Return to BSD News archive
Path: sserve!manuel!munnari.oz.au!mips!mips!sdd.hp.com!caen!zeus.ieee.org!rutgers!igor.rutgers.edu!dumas.rutgers.edu!hedrick From: hedrick@dumas.rutgers.edu (Charles Hedrick) Newsgroups: alt.suit.att-bsdi,comp.unix.bsd Subject: Re: UNIGRAM's article on the USL-BSDI suit Message-ID: <Aug.8.19.14.25.1992.26820@dumas.rutgers.edu> Date: 8 Aug 92 23:14:26 GMT References: <45961@shamash.cdc.com> <25138@dog.ee.lbl.gov> <1992Aug3.143259.23897@crd.ge.com> <5486@mccuts.uts.mcc.ac.uk> <1992Aug5.003803.28740@engr.uark.edu> Followup-To: alt.suit.att-bsdi Organization: Rutgers Univ., New Brunswick, N.J. Lines: 56 tep@engr.uark.edu (Tim Peoples) writes: > I just finished reading the ammended complaint that added the Regents >to the suit and from what I got out of it USL is not claiming that there >are SPECIFIC PARTS that contain infingements, but that NET2, as a whole, >is an infringement. So, go figure..... Normally, when someone wants to implement a system that will compete with an existing system, they do everything possible to isolate the development effort from the existing system. If they need to look at the existing code, they have someone do it who is not part of the development team, and that person produces only functional specifications, what are then used by the developer. This is the "clean room" approach. Berekeley has done something interesting, for which there may be no legal precedent: they have started with another vendor's product, modified it, and then attempted to remove vestiges of the original. The reason this is legally dangerous is that when you start with someone else's product and modify it, what you have is a "derived work", which is still covered by the original copyright. The question is whether you can continue modifying it, and at the point where none of the original is left it becomes underived. If this has been litigated before, I haven't heard of it (which might not be surprising, since I'm not a lawyer). I could imagine that the status as a derived work would continue. Note also that Unix is made up of a number of files. Is the copyright only on the individual files, or on the work as a whole? If it's on the individual files, then one could purify the code by rewriting it file by file. But I would guess ATT will claim that it is on the work as a whole. Thus rewriting it file by file might not remove the status of BSD as whole as a derived work. If this is the argument they are going to use, then even a comparison that showed there to be no similarity might not be conclusive. This may be the reason that the suits do not contain detailed lists of where copying occured (though it's not clear to me that they would in any case), and the reason that ATT is not interested in comparing only snapshots. The issue is also complicated by the fact that ATT seems to be claiming both trade secret and copyright protection for the same work. Since copyright protection is intended only for published works, it's not clear whether this is valid. In my view, this case is full of strange and interesting legal issues, which may make much of the discussion here irrelevant. I have a suspicion that the decision may turn on details of these issues, and not on what computer scientists would regard as the substance: whether BSD has any longer any ATT code in it. I've looked at some decisions in areas like this, and about all one can say with any certainty is that many judges have a very different view of the world than we do, and that it's hard to predict even what questions they will consider relevant, much less what conclusions they will come to. There are some signs that the much of the legal profession is beginning to realize what a disaster they are creating, and that they are moving towards a sensible view. It would be nice to think that this case would be handled in such a light. But I'm afraid I don't trust our legal system that far.