Date: Sun, 23 Dec 2001 15:32:32 -0800 From: Dave Walton <dwalton@acm.org> To: chat@freebsd.org Subject: Does Linux violate the GPL? Message-ID: <20011223153232.4b562a74.dwalton@acm.org>
next in thread | raw e-mail | index | archive | help
Even though they can get a bit unruly at times, I enjoy reading these licensing threads, because they tend to give me new perspectives on the issues and help me understand them better. This most recent discussion made me realize that I've been suffering from a major GPL-induced misconception. Thanks to those who described the problem in ways that made this clear. The misconception comes from the viral nature of the GPL. Even on this list, people speak in terms of the GPL "infecting" other code, as if combining GPL code with other code automatically causes the other code to be under GPL. I believe the FSF encourages this muddy thinking, and this misconception is behind the attitude of many GPL zealots who believe they can alter the license on code they did not create. The truth of the matter is that the GPL does not "infect" code, but instead attempts to force authors to take action to license their code under GPL. In practice however, I believe the muddy thinking it encourages may be causing numerous copyright and/or GPL violations. There are three obvious scenarios that spring to my mind. Let's have a look at them... A creates software S under BSDL. (Note that the BSDL does not grant permission to alter the license, a right which is reserved for the author by default, contrary to popular slashdot belief.) B creates software T under GPL. 1. A decides to include T into S, but does not change the license on S. One of two possibilities exists: a. A is in violation of the GPL by not relicensing S and may not legally use T. b. Because A is the author of S, A might be considered to have implicitly licensed S under GPL, regardless of A's intentions. This example illustrates the dangers that FreeBSD flirts with and (hopefully) avoids. 2. B decides to include S into T, but does not arrange for A to release S under GPL. This causes two violations: a. B has violated the copyright of A by altering the license and may not legally use S. b. B has violated the GPL by not releasing ALL the code under GPL. 3. C creates a new software U, which is a combination of S and T, but does not arrange for A to release S under GPL. Once more, two violations: a. C has violated the copyright of A by altering the license and may not legally use S. b. C has violated the GPL by not releasing ALL the code under GPL. I find it mildly entertaining that the GPL makes it so difficult to use GPL software without violating GPL. This is freedom? And given that the GPL requires an entire work to be published under GPL, even the parts from other authors, how can any license be considered "GPL-compatible"? But I have to wonder... Linux (among many other projects) has borrowed code from BSD. If, as I suspect, relicensing was not arranged, doesn't that mean that Linux is in violation of both the authors' copyrights and the GPL itself? Dave -- ---------------------------------------------------------------------- Dave Walton dwalton@acm.org ---------------------------------------------------------------------- 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?20011223153232.4b562a74.dwalton>