IP: GNU license controversy

[David Farber <dave@farber.net>]

>Date: Mon, 7 May 2001 13:49:08 -0700
>From: Brad Templeton <brad@templetons.com>
>To: farber@cis.upenn.edu
>Cc: ip-sub-1@majordomo.pobox.com
>Subject: Re: GNU license controversy
>User-Agent: Mutt/1.2.5i
>Organization: http://www.templetons.com/brad
>
>
>Indeed, I've often wondered if this is an as-yet-unexploited bug
>in the GPL.
>
>This is because computers and software don't obey or disobey
>licences, only humans can do that.  Humans give computers orders.
>
>It should be possible, as I read the GPL, to prepare a set of
>proprietary modifications to some GPLed program, and then produce
>a software package with two components:
>
>         1) The original, freely distributable GPLed program, with source
>            where changes are to be done, and otherwise object code with
>            source available online.
>
>         2) A proprietary, binary module which takes the source above,
>            applies the changes to it, and also localizes the result for
>            the user and system in question.  It then compiles it into
>            object code that will only run on the user's system, and of
>            course does not leave the modified source code behind on
>            the disk.
>
>
>The result would be a binary-only, enhanced version of a GPL program
>which while it might arguably be freely distributable as a program
>derived from GPL code, but which won't run anywhere else making the
>point moot.   It might also contain severable binary components which
>are not distributable at all.
>
>Of course, the GNU project is not likely to modify the licence to stop
>this, since forbidding technology is not a popular philosophy there.
>
>So I think if a company were determined to make a set of proprietary,
>binary only packages that it sells which are based on GNU packages, it
>could.  Of course, they would get a lot of bad publicity over it but I
>doubt the law could stop them.



For archives see: http://www.interesting-people.org/