Re: digital images

On Thu, Jan 27, 2000, Lynn Winebarger <owinebar[_at_]free-expression.org> wrote:
>
> On Mon, 24 Jan 2000, Karsten M. Self <kmself[_at_]ix.netcom.com> wrote:
> >
> > Interesting. I was about to tell Peter he was wrong and that it was
> > the *libraries*, not the compiler itself, which were licensed under
> > the GNU LGPL (the GNU Lesser (formerly Library) Public License). But
> > I checked, and in fact both gcc and g77 (the GNU Fortran-77 compiler)
> > are licensed under the LGPL. My understanding was that while the
> > libraries needed to allow incorporation into other works, the compiler
> > itself didn't. Curious.
>
> Where did you hear this from? AFAIK, the compilers are specifically
> under the GPL. How else do you think they got NeXT to donate the
> Objective C compiler?

Good question. I'm scratching my head.

I thought I'd checked the 'copyright' files under /usr/doc for the gcc and g77 packages on my Debian GNU/Linux (2.1/potato) box and read that both programs were licensed under the LGPL. Now I'm reading same and finding GPL for both gcc and g77. libg++ is licensed under LGPL. Must've been a flashback or somethin'. Got to check the coffee beans more carefully....

In either event, NeXT would have been compelled to contribute their code or remove the GNU gcc code. LGPL allows works which utilize the library only in linking to be considered independent. Modification of the library itself produces a derivative work and is subject to the relicensing requirement:

    5. A program that contains no derivative of any portion of the     Library, but is designed to work with the Library by being compiled     or linked with it, is called a "work that uses the Library". Such a     work, in isolation, is not a derivative work of the Library, and     therefore falls outside the scope of this License.

http://www.fsf.org/copyleft/lgpl.html

BTW: Peter, you were wrong. <g>

> The object code produced by a compiler is a derivative work of the
> compiler, AFAIK, but I believe the FSF walks a fine line (and knows
> it) between (a) wanting to preserve the spirit of the GPL, and (b) not
> strengthening copyright law. So I don't know whether they would ever
> try to assert such a thing, or even be interested in asserting such a
> claim.

I'm agreeing with your conclusion but not the logic here.

Claiming authorship to programmatic outputs of one's own program, when run by someone else on inputs of their creation, seems to fall outside the scope of copyright: "original works *of authorship* fixed..." (emphasis added). The authorship requirement appears to be missing in this case on the part of the compiler writer as applied to compiler outputs. Rather the compiler is generating a mechanical translation of a work, presumably of authorship, written by the programmer of the now compiled work.

Legal interpretation of "of authorship", anyone?

-- 
Karsten M. Self (kmself[_at_]ix.netcom.com)
    What part of "Gestalt" don't you understand?

SAS for Linux: http://www.netcom.com/~kmself/SAS/SAS4Linux.html
Mailing list:  "subscribe sas-linux" to mailto:majordomo[_at_]cranfield.ac.uk