Slashdot Mirror
Apache License Updated to 2.0
Roy_Fielding writes "The Apache Software Foundation has approved an update
to the open source
Apache License (Version 2.0) that will be mandated across all Apache projects
starting on March 1st.
I have been working on variations of this
license for the past three years, trying to
balance the many different goals of the revision. That includes making the license easier for
non-ASF projects to use, improving compatibility
with GPL-based software, allowing the license
to be included by reference instead of listed
in every file, clarifying the license on contributions, and requiring a patent license on contributions that necessarily
infringe a contributor's own patents.
The result is a license that is compatible with other open source
licenses, such as the GPL, and yet still remains true to the original
goals of the Apache Group and supportive of collaborative development
across both nonprofit and commercial organizations." While it has yet to become OSI-certified, it will probably will be so Real Soon Now, and in the meantime it's fun to compare licenses.
If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed.
I'm no lawyer, but this seems to be saying that if you sue anyone for breach of patent for something in apache, then you lose your patanet license?
Can someone explain this onein plain english, please?
"How fine you look when dressed in rage."
Can someone explain to me some specific problems there were with the old license? Not having it in every file I am sure makes things a bit easier, but what else does this do exactly? The original poster didn't make this too clear, and I don't exactly go around studing licenses unless I know it conflicts with something for me personally.
in the meantime it's fun to compare licenses.
I don't know about CowboyNeal, but I sure have a different definition of fun than he does!
Doh!
After something of the same journey with licenses for my own open source work, I finally came to the conclusion that Richard Stallman had seen the inevitable truth clearly when he designed the GPL, namely that free software thrives best when there is a definite barrier between it and commercial software.
Our software now uses a dual license model in which it's either licensed for free under the GPL, or licensed for a fee under a standard commercial license.
Without exageration, nor wanting to start a religious war, I believe the GPL is an astonishingly robust answer to the question of how to share creative works without subsidising commercial interests that inevitably seek to quash the independent creative spirit.
Ceci n'est pas une signature
The Apache License 2 is just a workaround for a real problem. Software patents are bad for development and bad for the economy. The US Federal State Commission called for change in a recent report, the benefits of software patents are falsified by emirical ressearch.
However, as software patents serve for the benefit of patent attorneys the institutions are intrested in an extension of the system, by widening the scope of patentability regardless of an economic foundation. Politicians like this quantitative patent approach, the result are many trivial patents of low quality and disfunction of the patent system atlarge. Many low quality patents endanger our information society. So it is nice to see that organisation like Foundation for a Free Information Infrastructure build a counter-force to patent lawyer interest groups in Europe. So far the lobby work against software patents and the Eurolinux petition were very succesful. In my opinion we need a world wide movement in order to avoid Eolas vs. MS oder Amazonvs.Barnes&Noble ecc. will happen again. The GNU Public License is incompatible with Patent law and most projects and SME cannot afford to get patents. They, the innovators, don't want or need software patents.
I am going to guess that it is because apache is one of the MAJOR reasons linux ever took off in any commercial setting in the first place. Preconfigured apache on linux boxes that were extremely cheap made very attractive web servers to many people and businesses. This was a major boost to linux awareness and usage. I think we all know how popular apache is as a webserver, and I think it is fair to say that the majority of apache servers out there are running on a *nix platform. Without apache, these platforms may not have ever got much exposure.
I couldn't find the Apache 2.0 license on the FSF license list. The only "official" remark seems to be the following email thread which says:
Maybe it was a bit premature to announce this license without waiting for OSI approval and requesting feedback from the FSF. Of course the Apache group can do whatever they want without asking for approval and blessing from other Open Source and Free Software groups. But it would have been nice to try to cooperate a bit more.
They tried "improving compatibility with GPL-based software", but is it compatible or not?
From a quick read it seems to be a valid Free Software license, but clause 4.d may make it incompatible with the GPL. This would be unfortunate for such a trivial condition. GPL doesn't allow placing additional restrictions on distribution, so is requiring a NOTICE file, and additional restriction?
The patents section might also be GPL-incompatible, but it might be GPLv3-compatible when GPLv3 comes out.
Has anyone seen of any comments from FSF about this?
Altogether, it's a good license, and vastly superior to the last proposal which was ~100 pages long. (slight exageration)
Expert in software patents or patent law? Contribute to the ESP wiki!
Is going to lawschool a requirement for beeing a geek nowadays?
Ciryon
Does the Apache license allow commercial distribution of code based on the apache code, in the case where the modified source code is included as part of the commercial package, but the end-user is prohibited from redistributing the source code?
I thought it was clear but I'll explain again.
My company writes software that we like to distribute as free software. I started doing this in 1995, and the tools and packages I make are quite widely ported and used. When I started my company our policy was that anything we could not rapidly turn into a product we would release as open source. Software that is not used dies, we figured.
Our license was a BSD-style license that basically allowed people to do what they liked with the software.
In 2001 I noticed that some commercial products incorporating our code were being sold. Very good, I thought, it's nice to see our work being used. But when I asked them to provide us with some of the extensions and patches they'd made, the answer was "no, this is not possible". Now, seeing people use the results of years of work then refusing to contribute anything back rather annoyed me. My company was selling support licenses for our products, and these were in fact our competitors.
The solution came in the form of an article by Richard Stallman which explained why using the LGPL was in fact giving help to closed-source developers who directly or indirectly compete with open source developers.
We decided to switch to the GPL, and in 2002 we moved all our OSS products to this license. At the same time, we had a number of commercial licensees. To give these groups a non-viral package, we developed a dual licensing model.
Since the code is ours, it's our right to license it to specific users under specific terms. The GPL is not incompatible with commercial licenses, so long as it's the copyright holder who decides what license to apply in each case.
To summarize: for OSS developers we use the GPL, for commercial developers we use a commercial license.
It works well. We've had no GPL violations, and enough commercial licensees to make it worth developing our core packages further.
Ceci n'est pas une signature
What is unusual about this license is that it grants a license to the copyright. Most licenses only grant a right contained and protected by copyright law. For example, a company may grant a license to distribute a good, or to reproduce elements of an original work of authorship. I have never heard of being able to license a copyright. We assign and transfer copyrights and we license trademarks and rights granted under tm, (c), and or patent. I wonder if this agreement is really valid. Lets say that someone modifies some code and then registers the copyright and does not inform the Library of Congress about the Apache license. Moreover, what if the "thief" then files suit against derivative authors. The problem here is that fed law trumps the law of contracts and under federal law you can share copyright in a work- it's called joint ownership. One suggestion is that Apache could grant a nonexclusive right to create derivative works so long as they meet the highest standards in the industry. It would be interesting to see if Apache filed the copyright through the LOC? It might be interesting to look into how foreign civil law treats software because it is more natural law/moral rights based and does not allow corporate entities to own the full ip rights of a creator.
Jax
IANAL but you sure sound like one :)
Of those to whom much is given, much is required.
The GPL requires to share the source code for modifications and also forces any software linked to GPL code to be licensed under the GPL as well. The Apache license does neither. From my POV the Apache License is more free then the GPL while the GPL tries to keep free software free. IMHO a good compromise is the LGPL which protected the freedom of your work but does not require others to open their source when linking to your stuff.
Software (et al.) patents are essentially:
- legalized monopolies, and monopolies are bad.
- legal minefields for software initiatives, and mines are extremely counterproductive in (real and analogous) terms.
- an involuntary subsidy by the software industry for the self-serving legal industry.
- an impediment to open (and commercial) software development.
- an idea based solution to an implementation driven market (ideas are cheap; implementation is hard).
- a backwards implementation benefitting manipulators of the system, not the brilliance in innovation.
- an upside down system benefitting those with big pockets (and fat lawyers) instead of the underdogs (for whom patents were very originally intended).
- a joke in any industry (e.g., software) where production is a mere disk copy away (i.e., there is no manufacturing impediment in software, so even the original concept of patent protection is absurd)
- completely absurd to implement in theory (due to impossibility of any patent examiner's job) and practice considering the infinitude of like ideas in different forms (with patents as with novels, plots, etc.)
- absurd since software is a written form of thought --- and you shouldn't be able to patent (anything but especially) thoughts.
That's just off the top of my head.
IOW, patents are a Lehman (prounounced "lemon") law perpetuated so the legal industry can forcefeed carcinogenic lemonade to the software industry.
This might be a tad offtopic but anyway, a while ago I was wondering how a big OSS project (like Apache) does a license change? If there are dozens of people involved who have contributed their work to the project under license X, wouldn't you need to ask each and every one of those people if you want to change that license in any way?