Grzegorz B. Prokopski writes: > On Thu, 2005-13-01 at 19:19 +0000, Andrew Suffield wrote: > > On Thu, Jan 13, 2005 at 08:02:57PM +0100, M?ns Rullg?rd wrote: > > > > Derivation is something that happens when you *write* the program. Not > > > > when you build it. > > > > > > How many times does it have to be stated that *using* an API does not > > > form a derivative work of *any* implementation of the API? > > > > More irrelevancy. Using an API is just a method for utilising part of > > another work. It may or may not be a derivative. This is just a > > variation on the 'linker' noise. > > Maybe, but then you should have your own, non-GPLed implementation of > these APIs agains which you're compiling your code. Ie. glibc has its > own collection of kernel headers. > > Now, what we observe here, is that a GPL-incompatible code is > explicitely compiled agains headers (.class files for that matter) > that are GPLed.
Which GPLed class files is Eclipse "compiled against"? ("import java.util;" does not count: java.util identifies an API with non-contaminating implementations.) > Even according to Linus such "use" is not permitted > > http://seclists.org/lists/linux-kernel/2003/Dec/1042.html Linux header files are different from Java packages in a number of ways. One difference is that C header files, especially those from Linux, contain function bodies as well as function declarations; Java packages define a functional interface but do not insert expressive content in a file that imports the package. Another difference is that Linux is the only implementation of many functions; using those means that you require Linux specifically, and cannot use an alternative such as the FreeBSD kernel. > It seems though you could compile your code against another set of > headers (.class files), that does not cause license conflict and it > would be OK. > > We do not have to push interpretation of GPL as far as to claim > copyrightability of the interfaces. Using GPLed implementation of > such interface to create your work does however create a derivative > work. Please cite law to support this hypothesis. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]