Menu
▾
▴
Posted by
2013-03-14
MIT license
After thinking much of it I decided to release BifurcumLib and Bifurcum under terms and conditions of the MIT license. It turned out for me that unmindful using the GPL just for everything could become more dangerous for our freedom than it defends our freedom.
Just one issue: If you don't apply the latest version of the GPL you can not use code released under the latest version of the GPL. For example, you release under the terms of the GPL version 2 since 2006, as everyone did in 2006. Now (2013) more and more projects put their sources under the GPL version 3. If you stay with GPL version 2 you can not use the sources from all the GPL version 3 projects around.
Even worser, If you did apply the "or any later version, at your option" clause you going to risk that your project is forked and put under GPL version 3. You may think: So what just use the GPL version 3 instead. But what if I think the GPL version 3, 4, 5 or whatever is a mistake? Or just can not change? One option would be to remove the "or any later version, at your option" clause ... a "nice" fracture in the community.
One could even just write "GPL" and no version number in her/his copyright notice. Very generous but that means that all versions of GPL are applicable - even the GPL version 1. Forget about using GPLv2 or GPLv3 code so long as this notice remain...
So it looks like (<=> share, --> exploitation, # a fence)
Project A: GPLv2+ <=> Project B: GPLv2+ // A and B share code
Project A: GPLv2+ <-- Project B: GPLv3+ // B upgrades to newer GPL version, now exploiting A
Project A: GPLv2 # Project B: GPLv3+ // A prevents from being forked and stops exploitation
I'm convinced that the FSF is neither ignorant nor evil but I think they have underestimated how popular Free Software became. It is very unlikely that all the 100,000+ Open-Source-programmers worldwide agree on the latest version to GPL. We are just to much different people (and often people with strong opinions too).
As I understood it, strong copyleft (and the GPL as an "implementation" of it) was invented to build a free operation system including a free developer toolchain. It fails its own goals when applied to every small piece of userland software because it may prevent people from working together, especially people with different cultural background, political opinions or economic background but causes fractures in the community.
I had my lessons in history and learned about so many movements which failed because they split and some being successful because bringing very different people together by fighting for only few goals which just makes sense.
By the way, a simple license like the MIT license still protects freedom in an amazing way: Every software-manufakturer in these days must have an Open-Source-management. Either he must make sure that no Free Software makes it into his products or he must ensure that the copyrights of Open-Source-programmers are respected. However he must educate or employ someone how knows what Free Software is. (I'm mixing both terms, Free Software and Open Source, with intend.) If the manufacurer says Yes! to Open Source he can not turn around and dub us "cancer"! Moreover, the problems of Free Software are also his problems i.e. software patent threat.
Appropos software patent threat: I toggled the MIT license over BSD license after reading this article:
http://www.scribd.com/doc/46088081/Closing-the-Loophole-Open-Source-Licensing-amp-the-Implied-Patent-License-Nadan
and this FAQ my be helpful too (is about Boost software license):
http://www.boost.org/users/license.html
About the advantages of simple licenses: http://www.copyfree.org
Edit 1: I want to clarify that there is nothing wrong with any version of the GPL as such in my opinion. I just think the idea behind is going to be overstretched. Beware that the opposite of wrong isn't automatically right: Strong copyleft / the GPL is however great and important.
What I would like to see is a living network of people who share ideas, code, work; not so much by terms of sophisticated licenses rather than a "gift economy" by spending and receiving. The most amazing things arise from different people working together and/or from bringing different things together. Think about bio-chemistry and others.
If we are convinced that Free Software is superior to non-free software - what are we afraid?! Victory is ours! Even with MIT license. The author saw enough closed source software to tell this, really.
Edit 2: Just wrote second sentence, "It turned out for me...", more precisely. Sorry for my limited English.
Edit 3:
Other people think far more ahead: The H: Why it's time to stop using open source licences - this article describes the advantages of releasing software in public domain. Thinking on SQLite - he may right.
Edit 4: Did you realize that shifting to MIT license enables this code to be used under all versions of the GPL, including future versions? (It was GPLv3-only before).
Same happened to my project JDecisiontable.