GPLv2 'web-app loophole'

Rod Dixon, J.D., LL.M. rod at
Tue Aug 7 12:13:23 UTC 2001

> On Monday 06 August 2001 08:42 pm, Rod Dixon, J.D., LL.M. wrote:
> > This sounds like "much ado about nothing." As is well-known, software is
> > not an easy fit within copyright doctrine. I am unsure whether
> there is a
> > relevant distinction between "use" and copy as far as software is
> > concerned. Copyright interests are invoked when one "uses"
> software as long
> > as the prevailing view is that a RAM copy is a *copy* as that terms is
> > defined and understood by reference to the Copyright Act. Hence, I doubt
> > whether an ASP or a "web-app" presents a case for a loophole in the GPL.
> If using the software constitutes copying, and thus regulated by the GPL,
> then does this mean I have to release private modifications I have made
> within my own home? I distinctly recall RMS arguing strenuously
> against this.

This is tricky because we need to distinguish what the license authorizes
from what the default rules of copyright law provide. I believe RMS is
referring to the terms of the GPL, which certainly may authorize in-home
personal use (and is likely to be consistent with copyright law regarding
*public* distributions). In MAI Systems Corporation v. Peak Computer, Inc.,
999 F.2d 511 (9th Cir. 1993), the Ninth Circuit held that unauthorized
"copying, for purposed of copyright law occurs when a computer program is
transferred from a permanent storage device to a computer's RAM." Many
commentators think the Ninth Circuit was wrong, but other courts have cited
this decision favorably. The upshot is that with regard to software, an
end-user's "use" of the program where the program is loaded into RAM
constitutes a copy within the meaning of the Copyright Act ( 17 U.S.C. 101).

> The GPL clearly states that the execution of the program is
> outside of the
> bounds of the license. And copyright law allows for the execution of the
> program in RAM by the legal owner of the copy. I don't see how anyone can
> argue that what I do in the privacy of my own home on my own
> computer with my
> own legal copy of GPLd software is the legal jurisdiction of the author.
> Public performance is another matter, and is unfortunately not
> covered by the
> GPL. Abe wanted to "feel out people's attitudes regarding this 'web-app
> loophole'", so I offered up my attitude. Regardless of what the authors'
> rights may be by law, I don't believe that Open Source / Free
> Software should
> make any attempt to restrict public performance.

Agreed. Public performance does not seem to be the most likely argument to
make in this type of case.
> But I don't think that web-apps even constitute public
> performance! When a
> book is read aloud or a film shown for public performance, the
> audience will
> actually receive a copy of the copyrighted materially aurally and
> visually.
> But such is not the case for a web application. The software will only be
> copied to the legal owner's servers, and the only thing the user
> receives is
> the output of the program, not the program itself.

Hmm... This is an interesting argument. You seem to be saying that you doubt
that some/all web-apps (scripts that execute on the server) load into an
end-user's RAM? If so, there might be a plausible way to distinguish some
web-apps, but this migth not lead to rigorous legal analysis (too many
exceptions). Of course, the person who posted the question was also
referring to the alleged initial impermissible copy by the user who
downloaded (copied) the script and uploaded (copied) it to his/her server.

> An analogous situation would be for an attorney using copyrighted
> law books
> to prepare a brief for a client. The law books have not been performed
> publicly.
> I myself have had one of my applications converted to a CGI
> program with a
> public interface on a public server. I could not in good
> conscience refer to
> my program as "free" if I placed restrictions on that type of
> usage. To me it
> would be beyond the bounds of propriety. If I wanted to restrict
> the program
> in such a way, then I would not make it Open Source, and instead
> lease it out.
> --
> David Johnson

I agree that the public performance issue is likely to be a non-starter in a
case like this. I am still unsure why this is viewed as significant.
Certainly, you do not mean to say that the GNU GPL or some other open source
license cannot be applied to a  web-app or any software application running
on a server because the public performance right is outside the scope of the
license as it iscurrently drafted? If so, then I am hopeful that the case I
cited at the outset should have some impact on your thinking.


More information about the License-discuss mailing list