> An assembly exception is sort of a way to neuter a > license. Suppose I > have two files, gpl.c and harpster.c. gpl.c is dual > licensed under CDDL > and GPLv3. harpster.c is licensed under the Harpster > license, a > proprietary license that solves world hunger. ;-) > > Now normally, linking gpl.c and harpster.c would > force harpster.c to > also be licensed as GPL. That's the "viral" nature > of GPL everyone > talks about. If I wrote gpl.c, I can place upon it > an assembly > exception that says, "when you link gpl.c with a > Harpster licensed file, > don't force the Harpster licensed file to be GPL." > Because I wrote and > wn the original gpl.c, I can modify the terms of the > license to be more > restrictive. (Or less restrictive depending upon > your point of view.)
How would FSF likely feel about GPLv3 modified by an "assembly exception"? And even if they wouldn't get publically upset about it, what about the difference between GPL being per-program while CDDL is per-file? And what happens even with a tear-off involving something dual-licensed with both licenses per-file? The original of foo.c may be dual-licensed, but someone could fork it (still dual-licensed) to now invoke extended functionality kept in the separate (non dual-licenced) bar.c. You can't pick up their change, then (at least not without reimplementing their bar.c file). In my seldom-humble view, the real solution would be just work more with folks using compatible licenses (the *BSDs, for example), and forget about trying to exchange code with the (however well-meaning) ideologues of the FSF and the fans of their viral license, until they grow up. They've more than achieved the realistic aspects of their objectives anyway, in that there's multiple open OS's and lots of open apps out there. Hoping to achieve more idealistic goals (i.e. opening _all_ software that is distributed beyond the originator's organization) is IMO unrealistic because even if every existing OS and app that was distributed was GPL (but with the shared libs or other environment interfaces LGPL), that still not only wouldn't preclude new proprietary apps, but wouldn't even preclude extensions of the environment (servers on a microkernel like HURD or interposers or LD_PRELOAD wrappers on something more conventional like Solaris or Linux). Look at libusb; effectively, it allows moving what used to be specialized kernel driver functionality into user space. The line between environment and application is blurred enough that one can no longer demand the former be open and yet tolerate the latter closed. And proprietary apps IMO will _never_ go away, insofar as open development models will not work with _every_ opportunity. So I think they've done just about all they can in terms of forcing code open. OTOH, I think they're right on track with wanting to extend the notion of open source to deal with patent and DRM issues. Choice of venue could be argued equally on behalf of the author and the user, but I see the author as the scarcer resource, so the whole DFSG position on that seems strange to me, unless it were refined to deal with specific problems, i.e. oppose only venues where the user had pretty much zero chance of winning. Their (DSFG) dissident test, although admirable in theory, also has problems when dealing with a composite work where some accountability is needed to preserve reliability and reputation, although for some specific works, it might be appropriate. And so on. So most of the more ideologically-motivated open source license terms have narrow circumstances where they're very useful, but mostly they're a barrier to collaborating communities (which of themselves probably result in more code being opened), and thus arguably self-limiting. I think there needs to be a recognition of that, keeping the more problematic terms confined to cases where they're actually an issue, without creating undue barriers to collaboration. In that regard, I think the CDDL reflects a greater state of maturity (pragmatism as enlightened self-interest, in this case) than the GPL, just as most people start out liberal but end up middle-of-the-road or slightly conservative (no, not just a different kind of ideologue still out to impose their ways on others!) later on in life. This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
