That uncertainty plus the viral nature of the GPL means that there's no way to know for sure that you won't be forced to open your whole codebase at some point in the future. That's more risk than a proprietary software company is willing to take.
The only way to resolve this problem is to get legal decisions like this one. One way or the other, this decision will clarify the picture slightly. If we add enough other decisions, maybe someday we'll have clarity.
Until then, my company is solidly in the "Absolutely no GPL" camp.
No one can force you to open your whole code base. They can only make you choose between opening the parts that do the unsafe things listed above (linking, running in the same address space, utilizing private data structures via some kind of shim that is supposed to work around the GPL), or they can force you to stop distributing the GPLed software.
Until then, my company is solidly in the "Absolutely no
GPL" camp.
It's so stunningly easy to use GPLed software properly, and there are thousands of companies doing it. Just don't link it to your code, or try to do some kind of hack that avoids linking it while giving you all of the same benefits of doing so.The problem with VMware is that they have been pushing all of the boundaries; rather than just taking the easy road and doing known safe things like ensuring that there's a clear separation between GPLed code and their code, or just releasing their code that is linking to the GPLed code.
Take MediaWiki, it is GPL licensed (not AGPL), so you can modify it and run an instance of it without sharing your changes. Except, when someone accesses it, you are redistributing MediaWiki's GPL-licensed JavaScript files to their browser, so you do have to share your changes[0]. Think how many sites there are violating that without even realising.
Now think how you configure MediaWiki; you modify LocalSettings.php to (at the bare minimum) add the database username/password, and set the secret key.
This means essentially every site running a MediaWiki installation (or Wordpress, or effectively any other GPL-licensed web application) is violating the GPL, and are only getting away with it because no one has bothered to take legal action over it.
[0] http://lwn.net/Articles/609709/ (section: "JavaScript and the GPL")
However, once you distribute the JavaScript files:
"You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License [...] These requirements apply to the modified work as a whole."
By modifying LocalSettings.php you have created a derivation. By redistributing the JavaScript files you have distributed part of the work and therefore must make the whole derived work available under the GPL.
If the thing you want to use is a DLL, how do you use it without linking to it? Writing some sort of external process and using IPC to talk to it might fulfill the letter of the license, but it's more work and less robust, and still violates the spirit of the GPL.
In the end, we picked the MIT licensed component and worked around the issues.
That's not so hard, is it?
It's just like the question you get when using proprietary software. If you don't want to pay the license fee or abide by its terms and conditions, you find an alternative or write an alternative yourself.
The obligation to release your code is just the license fee for the GPL. If that's too steep for you, look somewhere else.
I'll quote LWN's summary [1] from the talk Bob Young of Red Hat gave at LCA this January:
He has long been a staunch advocate of the GPL, he said, but not because he has detailed opinions of the differences between it and other, similar licenses. Rather, he said, he learned early on that customers do not care about the details of free-software licenses, and that diving into an explanation of them was the fastest way to send a potential customer running toward the competition. ...
With the GPL, everyone knows what the rules are, more or less, so the debate is over quickly. They might not actually know the rules, he added, but they think they do, which amounts to the same thing. As a "business guy," he said, he is always trying to simplify the "pitch" - the GPL tells everyone "what we're about," so it has the greatest impact.
[0] We did this when we wanted to publish a royalty-free patent license tied to certain code, and wanted to be sure it would not keep people from using that code in GPLv3 projects. We were reasonably sure of our interpretation of the terms, but it's always nice to double-check.
Outside of a business that can hire lawyers, it would be helpful for the FSF and sites like http://opensource.org/licenses/BSD-3-Clause to have examples right beneath the licenses. Something like: "Here is a JS plugin with BSD-3 here is another plugin based on it. Note how the license did/didn't change." "Here is an MIT license and an example of a company adding that code to their app with license intact on the app's about screen". Stuff like that. http://choosealicense.com/ is helpful, and part of the way there.
They can't do that. The law doesn't work like that. The GPL FAQ[1] has to be deliberately vague about some things, because judges and juries themselves are vague, as are the laws they work to interpret. You can't have cut-and-dry examples of what can be done and what can't, because there can be many other externalities that could make a different case work in a different way, despite similarities to your proposed list of examples.
There are can be broad strokes, but the law is not a programming language. Recent xkcd comes to mind.[2]
Now, if individuals would be willing to work together to put up such a site, as a collection of information and opinions, with necessary disclaimers so that its limits are well-understood, and add examples, thoughts about them, critique and disagreements about them too, that would be different.
If Apple can afford to use the GPL, i.e. they think the GPL risk is low enough for them and won't make them free their entire codebase, then anyone can also make this assessment.
They seem to have no problem with GPLv2, but since they are using the last and aging GPLv2 releases of a bunch of GNU packages, they seem to be moving to non-copyleft alternatives purely because there are no more updates to the GPLv2 versions.
They had no apparent problem in adding git to Xcode despite their overall GPLv2+GPLv3 purge.
That sounds like a fun story. What happened?
Consider GNU Objective C. NeXT initially wanted to make this front
end proprietary; they proposed to release it as .o files, and let
users link them with the rest of GCC, thinking this might be a way
around the GPL's requirements. But our lawyer said that this would
not evade the requirements, that it was not allowed. And so they
made the Objective C front end free software.
https://www.gnu.org/philosophy/pragmatic.html Specifically, as I've spoken about in my many talks on GPL
compliance, the earliest publicly discussed major GPL violation
was by NeXT computing when Steve Jobs attempted and failed (thanks
to RMS' GPL enforcement work) to make the Objective C front-end to
GCC proprietary. Everything for everyone involved would have gone
quite differently if that enforcement effort had failed.
http://ebb.org/bkuhn/blog/2014/01/26/llvm.htmlThat doesn't happen. A violator would be ordered to pay compensation and remove the offending code, though they could reach a settlement to do something else.
Simply providing the source of libhoge is not sufficient, and most people do not understand that about the GNU LGPL.
Yes, you have to provide the source and a build process for the (L)GPL'd software in question. That's standard for both the GPL and LGPL; no difference. From there, linking is just another step in the build with ld, or at runtime with LD_PRELOAD.
The LGPL is incredibly under-appreciated. Developers love to complain about GPL'd software, but who actually takes raw chunks of code, or entire applications, and re-packages it as their own? Most developers don't have a reason to do this -- the majority of our work with third-party code consists of interfacing with libraries and that is where the LGPL comes in handy.
If you're uncomfortable committing completely to the GPL but still want to support its basic tenets, try the LGPL.
I personally recommend against using the GNU LGPL because of this confusion. If you want something that does what most people think the GNU LGPL does, use the MPL with a GNU GPL exception. (IMO, the MPL is a superior licence to anything in the GNU GPL family, but it is also not my licence of choice—that is the MIT-style licence.)
The language of most of common software licenses is straightforward; blindly using any license without being aware with their implications, especially in a commercial capacity, is... stupid.
Beyond that, we'll have to agree to disagree. If you're familiar with the four freedoms of the GNU free software definition [1], then licenses that are nonfree and non-copyleft do not ensure the four freedoms for users of the software. The (L)GPL has been carefully crafted to ensure those those freedoms. Although that might not matter to you personally, preserving those freedoms is important for many people.
The MPL is a superb licence that is under-appreciated. It usually acts similar to the MIT licence in that it never taints your software that uses MPLed software. It has, at best, a “modular” taint. If you depend on an MPLed library, you can combine that library without revealing your source at all. If you modify that library, you only have to release your modified source. You can even modify it so that it goes from proprietary -> MPLed library -> proprietary, but you have to document the nature of what you’re doing in that call out from the MPL and you may need to release that (I haven’t dealt with the MPL in years, but I think that describes that particular case).
GNU activists were loud enough for years (despite having had the exact same debate at the beginning of the licensing project) that ultimately Mozilla caved and dual-licensed its projects just to shut them up. (It didn’t help with some folks in the Debian community because of the constraints on the trademarks, but some people will never be satisfied unless you’re receiving your web pages by email in emacs.)
You may want to know that in the past few years, MPL 2.0 was released, which embeds a so-called "GPL compatibility" by default. In fact, it's a relicensing allowed by MPL 2.0 licensor.
MPL 2.0 still keeps an option for the licensor's software to never be relicensed, but they have to specifically use an "incompatibility" clause.
Ignoring the structure of actually related areas of law, as ditching intellectual property as a category would be doing, is great if you want to deceive people, but very bad if your intent is to educate and help people understand the law.
IP is a subcategory of intangible personal property, which is a subcategory of personal property, which is a subcategory of property. There are legal principles that are generally applicable to property, others that apply to personal (as opposed to real) property, others that apply to intangible (as opposed to tangible) personal property, etc.
IP isn't particularly closely related to "physical property" (which isn't a particularly useful category); nor does the term suggest that it is. Real property is in a separate branch from IP at the highest level of categorization of property rights, and tangible personal property is at the next level down. It is more closely related to things like debts and choses in action (rights to sue), which are, while not IP, also intangible personal property.
And there's no need to make it an us-vs-them issue. Nobody is defining terms to fit their agenda, criticism of IP is simply pointing to differences that are already codified in law. So if you don't like it, take it up with your legislators. You don't have to be either for or against copyleft to want say the big scary copy-r-word to your developers. Even the most ardent copyleft haters and copyright abolitionism idealists are shooting themselves in the foot when they make snide remarks toward lawyers about their "legalese", before they know it reality will catch up with them and they'll be hit with overly broad patent suits or some jerk will infringe on their trademarks.
In one way, this seem to be suggested. How can we expect employee of software development companies to understand software licensing, copyright, or that downloading stuff from the internet doesn't automatically make it free-for-fall. On the other hand, the software industry seems to be very aware of piracy, that downloading stuff from the pirate bay is copyright infringement, and that its not free just because its on the internet.
I am thus torn to the suggestion that this is a education problem.
I like to believe that corporations would generally agree to the above, though some employees may accidentally violate open source licenses or other forms of IP without realizing (and thereby potentially putting the company's business or own IP at risk). Making employees aware of the licensing issues would help.
Interestingly though also, people go through such corporate IP training treating it as bureaucracy. Elsewhere too, I see people copying things from the Internet (e.g., images, music) without adhering to copyright laws. People click on "I agree" without reading, sign employment agreements without having slightest idea of what they says, and so on. I have myself tried to educate people about these several times on Hacker News.
Not to defend VMWare here by any means, fixing/mitigating the bigger problem would require increasing general awareness of IP issues at forums like HN and elsewhere, such that the corporations can sanely honor the IP licensing without high costs of employee training and maintaining a sizeable legal department.
It also seems to me that lawyers at big corporations can also help by making the legal processes easier. They are often so risk averse (to save their own back, not necessarily for the benefit of or on behalf of the organization) that they won't approve even reasonable IP requests sent their way.