Survey Says GPLv3 Is Shunned

willdavid writes in to note a survey of open source developers conducted by Evans Data that indicates a real rift in the community over GPLv3. The survey was based on in-depth interviews with 380 open source developers and no estimated margin of error was given. "Just 6 percent of developers working with open-source software have adopted the new GNU General Public License version 3... Also, two-thirds say they will not adopt GPLv3 anytime in the next year, and 43 percent say they will never implement the new license. Almost twice as many would be less likely to join a project that uses GPLv3 than would be likely to join... [Evans Data's CEO said] 'Developers are confused and divided about [the restrictions GPLv3 imposes], with fairly equal numbers agreeing with the restrictions, disagreeing with them, or thinking they will be unenforceable.'"

Slow adoption is to be expected

[A beautiful mind (821714)]

It is much easier for new projects to start out with GPLv3 than old projects to convert. Unless the committers transfer copyrights to a central body like in the case of the gnu tools and FSF, it is hard to move to another license if not bordering on impossible.

Re:Slow adoption is to be expected

[stinerman (812158)]

Not to mention any project with files licensed under GPLv2 or later is, for all intents and purposes, GPLv3 anyway.

Re:Slow adoption is to be expected

[fotbr (855184)]

Not necessarily. GPL version 3 only provisions do not apply to it, unless it is changed to be licensed under GPL 3 (only, or "or later").

Re:

[swillden (191260)]

Not necessarily. GPL version 3 only provisions do not apply to it, unless it is changed to be licensed under GPL 3 (only, or "or later").

One key clarification here: You can't change the license on a piece of code unless the license gives you permission to sub-license. Otherwise, only the copyright owner can change the license on a file from, say, GPLv2+ to GPLv3+. Someone else can come along and add some GPLv3+ code to the GPLv2+ file, and the result will only be distributable under GPLv3+ but this doesn't mean the original GPLv2+ code has been relicensed. If the GPLv3+ code is removed, then the result will again be distributable under

Re:

[fotbr (855184)]

They SHOULD all be counted as GPL2, because until they are explicitly moved to GPL3, they are not GPL3.

Re:Slow adoption is to be expected

[DragonWriter (970822)]

My point was that all these projects can be counted as GPLv3 projects, or is it that important that I formally fork such a project to be counted in the numbers?

Well, yes, if it is not currently licensed with the restrictions in the GPLv3 (but merely allows other people to relicense their own redistribution that way), it is inaccurate to describe it as a GPLv3 project. I mean, by your argument, every project under a GPLv3-compatible license (or, presumably, in the public domain like SQLite) should be counted as a GPLv3 project because someone could conceivably redistribute a derivative of it under the GPLv3 at some point in the future.

Re:

[dosius (230542)]

Important, yes, important enough that I insist on staying with 2, or using a more liberal license.

-uso.

Re:

[JesseMcDonald (536341)]

Unless the committers transfer copyrights to a central body like in the case of the gnu tools and FSF, it is hard to move to another license if not bordering on impossible.

Unless, of course, all the commits were "GPLv2 or later", in which case the project was effectively already under the GPLv3 from the moment it was released.

Re:Slow adoption is to be expected

[hansonc (127888)]

wrong.

If I'm the user of the code e.g. Tivo and I don't decide that I want to comply with gpl v3 I don't have to in that case. For you to force me to comply with v3 you have to relicense it as v3 (or later) it's not a retroactive license which probably wouldn't be legally enforceable anyway.

Not exactly

[paladinwannabe2 (889776)]

Something licensed that way can be used by both GPLv2 and GPLv3 projects, but can't use GPL3 code itself without converting to GPL3. It's still under GPLv2 until then.

Re:

[sumdumass (711423)]

What? How can you claim something falls under a license when I can ignore a good portion of the limits the license imposes?

No, Section 9 says you have the option of following either version. You would have to change the license in order to have it count as GPLv3. Otherwise there would be nothing stopping me from ignoring your GPLv3 restrictions and just using the GPLv2. The GPLv2 says that derivative works much use that license (GPLv2). It also says no further restrictions so I'm not sure if you can even us

Re:

[huckamania (533052)]

I think this attitude is why so many developers are turned off by the GPLv3.

Is there a clause in the GPLv3 that makes the "or later" mandatory? If that's the case, might as well sign it all over to the FSF or better yet just put "This software is released in whatever manner RMS decides at any time now or in the future".

Still, I wonder about the legality of enforcing a license that doesn't exist or didn't exist when you first got the source. "This software is released under a future license which we will

Re:Slow adoption is to be expected

[Experiment 626 (698257)]

Unless, of course, all the commits were "GPLv2 or later", in which case the project was effectively already under the GPLv3 from the moment it was released.

Wrong. The "or later" does not mean that whatever the most recent version of the GPL has been published is the one that applies. It means someone wanting to copy / distribute / whatever the software is free to do so under the terms of the GPLv2, or any later version that they might prefer the terms of. If the GPLv4 came out next week and said "to distribute software under this license, you have to send RMS a case of beer", you could distribute "GPLv2 or later" software by either providing its source (the GPLv2/v3 option) or by sending RMS a case of beer. New versions of the GPL give you more choices in licensing "or later" code, they don't retroactively change the terms of the deal like some shady EULA.

Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

Re:

[einhverfr (238914)]

The problem is far more subtle than most realize.

The basic issue has to do with whether or not the BSD code can be "relicensed" (in RMS and Eben Moglen's words) as GPL code without making any changes. While it is clear that copyright-worthy changes can be under any license, the question is whether the original code licensed under the BSD license can be. IMHO (IANAL) this talk of "relicensing" seems like legal mumbo-jumbo devoid of any accepted meaning.

The problem is that the GPL 3 is only compatible with

Re:

[donaldm (919619)]

Actually the Article is not that clear who the 6% of developers that are adopting the GPL3. If they are Redhat and IBM and they rather like the GPL3 http://www.news.com/8301-10784_3-6171921-7.html [news.com] then you have nearly 40% of Linux development. Throw in HP and possibly Novell then you can add another 5% to 10% more. I can't see SUN going down the GPL3 path but then you never know. Can anyone shed any light on this because the article does not really say that much although I did find the source of the data at

Confused...

[advocate_one (662832)]

who actually commissioned the study??? cos studies don't just happen by themselves... they cost money

Re:

[BibelBiber (557179)]

Really? What about student projects? They do studies all the time and they even have to pay for it (so they get some feedback).

Re:Confused... M$? (or not)

[joe 155 (937621)]

couldn't say who actually coughed up the money for this one, but they do list M$ as clients. We all know M$ aren't above what we might (generously) call "interesting" techniques when it comes to dealing with the GPL (not least, IIRC, calling it a "cancer"). Evans also list some (what I would call) nicer companies though - especially from the open source POV - including but not limited to RedHat and Sun. You can check out the full list here;

http://www.evansdata.com/company/clients.php [evansdata.com]

Re:

[Fred_A (10934)]

If you want more specific than this; write to them:

740 Front Street, Suite 240
Santa Cruz, CA 95060

Wait; you means Evans corp is a Santa Cruz Operation ??

Fork

[BibelBiber (557179)]

To quote an earlier article: "If you don't like it fork it."

Re:

[VGPowerlord (621254)]

Also, don't call the Operator unless you're suicidal.

Does Simon still work there?

Remember!

[heinousjay (683506)]

Those restrictions are for your freedom. It is important to take freedom away to protect it. Truly allowing freedom would allow freedom to be taken away, and we can't allow that, so we've taken away some freedom to allow true freedom to flourish.

I can't imagine why anyone wouldn't understand that perfectly.

And I'm sure I'll get modded down, but before you do that, read through my first paragraph carefully and tell me what I've said differently than the GNU people.

Re:Remember!

[cromar (1103585)]

GPL protects the freedom of the code, not the freedom of the developer. Big difference! If you want developer freedom, use the BSD license or some such. Different tools for different problems :)

Re:Remember!

[Lost+Found (844289)]

What the GPL is really concerned with beyond the code is protecting the freedom of the code's user. BSD aims to give the initial recipient of BSD-licensed code the freedom to make copies and changes in virtually any way they want whereas GPL aims to give those same freedoms and enforce them in second and third and fourth order copies, etc.

Re:

[synthespian (563437)]

Freedom is a category that relates to people.
Freedom can never relate to an inanimate object, such as code.

Re:

[Rycross (836649)]

I don't think the characterization is fair. You are talking about different freedoms, and the trade offs between them. Not all freedoms are equal, and not all freedoms are good. If you'll excuse the hyperbole, in an anarchy, you have the freedom to murder and steal, but I don't see many people clamoring for those freedoms in our society. I'd imagine the FSF has made a decision about which freedoms are more important. "Software freedom" isn't quite as clear cut a decision as my example, and is largely d

Re:Remember!

[Bogtha (906264)]

I'm not sure if you are being sarcastic or not, but if you are, then please be aware that taking freedoms away to protect other freedoms is the basis of all law. You aren't free to hit me because you've had that freedom taken away from you.

Unless you are an anarchist, you really have no basis for criticising the GPL in this regard, because you agree with this logic applied to different areas.

Re:Remember!

[ScrewMaster (602015)]

Regardless, if the GP hits me I'm damn well going to hit him back.

Re:Remember!

[maxwell demon (590494)]

It doesn't restrict hardware. You can make your hardware DRMed like hell. You just can't run GPLv3ed software on that DRMed hardware. That's a restriction on the software (don't run it on DRMed hardware).

Re:Remember!

[Ginger Unicorn (952287)]

it also doesn't restrict how you *run* it. Run it on DRMed hardware to your heart's content. It just specifies what information you must include should you wish to redistribute it.

Re:

[A beautiful mind (821714)]

Ultimately, the GPL has begun the process that all legalese follows - it is now becoming too complicated to understand without paying a lawyer a very large sum of money. Given the target audience and the goals, this is not a good thing.

Your beef is with the legal system, specifically legalese. It is the way to enter into official, legally binding agreements. Without the GPL using legalese, it would be worthless for it's goals. If you want to know what the GPL is about, be sure to read the documents on FSF'

Re:

[A beautiful mind (821714)]

Actually, the GPL was always good at using the opponent's tools against them. It is after all using copyright to achieve it's goal. Without copyright, there would be no need for GPL, as we'd have a pretty level playing field, as all code would be practically BSD licensed. In such a world the advantages GPL provides for free software authors wouldn't be necessary anymore.

Oh dear!

[mmcuh (1088773)]

Oh dear! Another rift in the community, etc. Really, how many articles of this type have been posted to Slashdot in the last few weeks?

And the statement "Just 6 percent of developers working with open-source software have adopted the new GNU General Public License version 3" is obviously false, since the vast majority of GPL-licensed software have copyright notices that say that the software is available "under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version" - which includes GPL version 3.

What is this "Evans Data Corporation"? It would be interesting to see any other press releases they have written.

Re:

[cromar (1103585)]

I am wondering who they are, as well. If it helps, they also did this study [zend.com] about PHP.

Ooh, and here's one [softpedia.com] where they say Linux is gaining ground over Windows in certain areas...

Re:

[RightSaidFred99 (874576)]

It's not obviously false. The GPLv3 is a more restrictive license, the copyright holder, if he wishes to enforce this higher level of requirements for distribution, must distribute his code under it. It's rather silly to say that since the person wishing to make use of the license can choose a more restrictive (Gplv3) license - what's the point?

The whole point of Gplv3 is that the FSF and several assorted righteous nutjobs thought they would "get one over" on Microsoft. They didn't, but Groklaw and oth

I'm confused...

[packetmon (977047)]

'Developers are confused and divided about [the restrictions GPLv3 imposes], with fairly equal numbers agreeing with the restrictions, disagreeing with them, or thinking they will be unenforceable The conundrum... If a packet falls in an OSPF forest of Spanning Tress does it send an ICMP-Unreachable

Maybe for Now...

[cromar (1103585)]

Anyone who pays attention to the discussions on ./ would know there is division of attitudes toward GPLv3. I'm fairly confident it will gain more support, what with the FSF behind it, and as new projects come out.

By the way, anyone have more information about just who exactly the Evans Data Corporation is, and whether they are a respectable source of research? (I noticed this is a press release and not an independent article...)

No Margin of Error

[DragonWriter (970822)]

The survey was based on in-depth interviews with 380 open source developers and no estimated margin of error was given.

No doubt because it wasn't a random sample in the first place, so a "margin of error", which reflects the sampling error, would be meaningless.

Fud-mongering

[Vainglorious Coward (267452)]

Only 6% of developers...have adopted GPLv3...Two-thirds say they will not be adopting GPLv3 anytime in the next year, and 43% say they will never implement the new license

Interesting to compare this "shunning" with Vista [slashdot.org] :

less than 2% of UK-based firms have already upgraded all their desktops to Windows Vista. Just shy of 5% said that they have begun a Windows Vista desktop upgrade program. 6.5% said they will upgrade in the next 6 months; 12.6% in the next 12 months; 13% in the next 18 months; and 18% in the next two years

Summary : GPLv3 is more popular than Vista

Vista?

[athloi (1075845)]

You mean the GPL3 is the Microsoft Vista of the open source licensing world?

Just wait

[Chris Snook (872473)]

Once adoption of version 3 slows, they'll just release version 3.5 and we'll all have to buy new books anyway.

Re:

[Dr. Photo (640363)]

I attempt to disbelieve!

Re:

[TheWanderingHermit (513872)]

By 2009? I think a lot of people have strong feelings about not using it.

It's now the Vista of Open Source Licensing.

Re:Maybe I want my code to be used Commercially!

[Grishnakh (216268)]

The GPL isn't about preventing commercial use. If you bothered to read the license(s) you would know that.

It's about preserving users' freedoms. If a commercial entity uses GPL code and distributes that to end users (even paying ones), they're obligated to give them access to the source code. It's that simple. GPLv3 just adds some extra clauses to prevent companies from weaseling around the spirit of these simple terms by using any software patents or the like.

If you don't care about commercial entities taking your code, making changes, distributing it to users, and then refusing to give those users (which may include you) access to their modified code, then release your code under the BSD license, or into the public domain. It's your choice. Stop complaining about other peoples' choices.

Re:

[Em Adespoton (792954)]

GPLv1 and v2 look at free software being sidetracked or misused by private interests who want to subvert it to their own goals. What GPLv3 has done is tackled the opposite problem... private interests attempting to subvert GPL'd code by contributing code to a GPL project which has other non-copyright restrictions attached to it (eg, patents). The idea here is that someone can contribute a key piece of code and have a submarine patent on the algorithm. The entire project becomes popular, things are modifi

Re:

[huckamania (533052)]

It's more likely that a patent infringement would be added unwittingly by a third (or fourth) party. The GPLv3 does what it can, but it can't magically give immunity from patent infringement. Well, except for the case you outlined which is probably the least likely to occur.

Still, SAMBA!

Re:

[Grishnakh (216268)]

The GPL is all about preventing commercial use. If you bothered to think about the license(s) after reading, you would realize that.

You're misinterpreting the words "preventing commercial use".

What you really mean is "steal someone else's code, make a few modifications, and sell it as your own." That's basically stealing, because you've added no real value.

What I mean by "commercial use" is "a company is allowed to use GPL code and incorporate it into their own product, as long as they release the changes

Re:

[phantomlord (38815)]

After all, how many projects still use GPL version 1?

GNU GENERAL PUBLIC LICENSE
Version 1, February 1989

GNU GENERAL PUBLIC LICENSE
Version 2, June 1991

GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007

Mind to venture a guess how many non-FSF/GNU projects were created in the 28 months that the GPLv1 was the current license versus how many projects were created in the 16 years that the GPLv2 was the current license? Predicting the dominance of the GPLv3 based on the current usage of GPLv1 is a little disingenuous.

As for the GPLv3 being the dominate license in a few years, I've read that RMS already wants to get a GPLv4 out soon. If it's within the next 3 years, the GPLv3 will barely have time to

There's a difference

[GroundBounce (20126)]

The change from GPL1 to GPL2 was more of a no-brainer. Even Linus adopted GPL2, and he's pretty much on the "business-friendly" side of the spectrum.

The thing that's new about GPL3, is that it tries to not only keep the code itself open and free, which I believe is a valid goal of a software license, but it tries to control *other* behaviors of an organization that are more marginally related to the code itself, such as patent cross-licensing agreements, etc. If a piece of software does not violate any kn

No.

[keeboo (724305)]

The copyright holder may license each different version as he pleases, he may even make a parallel commercial release.
The one able to violate such license is the one who receives the code already under such license.