Darren Oh wrote:
I still think the wrong question was answered. The GPL is a one-way infection: non-free software cannot GPL components, but GPL code can use non-free components.
I'm not a lawyer, but I think that's impossible. The GPL cannot tell me what I can and cannot do with my own, proprietary code, even if I provide instructions to end users telling them to compile my code with a GPLed library. And unless calling a library's functions constitutes a derivative work (which it doesn't seem to, see <http://www.rosenlaw.com/lj19.htm>), the GPL cannot prevent me from downloading a GPL library and working on my own proprietary code that compiles against the library. I can then distribute that proprietary code with the aforementioned instructions (though not the library itself). I know this isn't how most people interpret the viral nature of the GPL, but I simply don't see the mechanism that the GPL would use to prevent proprietary software developers from compiling their code against GPLed libraries in development and distributing the proprietary source code in a non-GPL form to be compiled with GPL code by end users. At what point would the developers be forced to accept the GPL's terms? You have to agree to the GPL if you (re)distribute GPLed works -- but not if you only download them, use them, or install them. If you download a GPL library and develop proprietary code that uses the library, you have not been forced to accept the GPL. For this reason, I think it's ridiculous when GPLed programs ask me to "accept" the GPL on installation. The GPL is quite clear about this:
Activities other than copying, distribution and modification are not covered by this License; they are outside its scope.
Moreover:
You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works.
Now, if the GPL required that you accept it to *use* code, it might be a different matter because you would be agreeing to the infection clause in the library's license. But this infection clause seems to go well beyond what normal copyright would consider a derivative work. I would love to have someone clear up any errors in my interpretation. While I love the FSF, they have a vested interest in presenting the GPL as viral as possible. So, I don't consider their FAQ to be a reliable, unbiased source.