Ownership in Software

Frank Heckenbach frank at g-n-u.de
Thu Apr 22 18:06:23 UTC 2004

Axel Schulz wrote:

> > > You are free to create software and you are free to share the
> > > source code but if you force someone to give away his work or
> > > doesn't allow him to do with it what he wants to do you violoate
> > > his rights and freedoms.
> > 
> > Yes, and that's exactly what Stallman advocates. E.g., he criticized
> > another license for requiring that each derived work must be
> > distributed or published.
> But doesn't that presupposes that you have the source code? I have
> to accept a license by M$ when I want to install their software.
> But how shall I distribute or be forced to publish a derived work?

It doesn't apply to such license. The case was about a license which
provided source code access, not completely unlike the GPL, but with
such extra clauses, e.g. that each derived work *has* to be
distributed (which I also don't like).

> If Stallman really says so, he goes a little to far I think. And
> this is my problem. I like his ideas. I make use of the FSF and
> the GPL [1] but to much freedom harm others, just as to much
> restrictions do nowadays.

I think there's a misunderstanding. This other license required it,
and Stallman objected to it. So what he requested is exactly what
you said in the first paragraph above. (Of course, as long as you
don't intrepret "to do with it what he wants" as "to restrict others
from doing certain things".) And that's just what the GPL says --
you can use the software for any purpose, you can modify it, and
you're *not* required to give away your modifications. (But *if* you
choose to, you cannot use them to restrict the rights of the
recipients any further.)

> > But as soon as you give the work to someone else, every
> > use of your power affects others' freedoms as well, so any arguments
> > based on "absolute freedom" fail at this point, and you have to
> > weigh the freedoms of both parties against each other.
> It is tricky to talk about "absolute" entities. Rights, duites,
> freedoms and so on can be absolute. An American philospher (Mackie
> [1]) claimed that the only absolute right in moral philosophy is
> the right to life.

(Even this is sometimes questioned in discussions about war, death
penalty, etc.)

> If we accept this, than we have to measure and weigh laws and our
> principles and conventions in respect to that right. But this is
> beyond the purpose and competence ;-)) of this list.

Thank you. (Well, fortunately, the list rarely takes part in
discussions itself. Usually it's people writing to the list. The
exception is me who is a robot. ;-)

> Back to the absolute freedom. I think the concept still holds. I
> do not accept that there is a natural relation between the
> "absolute freedoms" of the "power-holder" and the one who takes
> the software. [...]

I don't have the time to redo the whole discussion that others have
done already. However, your argumentation sometimes looks like
asserting absolute (WRT the given software) freedom/power to the
"owner", including the right to restrict others in any way he likes.

Since we agree that absolute principles aren't applicable here, I
think your argumentaion isn't very sound. Even if you assume that
the "owner" can impose some conditions on the receipient, where
would you draw the line? Some examples for you to think about (all
are very real with proprietary software):

- The right to fix problems (especially if the "owner" doesn't do

- The right to get compatibility (if interfaces and data formats are
  undocumented, sometimes this requires reverse engineering).

- More generally and related to the previous point, the right to
  your own data. (If you store them in a proprietary format, you
  might not be able to do with them what you want, or you might even
  depend on the proprietary vendor (repeated license payments etc.)
  just to be able to access your own data on another computer, or in
  the future, ...)

- The right to check if the software does anything nasty which could
  harm you (intentionally as in backdoors, or unintentionally as in
  security holes). This requires source code or reverse engineering.

- The right to your computer at all. (Think about DRM, or about
  required remote access to the "owner" for checking your license
  compliance, etc.)

- The right to do unrelated things. (I think some proprietary
  companies once tried to restrict you from using certain free
  programs at all on the same machine. I don't know, though, if this
  restriction was legally valid.)

I hope you agree with at least some of the points. So now it's up to
you to provide a definition where the "owner"'s rights should end.
Just saying the "owner" can put any conditions on the recipient
would deny all of the above rights.

Axel Schulz wrote:

> Simo Sorce wrote:
> > On Thu, 2004-04-22 at 18:27, Axel Schulz wrote:
> >
> > > The GPL is a very smart solution. It restricts also the "taker"
> > > of the license but it does something very good to society. But
> > > this software has an author. That is way I hold that Stallman is
> > > a little mistaken.
> > >
> > > What do you think?
> >
> > Never heard that stallman said software should not have authors (from
> > which derive authorship).
> >
> > Can you show me where Stallman states that software should have no
> > authors please? Can you give me any pointer?
> >
> > authorship != ownership, authors != owners
> You consider it this way. Many people here consider it this way. 
> I do not. My justification is that only authorship in software is
> insufficient because of the dual nature of software (-> the ACM
> artilce I quoted).
> So, for me  "authorship != ownership" is not true when it comes to
> software.

Just to point out a logical flaw in your argument: Stallman says
(and you disagree) that software should not have owners. You say
(and Stallman probably disagrees, like most of us here do) that
authorship = ownership. Combining these two statements results in a
statement ("software should not have authors") that neither one
necessarily agrees to, i.e. it's meaningless.

> Authorship is not made for software, but applied to software it
> becomes ownership. This is sufficient. Because than we do not nee
> patents and stuf like this.

*We* don't need patents, anyway. But those who (think they) need
patents, will still do, regardless of how strong copyright is made.
With copyright they can't "kill" independent reimplementations, but
with patents they can. (And that's what they want.)


Frank Heckenbach, frank at g-n-u.de
GnuPG and PGP keys: http://fjf.gnu.de/plan (7977168E)

More information about the Discussion mailing list