Date: Sat, 15 Dec 2001 03:54:39 -0800 From: Terry Lambert <tlambert2@mindspring.com> To: "Gary W. Swearingen" <swear@blarg.net> Cc: chat@FreeBSD.ORG Subject: Re: IBM's intentions with JFS (was: IBM suing (was: RMS Suing was [SUGGESTION] - JFS for FreeBSD)) Message-ID: <3C1B39FF.C0F06938@mindspring.com> References: <3C186EA5.4EA87656@mindspring.com> <20011213093555.76629.qmail@web21107.mail.yahoo.com> <3C186381.6AB07090@yahoo.com> <3C1875D6.5DE4F996@mindspring.com> <1id71idej9.71i@localhost.localdomain> <3C1875D6.5DE4F996@mindspring.com> <20011213051012.Y56723-100000@turtle.looksharp.net> <3C186381.6AB07090@yahoo.com> <3C1875D6.5DE4F996@mindspring.com> <3C186381.6AB07090@yahoo.com> <20011214122837.O3448@monorchid.lemis.com> <3C19807D.C441F084@mindspring.com> <5ipu5i9u0w.u5i@localhost.localdomain> <3C19D716.3FC77047@mindspring.com> <c58zc5a68o.zc5@localhost.localdomain> <3C1A6E7F.3CF2E0EB@mindspring.com> <c48zc496pp.zc4@localhost.localdomain>
next in thread | previous in thread | raw e-mail | index | archive | help
"Gary W. Swearingen" wrote: > Terry Lambert <tlambert2@mindspring.com> writes: > > "Gary W. Swearingen" wrote: > > > ... I was just trying > > > to say that BSD-licensed code is always BSD-licensed code, even if a > > > derivative in which it appears is GPL-licensed (or even closed-source- > > > licensed) and the two licenses are not incompatible. > > > > The problem with this is that you are only licensed to use the > > GPL'ed code if you meet the terms of the GPL, which means that > > the code it is linked with is GPL'ed. > > But that statement doesn't say enough. The GPL requires the deriver > to license the derivative under the GPL, and thus his own code under > the GPL, but nothing in the GPL prevents him from also licensing his > own code (that which is NOT a derivative of the GPL code) under any > other license. (Some other licenses would be of no practical value > (like a closed-source license), but the BSDL would allow broader use.) I think you are not geting it. My (and others) concern is not that FreeBSD code would be GPL'ed (though that is a valid concern), but that the various licenses which the FreeBSD code is distributed under are incomptible with the GPL (e.g. the :BeerWare License" and the "4 clause BSDL", the "Bill Paul Voices in the head license", etc.). You can't argue that an OS consisting of a GPL'ed JFS and a BSD licensed everything else is _not_ a derivative work of both the GPL'ed JFS and the BSDL'ed everything else, particularly since it is widely acknowledged that an OS needs an FS to be useful at all. If all of FreeBSD were under the 2 clause BSDL, then the FSF would claim compatability of the license with GPL. It's not all under that license, however. But let's say for the sake of argument, it was. I would argue that the FSF's legal theory was wrong, since you can not license BSDL'ed code under a different license (e.g. the GPL) unless you are the copyright holder, or have an assign from the copyright holder(s): the BSDL is what permits you to use the code, and you are not permitted to change the license. The closest you could get would be an aggregate license, like the UofU "OSKit" GPL license. > The non-derivative code could even be put into the public domain. I believe that the BSD code would all be in the Public Domain already, if there wasn't a need for a "hold harmless" to stave off litigation. If the Congress could see fit to imply a "hold harmless" for works placed in the public domain, most of my code written on my own time for the purposes of reference would be placed in the public domain already. The GPL _went out of its way_ to make itself incompatible with the MIT Project Athena license, and therefore the BSDL, as well. I think you are seeing here the crux of the fundamental philosophical difference between the users of the GPL and BSDL: the intent is totally different. > Consider the opposite case of Linus putting a BSD-licensed driver into > his GPL'd kernel. He doesn't even own the copyright of the driver and > therefor certainly can't change the licensing of the driver, yet we > know that the kernel as a whole (being a derivative of the driver and > the old kernel) is under the GPL. There is no conflict. Both licenses > permit sub-licensing, the difference being that BSDL allows stricter > (but not less strict) sub-licensing, while the GPL allows only GPL > sub-licensing. Nothing prevents BSDL code, or even public domain code, > from also being part of a GPL derivative and being part of it doesn't > remove the BSDL from the BSDL code or remove PD code from the PD. For the 2 clause license, this is true. However, I would argue that doing this would prohibit the preparation of derivative works of the driver code in particular. Most BSD code is under the 4 clause license (the "claim credit clause" counts as an additional restriction, in the GPL's book). THis is because in order to distribute the derivative work, you would need to license it under the aggregate license, but the code you are eriving it from must be licensed under the 2 clause BSDL. At best, they can coeexist in the same kernel, but they are immiscible. I think that you would end up with a derivative work of BSDL'ed code that lived in the Linux kernel _only_ because of the exception in the Linux license granting treatment of the kernel as an LGPL'ed library for the purposes of access to kernel services. > > We are not talking "mere aggregation" when we talk about linking, > > I knew that. I've not been even considering aggregation here, except > maybe in my off-target anthology-book example. Your 2 clause BSDL'ed driver in a GPL'ed kernel is an aggregation example, IMO. You also neglect to note that, for drivers, the Linux kernel is treated as an LGPL'ed library due to a specific exception granted to licensees to allow the use of commercial third party drivers. > > They disagree with your conclusion? 8^) 8^) > > Again, I didn't understand, but one problem is that I don't have a > conclusion and I'm fairly sure only a court (or some lawsuit-threatening > licensees) will ever come up with a conclusion that should reduce our > uncertainty significantly. Again, feel free to present youself as a test case; I have better things to do with my time and money (like getting every oddball piece of hardware NetBSD claims to run on up and running -- and that's way in the heck down my list of useful pursuits). -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3C1B39FF.C0F06938>