Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 12 Feb 2003 21:32:07 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 12 Feb 2003 21:32:07 -0500 Received: from mail.webmaster.com ([216.152.64.131]:28861 "EHLO shell.webmaster.com") by vger.kernel.org with ESMTP id convert rfc822-to-8bit; Wed, 12 Feb 2003 21:32:05 -0500 From: David Schwartz To: CC: , Linux Kernel Mailing List X-Mailer: PocoMail 2.63 (1077) - Licensed Version Date: Wed, 12 Feb 2003 18:41:53 -0800 In-Reply-To: <20030213022127.GB13897@bjl1.jlokier.co.uk> Subject: Re: Monta Vista software license terms Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT Message-ID: <20030213024154.AAA23702@shell.webmaster.com@whenever> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4653 Lines: 110 On Thu, 13 Feb 2003 02:21:27 +0000, Jamie Lokier wrote: >David Schwartz wrote: >> 1) I have all the rights to the difference between the original >>work >>and the derived work. (Since I made the derived work, I must.) >The logic of derivation does not work like that. >You don't have all the rights to the difference. >It is not a case of B = C - A. People do not own individual lines >or >bytes of the source code. (Sometimes they do, but often they >don't). >The "difference" between the original work and the derived work is >_itself_ a derived work, because you can't possibly *create* a diff >without deriving from the original work. I don't think you understand what I mean by "difference". It's not a precise mathematical term and it doesn't refer to 'diff' output. It means the rights to the derived work apart from the rights to the original work. >In the act of you creating the diff you must have worked from the >original GPL'd work, so the diff is derived from that, as well as >from >your original work. When you write "diff -ur orig-kernel my-kernel >>file", the output from that command is a derived work of the >orig-kernel directory. Certainly, but that's not what I'm talking about. >This is true even if you didn't copy any context lines from the >original tree. > >That means your difference is covered by the GPL, even though you >wrote the changes. Of course the original work is covered by the GPL and of course the derived work as a whole is covered by the GPL to the extent that the original work is covered by the GPL. >(If you did manage to write the same diff working from other >sources, >or if the diff only makes "fair use" of the original, that would be >different. But any large, complex diff such as a patch to the Linux >VM simply cannot be written without starting from the original Linux >VM >code. It is a question of how well the glove fits the hand, as it >were). When I say 'difference', I don't mean the output of a 'diff'. I mean the literal subtraction. I mean the derived work less the original work. Or, to put it another way, the rights to the output of the additional creative effort added by the person who made the derived work. >(If the difference was just whole files using a well defined >interface, such as a device driver or filesystem, your point seems >quite applicable though.) Suppose you write a play and then I write an adaptation of that play in the form of a screenplay for a motion picture. To produce a motion picture based upon my screenplay, you would need certain specific rights to your play and certain specific rights to my adaptations. By "difference" I mean the rights to the adaptations themselves. I do agree that it is nonsensical to have, say, distribution rights to the adaptations without rights to the original. In other words, to have distribution rights to the derived work, you must have distribution rights to the original. You must also have distribution rights to the "difference". This is a notional legal thing, there does not have to exist such a "difference" as an actual object. Consider if I translate a book you wrote. To distribute the translation, you need the right to distribute the original work, the right to make the translation, and other rights to the "difference", that is, the right to distribute the translation apart for the right to distribute the original work. I think you're hung up on my use of terminology. If you don't like the word "difference" substitute "derived work apart from the original work". For example, if I translate your book into another language, the thing you need to get from me to distribute the translation is what I call the "right to distribute the difference", but of course there is no actual "difference". Call it the "right to distribute the translation apart from the right to distribute the original work" or the "right to distribute the additional creative work added in the derivation process". It's a complex and technical legal issue, but I'm not doing anything unusual with it. See, for example, section 103 which states: "The copyright in a compilation or derivative work extends only to the material contributed by the author of such work, as distinguished from the preexisting material employed in the work, and does not imply any exclusive right in the preexisting material." DS - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/