Re: the FSF's definition of Free Software and its value for Debian

[Jakob Bohm <jbj@image.dk>]

On Thu, Mar 20, 2003 at 12:47:48AM +0100, Henning Makholm wrote:
> Scripsit Jakob Bohm <jbj@image.dk>
> > On Tue, Mar 18, 2003 at 09:49:03AM -0500, Branden Robinson wrote:
> 
> > >   Each time you distribute the Document (or any work based on the
> > >   Document), you grant to the recipient and all third parties
> > >   in possession of the Document the authority to gain access to the
> > >   work by descrambling the work if it is scrambled, decrypting the
> > >   work if it is encrypted, and otherwise avoiding, bypassing,
> > >   removing, deactivating, or impairing any and all technological
> > >   measures effectively controlling access to the work.
> 
> > I think it is still too widely scoped.  If I place a copy of the
> > document in a gpg encrypted e-mail along with something not for
> > public consumption, all the ISPs on the way are in possession of
> > a copy.  This clause allows them to crack it open.
> 
> Yes, but if gpg works as intended (and I'm reasonably certain that a
> lot of smart people have tried without success to demonstrate that it
> does not), they *can't* crack it open, irrespective of the right they
> get. Notice that it is not a right to demand that you supply a key.
> 
> > Also even in the intended case of allowing people to decode CSS
> > like protection on a copy, the clause may make the whole
> > document non-distributable due to a conflicting obligation to
> > obey a DMCA-like law.
> 
> Is there any "DMCA-like laws" anywhere that say that a copyright
> holder can *not* authorize other people to access his work?
> 

I don't know, but if there are not, and a lot of people start
using such licenses, the big media companies are likely to get
their supporters in government to enact an amendment stating
that just because the copyright holders of *some* works
protected by a technology authorize locksmithing, this does not
allow ordinary citizens to create, distribute or possess tools
which can also crack the works copyrighted by the big media
companies.

Looking at it from another perspective, anti-copy-protection
clauses in free software licenses are intended to deal with two
unrelated issues:

   1. Someone might store a copy of the free software in a
copy-protected format thereby subverting the virality of the
license.  My alternate method of defining such non-reversible
forms as "binaries" not "source" (= the preferred for for making
changes) solves this by forcing the person putting the free
software under copy protection to also provide an unprotected
copy of the source.

   2. If a copy protection scheme relies on the inability of
users to change certain aspects of the programs legitimately
processing the format, then it becomes a legal impossibility to
provide or develop such programs as free software (because free
software by definition empowers the users to arbitrarily change
program behavior).  While the previously suggested clause does
not deal directly with this issue, it may be part of a grander
scheme to create enough excuses for someone developing a new
DeCSS under the pretense of gaining access to free software
explicitly authorizing such activity in its license.

Jakob

-- 
This message is hastily written, please ignore any unpleasant wordings,
do not consider it a binding commitment, even if its phrasing may
indicate so. Its contents may be deliberately or accidentally untrue.
Trademarks and other things belong to their owners, if any.