On Fri, 17 Apr 2009, Anthony W. Youngman wrote:>I was under the impression that the FSF thinks that if it's illegal to>link a program with GPL software and distribute that, it's also >illegal if you>just distribute the other program and have the user do the link. HOW? I hope the FSF doesn't think this, because imho it is so sloppy legal thinking as to be incompetent!http://sources.redhat.com/ml/guile/1999-02/msg00151.html
This talks about "static or dynamic linking". I don't actually see how it applies, because if it's statically linked it's a clear violation - the person distributing the program has to distribute the library as well. But if it's dynamically linked and the program - as distributed - merely EXPECTS to find the library on the target machine, I don't see any violation.
http://www.gnu.org/licenses/lgpl-java.html
I don't understand this.
http://www.gnu.org/licenses/gpl-faq.html#GPLPluginsInNF http://www.gnu.org/licenses/gpl-faq.html#GPLIncompatibleLibs Also http://www.fsfeurope.org/projects/gplv3/bangalore-rms-transcript : Eben Moglen: As when, for example, people tried to draw a line between static linking and dynamic linking under GPL version two, and we had to keep telling people that whatever the boundary of the work is under copyright law, it doesn't depend upon whether resolution occurs at link time or run time.
Ummm...This whole thing is a rather grey area, but I still stick by what I said. You may have noticed references to the "system library exception". Is that there as a valid exception, or because they're not sure whether it'll stick in court?
At the end of the day, if the proprietary program does not contain any GPL code *as* *shipped*, I find it hard to see a "copyright violation" suit sticking. Who is violating the GPL? The FSF would like to say it's the proprietary vendor but ... (and it's certainly not the user, the GPL explicitly says they're in the clear).
Cheers, Wol -- Anthony W. Youngman - anthony@thewolery.demon.co.uk