[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: What makes software copyrightable anyway?

On 5/12/05, Raul Miller <moth.debian@gmail.com> wrote:
> On 5/12/05, Michael K. Edwards <m.k.edwards@gmail.com> wrote:
> > And, I might add, this is another respect in which the FSF FAQ verges
> > upon the dishonest.  Since 17 USC 117 explicitly limits the scope of
> > what can be considered infringement under section 106, it also
> > nullifies any claims of contributory infringement when a distributor
> > arranges for things to be combined at run-time.
> This doesn't do anything for the distributor of copyright infringing
> software.  17 USC 117 only protects users of that software.

Stipulate that, _contrary_to_law_, we read "mere aggregation" to mean
_only_ on storage media and _only_ at the level of putting debs or
tarballs side by side within a CD image, and permit no other
"collective works" AT ALL, for lack of sufficiently explicit language
in the GPL.  In the dynamically linked case, the only time at which
even an uncopyrightable collection-with-trivial-selection-criterion is
created, other than under the stipulated "mere aggregation" exemption,
is at run-time.  Under 17 USC 117, that is NOT AN INFRINGEMENT.  And
note that 17 USC 117 uses the phrase "copy or adaptation"
consistently, so it's not an infringement even if the dynamic linker
munges things around to create your _legally_incorrect_ idea of a
derivative work.

So there is no "copyright infringing software" involved, period.  And
read up about "contributory infringement":  no infringement by
recipient, no contributory infringement by distributor.

> > And that's a genuine respect in which it is safer to link dynamically.
> >  Even if a crack-smoking court ignores all precedent by proceeding
> > directly to copyright law AND construes "work based on the Program" to
> > include collective works (no, Raul, that would not be an automatic
> > consequence of the former), you've got an unassailable defense under
> > 17 USC 117.
> Of course, if the court nullifies the copyright then protection is not
> needed.  But the court never has, at least not in the case of the GPL.

Sigh.  "Nullifies the copyright"?  What is that supposed to mean?

> The closest the GPL has come to being nullified is the preliminary
> injunction in Progress Software v.  MySQL, where the judge granted
> a preliminary injunction preventing Progress from distributing or
> sublicensing MySQL, and deferred other consideration to MySQL pending
> a full trial.

This is hogwash.  The judge granted a preliminary injunction against
use of MySQL's _trademark_ without license in Progress's promotional
materials.  She carefully stated that Progress was entitled to
continue to use the trademark in certain ways that do not require
license, such as "stat[ing] that its product operates with the MySQL
program".  Then, she evaluated the separate request for preliminary
injunction on GPL/copyright grounds, and denied it, explicitly
sanctioning Progress's continued "use of the MySQL program under the

It is not perfectly clear how the judge would have ruled had Progress
subsequently continued to _distribute_ the MySQL program under the
GPL, and had MySQL AB claimed contempt of court in a later proceeding.
 That depends entirely on whether such conduct would require trademark
license and/or would violate the letter of the preliminary injunction
(which could be read to enjoin Progress, based on past misconduct,
from ordinarily non-infringing uses of the MySQL mark).  GPL and
copyright issues would have absolutely no bearing on that analysis.

> His reasoning was that granting a stronger preliminary injunction
> against Progress would destroy that company and thus render the
> trial irrelevant.

_Her_ reasoning (don't you even read the opinions you mangle?) was a
great deal more conformant to applicable law and procedure than you
imply, as I have repeated time and time again.


- Michael

Reply to: