Chiken
Don't Let Your Walls Down
this post has been archived.
so let's talk about bsdl versus gpl.
if you're debating which license to open source (OS) your software under, it seems to me that there are two potential effects of choosing gpl:
a. if your code is gpl, other people who use your code will have to make their software free. so in this sense, gpl has more social benefits than bsdl. if you just used bsdl, other people who use your code may not OS their software as well.
b. if your code is gpl, less people overall will use your code, because they cannot or simply do not want their software to be OS. in this sense, gpl has less social benefits than bsdl. if you just used bsdl, other people could use your code regardless and benefit from it.
so which is the bigger effect? the effect that the gpl has on making others OS their software? or the effect that the gpl has on making others not want to use the code?
i think an interesting economics cost-benefit analysis (
http://en.wikipedia.org/wiki/Cost-benefit_analysis ) could be done on this.
this post has been archived.
Chiken
Don't Let Your Walls Down
this post has been archived.
Chiken
Don't Let Your Walls Down
this post has been archived.
I think the licensing decision is very situation-specific. phoenix made a good post about this on BSDnexus a while back:
This is why the BSD license (or similar) should be used for architecture-level code (protocol stacks, spell checkers, file formats, etc) stuff that follows the standards and is needed for interoperability. That way, people can use it in their products, without being forced to release all the rest of their code if they don't want to.
Applications built on top of these, though, can be released under the GPL or other not-quite-as-open-as-BSD licenses. High-level apps like OpenOffice.org, or Apache Tomcat, or whathaveyou. These can be closed source or GPL'd or SharedSource as they are the money-makers, the apps you don't want your competitors "stealing" from you. (Hmmm, that para doesn't really make as much sense as I thought.)
Basically, the low-level stuff should be as open as possible, and be usable by as many as possible, with the least amount of restrictions possible, to allow the most people to use it in the easiest possible manner. That way, we all use the same foundation, and our products can all work together seamlessly.
I would tend to agree with this, but it really depends on what my goals with the software would be. do I want to make money with it? do I just want as many people as possible to make use of it? do I want to keep others from modifying and closed-sourcing my code?
I feel like I'm just stating the obvious here, haha.
whoa this thread got shorter in the time I typed that. archivin' shit.
Chiken
Don't Let Your Walls Down
also, you made an interesting point in the original post lucas. weighing the benefit of the sheer number of people using or modifying your code vs. the benefit of modifications to your code being open sourced.
I would tend to weight the former benefit more heavily because even if someone takes my BSD-licensed code, modifies it (or doesn't!), and releases it as closed source software, whoever uses that software is still benefiting from my work and — the most important point, I think — the original source code that I released remains open source. basically, someone got to use my work and the original work wasn't significantly harmed by it in any way that I can conceive.
(and typically, people benefiting from and using my work is what I'd be concerned with, not money or fame. this may change depending on the context in which I'm writing the code.)