We Don't Need the GPL Anymore

jpkunst writes "In a lengthy interview with Eric S. Raymond by Federico Biancuzzi at O'Reilly's onlamp.com, ESR defends his position that 'Open source would be succeeding faster if the GPL didn't make lots of people nervous about adopting it.'" From the article: "I don't think the GPL is the principal reason for Linux's success. Rather, I believe it's because in 1991 Linus was the first person to find the right social architecture for distributed software development. It wasn't possible much before then because it required cheap internet; and after Linux, most people who might otherwise have founded OS projects found that the minimum-energy route to what they wanted was to improve Linux. The GPL helped, but I think mainly as a sort of social signal rather than as a legal document with teeth."

Comments

He's right, of course

[A nonymous Coward]

If it were under the BSD license, Microsoft would have adopted it by now, under the hood, invisibly. Windows popularity would soar even more, and its reputation for stability and speed would have made Linux distributions obsolete, thus putting a stop to all independent peer-reviewed Linux development, leaving it to Microsoft, where it belongs. Then, with the lack of competition, Microsoft would stumble, dropping the ball, possibly scoring yet another own goal, and another Unix-lookalike would spring up, only this time the developers would be so mad about Microsoft's embrace extend extinguish of Linux that they would adopt a new license, called ... the GPL!

And ESR would have another chance to get it right.

Re:He's right, of course

[dj28]

That's a ridiculous argument.

FreeBSD, OpenBSD, NetBSD, and all the other operating systems based on the BSD license were not made obsolete by Microsoft, even if they do use some BSD code. The introduction of BSD code into Microsoft Windows did not magically make it a superior operating system. Neither would Linux code in Windows.

Your entire argument is a straw man.

Re:He's right, of course

[91degrees]

Yes, there's several bits and pieces of BSD code in Windows, but that's not the point. They never decided to simply extend BSD for their own OS. Presumably they wouldn't have done the same with Linux even without the GPL. The difference between the two operating systems isn't that great.

Apple did simply embrace and extend BSD. Nobody seemed to mind, and Apple ended up with a rather good OS.

Re:He's right, of course

[pesc]

So the GPL is good because it keeps Microsoft from adopting open standards?
No, it is good because if you improve GPL software, like M$ did with the BSD TCP stack, you must give the improvements back to the community. M$ kept their improvements to themselves, legally and according to the BSD license. M$ weren't stealing. They just want to take and not give back.

GPL is very much needed

[rajeshgoli]

Agreed that GPL may not have been the most important ingredient in linux's success. But can you imagine how many people would take away your code and claim as their own, sell it and not give back to the commnuity had it not been for the copyleft "GPL"?

It's not the GPL that makes people nervous

[SpaceLifeForm]

It's the attack against GPL via FUD and software patents that make people nervous.

Counterpoint

[vondo]

The part of his interview summarized by the post is that he essentially argues that since open source software is so popular now, it can be BSD licensed because no one has the resources to outperform the OS community with their own fork. This is demonstrably false, just look at the KHTML/Apple situation. If they could truly go their own ways without Apple showing anything they did but KDE showing everything, I think it's pretty clear Apple could run ahead of KDE.

KHTML isn't the biggest project out there, but it's in the top few % for size and complexity, I'd bet. Imagine what a private company could do to a smaller project.

Re:Counterpoint

[argent]

just look at the KHTML/Apple situation

Yes, do, because this is a great example of where the GPL doesn't make any difference. Apple effectively forked KHTML (as near as I can tell, accidentally), but it was legal to do so under the GPL... they didn't need to do anything more than release the source code. Instead, in response, they opened up the CVS and the bug database. Not because the GPL forced them to, but because the chose to.

If KHTML had been under the BSDL, would Apple have taken it away completely? Legally, they could have, but they haven't done that for other open source components in Mac OS X... the source trees at opensource.apple.com and opendarwin.org include code under BSDL, APSL, GPL, and more.

AT&T took BSD code and forked it, and nobody cared until USL tried to shut down the open-source BSD... and that BSD code turned out to be just the lever that Berkeley needed to bring USL to heel.

Microsoft's using GPL code and BSD code in Interix, and that has neither let them "outperform" Cygwin nor forced Microsoft to open Interix one skerrick more. Microsoft's been using BSD code in Windows for years, but that same code was re-implemented in Linux... if the code had been GPLed, would Linux somehow be more outperforming BSD in the market, would NT have been less successful? Personally, I wish Microsoft had used more of the BSD stack rather than mostly borrowing userland tools, it would have made socket programming in Windows a lot easier... and more compatible.

So, over and over again, we see that it's not the license that matters, it's the attitude of the people using it.

The GPL doesn't stop you from forking the code base. The GPL doesn't stop different open source groups from forking the code base. The GPL doesn't stop groups using the same code base from developing functionally equivalent packages on top of that GPLed code. Heck, sometimes the only way to bring a code base forward is to fork and switch, and Nokia at least seems to think that's a great idea...

If they could truly go their own ways without Apple showing anything they did but KDE showing everything, I think it's pretty clear Apple could run ahead of KDE.

But instead, Apple is voluntarily choosing to take part on the open market of ideas to a far greater degree than any license commits them to. They could easily pull a Sveasoft and release source code grudgingly enough that KHTML would forever remain the junior fork.

What ESR's saying now is what BSD advocates have been saying for years. Companies that are interested in being productive partners will be productive partners no matter what license you use, and companies that aren't will find ways to stick to the letter of the license while completely gutting its spirit.

Without GPL were doomed!

[MindPrison]

Thanks to GPL weve got thousands of pieces of codes that the community can both learn from and distribute amongst each other.

Dont even think for a minute that the world is so "well-adapted" and would play nice if we took away GPL.

Let me take http://www.blender3d.org/ as an example. The community bought this excellent piece of 3d software free from the grasp of shareholders and re-licensed it to GPL.

Thanks to that, its relatively safe from its actual competitors such as Discreet(AutoDesk), Alias etc. This program is so powerful that it actually can compete with the big ones, I know... I use it commercially today to develop artworks for ad-campaings that bring food on the table, but the GPL license made it affordable for me to get a "start" on my own instead of having to invest thousands of dollars into expensive 3d-software.

The big companies see us as potential customers as long as Blender where inferior to their software, but now as it has grown bigger...and more companies/personal users etc. are using it...

Dont go thinking theyll play it nice forever...losing customers theyll look for an "edge" somewhere...such as a license infringement...maybe code or functions that are equal to theirs SUE SUE SUE!

Darl McBride anyone?

We need GPL, now more than ever!

Naive Optimism

[sjbe]

"I don't think the GPL is the principal reason for Linux's success..."

Someone's been hanging around too many honest engineers. This statement grossly underestimates the selfishness of people and corporations as well as the impact of a strong legal system. Look, I'm not saying the GPL is the only important factor but I can't logically see linux existing in anywhere near its current form without it. Even if most individuals would respect other people's work (and that's retardedly naive) some people and most corporations will not. In fact, corporate management has a fiduciary duty to make as much profit as possible for their shareholders and they're under a lot of pressure to do it. There are MUCH easier (and proven) ways to make high margin profits with software than the open source model. Without legal teeth to enforce keeping software in the community it simply wouldn't happen. It's pretty safe to assume that nearly all people and companies act in their short term self interest first and foremost. Always. No exceptions.

Actually ESR misses the point.

[Anonymous Coward]

I've read the article and some of what he said is just plain wrong. For instance:

> NetBSD is a worthy project, but, let's face it,
> the fan base for it simply is not large enough to
> justify spending marketing effort to recruit them.

I agree that NetBSD is cool and appreciate all their hard work. It's allowed me to have a modern desktop on my Solaris 8 system at work without having root privileges. No Linux, not even Gentoo can claim to be able to do that.

That being said, the xBSDs were actually ahead of Linux in the late 1990s. The xBSDs were more widely deployed for enterprise systems. But Linux still overtook them. The initial fan base isn't really an issue.

It's also not the applications issue. NetBSD can pretty much run any app that's on Linux. There may be a bit lag (since the apps are developed on Linux most of the time and there's a bit of a porting effort), but the apps get there without too much time.

It's not the compile your own source code culture of the xBSDs since pkg_add supports binary packages, and Gentoo has more popularity than the xBSDs. There is also version of Debian for the xBSDs.

It's not even the kernel. A few years back, the BSD was superior in many ways, but Linux still outstripped it.

When all is said and done, there is only one key difference between Linux and BSD, the license. Companies like IBM don't mind GPLing their technology for the same reason TrollTech doesn't mind GPLing Qt....If anyone wants to use it in a commercial product, they have to pay IBM, TrollTech, Sleepycat, etc for the right to take the code prorietary. And although your competitors may have access to your source code, they can't do anything with it without releasing their changes so you can benefit from it. When a company GPLs their product, they haven't really given it away.

GPL is a quid pro quo license (you scratch my back, I'll scratch yours). Businesses understand quid pro quo and use it every day as a means of getting things done.

BSD is a charity license. As far as businesses are concerned, charity is good, but business is business and the last thing you want to do is give charity to your competitors.

It's not politically correct to say this, but "it really is the license, stupid".

BSD is a great example of what doesn't work

[jimfrost]

Actually BSD is the project I use to show exactly the opposite. While it's true there have been many individuals who have contributed to BSD, many major corporations have taken very significant code out of it and given back ... nothing.

Consider Sun Microsystems, whose SunOS operating system was based on BSD. What did they give back? Other than a few bug fixes early on, nothing.

Ultrix, from Digital Equipment, was BSD-based. Little to nothing came back to BSD from DEC.

Remember OSF/1, which was based on Mach/BSD? How much of their work went back? Next to nothing.

Microsoft used the BSD TCP stack as the basis of their TCP stack. What did they give back? Nothing.

FTP software based their whole product suite on the BSD codebase. How much came back? Nothing.

I don't know of any major corporation which has made significant donations back to the BSD core. There may be the rare exception, but the bulk of corporate back-donations has been some bug fixes. That has left the development almost entirely to individual developers or very small groups, and thereby limited how much could be done.

Lots of people think of the GPL as a "communist" license, but in fact it is BSD that is the free-for-all. The BSD license attaches no value to what it is licensing, and as a result you a software "tragedy of the commons" where everyone is happy to use it but almost nobody ever gives anything back. I know that there are going to be people who vehemently disagree with what I'm going to say, but: It has been my observation that the BSD source base has been relatively stagnant over more than a decade. If you look at what a modern BSD provides and compare it to what BSD 4.3 provided you'll find little that is new. A similar comparison with any major commercial UNIX will yield a great many such features (like working SMP support, journalled filesystems, NUMA support, logical volume management, realtime support, etc).

The GPL, on the other hand, leverages the fact that the source base is valuable. It is not a "give away" as so many people claim but rather an intellectual property trade very much like the patent sharing agreements so common in the proprietary world. While businesses would rather get something for nothing, if what they're getting in trade is valuable enough it is an incentive to give up some of their own rights.

If you think of the GPL as an intellectual property collective agreement you have the right idea. The thing about that kind of agreement is that the more IP that is covered by it the more valuable the collective becomes -- and therefore the more likely others are to join it.

In Linux' case the source base is exceptionally valuable at this point, worth literally billions of dollars, and for the better part of a decade has been receiving significant code donations from corporations. Remember the list of features modern UNIXen have that BSD doesn't? Did you notice how many of them Linux does support? All of them. For something like a decade corporations have been making major code donations back to the Linux codebase and it has advanced tremendously as a result. While Linux certainly has its rough edges it has seriously outgrown its tinkerer beginnings.

So Raymond could not be more wrong about this point. Oh, I agree that the development structure that Torvalds set up was a principal contributor to its success. To be sure, one of the major limitations in the BSD codebase has been the reluctance of the BSD principals to accept code they didn't write. But BSD has branched enough times that it has also seen conditions similar to what Linux enjoyed and it still never turned the corner.

What made Linux win was simply that large corporations had to give to get, and the more times that happens the more likely it becomes.

Re:BSD is a great example of what doesn't work

[greggman]

First of all your flat out wrong, lots of companies have contributed to FreeBSD.

Secondly that's only one example. There are plenty of other succesful BSDed open source projects

Apache
Python
PHP
Boost
Subversion
Lua
XFr ee
libpng
libtiff
libjpeg

the point is that BSD works just fine. you don't need to try to force people to contribute. In fact, the whole point of ESR argument is that most companies don't like the viral nature of GPL so they'd perfer not to contribute at all but given a BSDed project they can contribute directly to that project and not worry about having to give up everything else they use it with. Everybody wins.