Other BSD license and programming language does it matter

Perhaps the only thing I would avoid is proprietary / non-free tools like (Unity, Flash) or those backed by big companies (.NET, Java) because once the company goes kaput, the tools always die too. Even if they are "free", they are often too large for communities to adopt or the licenses too restrictive for another company to take forward.
I am not a fan of Java, although I do think that it should have its advantages, because all what Sun Microsystems supported was first class. Java is omnipresent, the demand is big since 20 years. Do
you really believe there is that risk?!

I am working on a new license based on BSD 3 Clause and BSD 2 Clause,
When one inspects a suftware, it is nice to read BSD, MIT, GNU, etc for not having to carefully read
all details. There are a lot of equivalent licences, many take into account laws in the U.S.A., but
each country has different laws. Do you really need there is a need for more "wordings" of the same
thing?

From all licences, I like MIT (wording) most: it is clearly an exchange, not a (unilateral) disclaimer.
 
You use BSD license, people can take it, add to it and place it under GPL. This DOES NOT affect your license, you will always own your code.
BUT
You can't take back those changes/enhancements made under GPL. It's a one-way street.
That's exactly the scenario that the CDDL tried to solve - The code is free and projects that use/import CDDL code can even be re-licensed, but CDDL-Parts still have to be under CDDL. That's what still prevents Oracle from taking OpenZFS code and including it in their proprietary ZFS. It also explicitly allows the co-existence of other licenses within the same project, which the GPL explicitly forbids (hence it's a viral license).
So if you want to make sure your code will always be freely available but can be used in other projects as long as your code stays free, the CDDL might also be a good fit. The CDDL even allows its modification as long as you rename it - so you could even remove/add clauses. IIRC there even was something like a "simplified CDDL" around that was more tailored towards small/private projects (i.e. removed all the business- and patent-related stuff).

This is quite interesting. Do you have a link to any specific license using that? Seeing the effects of GPL LibreOffice unrelentlessly sucking dry OpenOffice (and then splurging all their donations on trying to build a subscription cloud version), it would be quite cool to know ways to counteract this with the BSD license.
The story with OpenOffice is not the GPLs fault - OpenOffice (as a project led by Sun) was absorbed by Oracle, salvaged for what they needed and left out in the desert to die, just like any other software or project Oracle acquired from/with Sun Microsystems. Most developers left the OOo project during the invasion from Oracle and joined the LibreOffice fork. OpenOffice was already dead from the moment Oracle acquired it - it mostly is only still existent because of the GPL which prevented Oracle from fully merging it into one of their proprietary products and just closing the project.
 
The story with OpenOffice is not the GPLs fault - OpenOffice (as a project led by Sun) was absorbed by Oracle, salvaged for what they needed and left out in the desert to die, just like any other software or project Oracle acquired from/with Sun Microsystems. Most developers left the OOo project during the invasion from Oracle and joined the LibreOffice fork. OpenOffice was already dead from the moment Oracle acquired it - it mostly is only still existent because of the GPL which prevented Oracle from fully merging it into one of their proprietary products and just closing the project.
Are you sure? OpenOffice isn't GPL as far as I can see so if Oracle wanted to close it, they could. The fact that it isn't GPL is also what allows LibreOffice to benefit from the more permissive license, close the code up a little more (to the GPL) and not return anything back.

I am not a fan of Java, although I do think that it should have its advantages, because all what Sun Microsystems supported was first class. Java is omnipresent, the demand is big since 20 years. Do
you really believe there is that risk?!
Compared to .NET I think the risk is slightly smaller but just imagine we are now in the year 2099 and Oracle has stopped any involvement with Java for the last 10-15 years. If you suggest Java to i.e your line manager, I can't imagine them being thrilled. There is no support, Java is too large for you to maintain yourselves. I think java will fizzle out slowly rather than being destroyed purposefully. Of course you can still fit a few projects in between now and then ;)
 
Are you sure? OpenOffice isn't GPL as far as I can see so if Oracle wanted to close it, they could. The fact that it isn't GPL is also what allows LibreOffice to benefit from the more permissive license, close the code up a little more (to the GPL) and not return anything back.
According to wikipedia OpenOffice.org was released under SISSL and LGPL, from Version 2 onwards under GPLv3. LibreOffice of course had to keep the GPL on the forked codebase and all newer contributions are licensed under LGPLv3 and MPL 2.0.
 
Compared to .NET I think the risk is slightly smaller but just imagine we are now in the year 2099 and Oracle has stopped any involvement with Java for the last 10-15 years. If you suggest Java to i.e your line manager, I can't imagine them being thrilled. There is no support, Java is too large for you to maintain yourselves. I think java will fizzle out slowly rather than being destroyed purposefully. Of course you can still fit a few projects in between now and then ;)
Java is open source, and as been for some time now. My current employer had already switched to Openjdk, so I didn't even have to advocate for a change.

Sun was the source of so much technical innovation. Unfortunately, the business side of the house was not stellar. Java is a perfect example. It was free so they couldn't sell it, but it was not open source, which limited its adoption. Fortunately they did the right thing before the Oracle acquisition and we all got to benefit from the hundreds of millions of dollars Sun probably spent on the JDK.

But yes, Java is already fizzling out, not because it's a bad platform, but just because it's showing its age. Threaded programming proved to be too complicated for most programmers, yet every Java object has its own mutex. Many crappy libraries have been layered on top of the Java threading primitives, and yet asynchronous programming is still winning the day. The monolithic JVM with its kitchen sink of obsolete technologies is creating some drag too.
 
But yes, Java is already fizzling out, not because it's a bad platform, but just because it's showing its age. Threaded programming proved to be too complicated for most programmers, yet every Java object has its own mutex. Many crappy libraries have been layered on top of the Java threading primitives, and yet asynchronous programming is still winning the day. The monolithic JVM with its kitchen sink of obsolete technologies is creating some drag too.
Yep pretty much. Especially the monolithic JVM part. I think that alone is going to make it very difficult to keep pulling along into newer and newer environments. Especially if things like Linux move around (i.e I don't think Javax.swing will be ported to Wayland for a while) and Windows moves into entirely locked-down UWP stuff. Porting JVM to C++/cx will be pretty horrid!

Java could well be a great example of just being open-source is not quite enough to keep lifespan going. Software needs to be simple and agile too.

(That said, I am fairly confident we won't need to worry. Perhaps our grand kids will have to migrate away from Java for us ;)
 
I am working on a new license ...
How does that sound ?
Awful. IP law is very very complicated. I wouldn't let a regular lawyer write a new license, only a lawyer who is specialized in this field, and has experience. Are you an experienced IP lawyer who specializes in open source?
 
I was thinking of something like an Apache license, but with a clause that it can't be absorbed by more restrictive open-source licenses, but can be linked to them. GPL will likely reject this license, while it may be LGPL compatible. If this were to work, it would be ironic, because it would pivot LGPL dually under it.

Open source licenses have to be addressed separately from proprietary use which the above theory could work. Apache already in theory protects code from copyrighting, and I believe from patent trolling by companies wishing to close source or restrict.

If possible, a simplified license that covers relevent bases.


About Open Office's license as mentioned. Apache bought Open Office, and made it under the Apache license. The way it can get contributions, in spite of them being taken and not being given back by LibreOffice, is either another company finds a (financial) reason to get behind OpenOffice and contribute to it additionally under the Apache license. OpenOffice can get away from it by getting a license adjustment which makes it incompatible or unfavorable to GPL. CDDL was a good one that GPL didn't like.

all newer [Libre Office] contributions are licensed under LGPLv3 and MPL 2.0.
That's good. It would be nice if they can be convinced to release new code to be compatible so OpenOffice can absorb it, but perhaps this may be unlikely. Old LibreOffice parts obviously can't be changed.
 
Back
Top