Date: Mon, 25 Mar 2002 04:08:09 -0800 From: Terry Lambert <tlambert2@mindspring.com> To: Hans Reiser <reiser@namesys.com> Cc: hiten@uk.FreeBSD.org, Greg Lehey <grog@FreeBSD.org>, Chris Mason <mason@suse.com>, Josh MacDonald <jmacd@CS.Berkeley.EDU>, Parity Error <bootup@mail.ru>, freebsd-fs@FreeBSD.org, reiserfs-dev@namesys.com Subject: Re: [reiserfs-dev]i Re: metadata update durability ordering/softupdates Message-ID: <3C9F1329.E4BFC6D4@mindspring.com> References: <20020318174641.A1153@hpdi.ath.cx> <3C9676B4.49A76589@mindspring.com> <3C9E1DA4.1090703@namesys.com> <3C9E6E28.9D0B8778@mindspring.com> <3C9F0500.4050206@namesys.com> <3C9F0730.418A94BE@mindspring.com> <3C9F1638.4020707@namesys.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hans Reiser wrote: > Terry Lambert wrote: > >The GPL is like "part B" of a binary nerve gas bomb. As long > >as you don't combine it with a "part A", there's no nerve gas, > >and no one gets hurt. > > But you did FUNCTIONALLY combine it (or prepare it for combining), which > is to say that you performed more than mere aggregation. This is the fuxzzy part. Yes, it was prepared for combining, but it was not prepared exclusively for combining. The difference is whether or not the code that it's being combined with was modified to support an interface supported only by the GPL'ed code, or not. There are three places where this might be an issue with an FS: 1) The OS code will only operate with EXT2FS, and the EXT2FS code is GPL'ed. FreeBSD's use of the EXT2FS code doesn't fall under this, because the majority of the code is not GPL'ed, only some of it, and the code which is not was modified to operate with FreeBSD, and not the other way around. 2) There are additional components that exist only to interoperate with the GPL'ed code, and therefore these components should be considered GPL'ed, as they are arguably derivative works. A case might be made that the non-GPL'ed components of the FreeBSD EXT2FS implementation, except that these compnents are shared with an ecxisting implementation, so they don't constitute code that was purpose-written. 3) The on disk format for the FS constitutes a data interface, and the implementation of that interface is GPL'ed, and therefore the interface is GPL'ed. This doesn't work if the data format is published, which it can't help but be, for GPL'ed code (in fact, Partition Magic supports EXT2FS partitioning, and has no GPL'ed code in it to implement it). There was actually a lawsuit brought by the FSF over a case of a software product for cryuptography whic used an interface to an LGPL'ed library which was unique. The case failed to go to trial, when the software vendor wrote a plug-in replacement for the LGPL'ed library to show that the code was not itself derivative of the library because of the unique interface. Actually, this would have been a hard case to win, since the interface copyright failure by Apple, Microsoft, and the earlier Ashton Tate vs. Clipper, Inc. case over the dBase III language definition, in which it was held that a computer language could not be copyright (this is als an argument against case (3) above, which is an equivalent data interface). Actually, there's a guy at the UofU who got his Juris Doctorate (PhD in Law) in intellectual property law on the basis of software license and copyright; if this thread goes on for very much longer, I might have to take his name in vane to get him to give legal opinion, if he's up to doing it without charging for the act (you there, Lee? 8-)). -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-fs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3C9F1329.E4BFC6D4>