, given a big enough incentive. Game theory is all about this. The GPL keeps the arseholes on a leash. The BSD license and its ilk do not.

Openness is a virtue. For example, the government of the U.S.A. is one of the most open in the world, in the sense that much of its process is documented and publicly available (disclaimer: I'm not American). That openness means that it's hard for the government to reduce the freedom of its people, but the openness itself is not enough. People also have to be active in their defence of their freedom.

Here's a natural law of human growth: to gain something, you must sacrifice something else. To gain freedom, you must sacrifice the comfort of ignorance and apathy (at least). To gain free software, apparently we must sacrifice its absolute exploitability. Hence the GPL. This is the point that Eric seems to miss.

Posted Jul 4, 2005 16:02 UTC (Mon) by Seegras (guest, #20463) [Link]

Anarchy in the original sense of "a state without a government" is actually also a hugely superior system of production. The few times it was tried out, like in cataluna in the 30ies, production rose to incredible heights, because everyone was doing what he liked (and could do) best. Still, anarchies didn't last, they were destroyed from within, or even more often, from the outside.

Just because something works in itself, and is more successful than everything else does not make it immune to attack and destruction.

There is no "free market" out there in which "demand determines supply"; instead there are hundreds of thousands of suppliers out there trying to fuck up their competition with whatever means they have access to, most notably by making the governement write laws which will kill off the competition. If you can't compete on the market, wage war against your competitors, or make them give you a monopoly, one on ideas for instance (it's called "patents"; and they have been used for crushing competition since they were invented in the 18th century). And so on. (I won't go into the patents debate, I only use this as an example to show that there is no "free market").

So the GPL is a necessary defense, for the writers themselves, and for the public as a guarantee that what is free will remain free.

Posted Jul 8, 2005 20:51 UTC (Fri) by dwheeler (subscriber, #1216) [Link]

A few random thoughts. I see no need to attack the messenger; let's instead discuss the issue.

Clearly, some choose the GPL, some the LGPL, and some the BSD-style licenses. There are successful projects representing each. I think it's untrue that "all" OSS/FS projects should choose a BSD-style license, so let me give a few justifications as to why I believe that. (Note: I've developed GPL, LGPL, BSD-style, and proprietary code. So I've seen all sides.)

First, the "best" license from the point-of-view of the original developers depends on the developers' goals. The "best" license from the point-of-view of the user may be different, but if the user didn't pay for the product's development, then the user is just whining. Those who pay the piper (by money or sweat) get to call the tune. If you (the user) don't like the license, then use a different product or develop it yourself (possibly by paying for its development). In the end, only those who invest in something get to choose its conditions. Different goals, different "best".

Second, there's some marketplace evidence that some projects do seem to do better with an LGPL or GPL license than a BSD-style, because they can (at least in some circumstances) attract more development work and/or more users:

1. For a specific case, WINE reports "[after] changing Wine's license from BSD-style to LGPL ... The number of patches steadily increased after switching to LGPL (after [decreasing] previous to that.) Now we're seeing between 4x and 5x as many [patches] as we did in late 2001 and early 2002." So switching from the BSD license can aid a project, not an expected result if the BSD license is always superior.
2. The *BSDs were ready before the Linux systems, but while they're quite capable and are not going away, they are definitely not the leader in development resources or in user market share. They could have copied Linus Torvalds' development process by now, it's widely understood... the process is unlikely to be the only reason. The constant "bleeding away" from the *BSDs to commercial projects that didn't give back (SunOS, BSDI, etc.) is a far more likely explanation for the current state. This isn't a knock on the fine BSD systems (I use them!). Indeed, the Linux kernel was a latecomer and zoomed past not only the OSS/FS *BSDs, but it's now savaging all the proprietary Unixes which used pieces of the BSD code. The Linux kernel now has contributors from rival companies working together, using the GPL as essentially a constitution for a consortium. Here we have a GPL'ed component storming past BSD-licensed products that had a head start; not expected if the BSD license is always best.

I suspect that in some circumstances the GPL or LGPL comforts more developer resources (with its guarantees) than it repels. This may be particularly important in the early stages of a project, enabling such projects to leap the "energy barrier" between the idea and the self-sustaining project with a useful product. There are also companies (like Red Hat, MySQL, Troll Tech, etc.) that clearly prefer the GPL. I suspect that the answer is not simple at all, and involves many factors.

I don't agree that "if OSS produces better code in a more efficient manner, then it doesn't ever need legal enforcement." Even if you accepted the premise in absolutely all cases, it is not true that all actors want the best code at the lowest development cost. Actors' primary money streams might not come from the code at all (think RIAA), or might come primarily from locking in customers so that the customers cannot later choose the best product. Evan Leibovitch stated years ago that many have decided that the GPL is necessary to deal with today's "not-so-kind computing environment".

And the same argument makes no sense for the proprietary world. Imagine a proprietary vendor who said "proprietary software always produces better products, so we're going to eliminate all those license conditions and just hope people do what we want them to do." Even if you accepted the premise, the "correlary" does not follow. Legal enforcement through licensing doesn't demonstrate that a particular approach is a failure, or is less effective. Licensing is simply a legal mechanism to enable a particular business approach.

In the end, it's the authors (and those who pay them) who choose the licenses they'll use. It's clear that a vast number of OSS/FS project authors choose to release under the GPL. And it's also clear that a vast number choose BSD, and that a vast number choose LGPL. Indeed, I suspect that there are many like me who choose different licenses depending on their motivations for their specific circumstance. I expect that no amount of "GPL vs. BSD" argument is going to make all OSS/FS developers choose any one license.

Posted Jul 14, 2005 16:45 UTC (Thu) by jimwelch (guest, #178) [Link]

Has anyone seen a graphic of the licenses (GPL,BSD,M$,...) and freedoms (change, view, hide,...) from the different perspectives(user, originator, developer(3rd party), ...)? One of those intersecting circle logic charts (Venn diagrams). Or even a table of freedoms vs license? I am preparing a talk on LINUX for novices (most over 65)