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

Re: Corel's apt frontend



Raul Miller <moth@debian.org> writes:
> > Sure, but a frontend isn't mere aggregation -- in this case if you take
> > out the GPLed part of the system, the performance of that front end
> > can't happen.

On Sun, Oct 31, 1999 at 08:36:39PM +0100, Henning Makholm wrote:
> A front end is a front end is a front end.

A variable is a variable is a variable.  That doesn't mean
that all variables are equivalent.

> It's capable of controlling any program that has a user interface that
> coincides with a certain subset of dpkg's user interface.

Clearly you're not talking about all front ends here -- you 
can't be talking about ddd, for example.

So you're talking about the Corel front end for dpkg.  And that's very
clearly not designed to be a program which interfaces to arbitrary
package managers, but a program which interfaces specifically to dpkg.

Corel would be distributing it in conjunction with dpkg and not, for
example, in conjunction with sun's package manager.  And I think it
would utterly fail if you renamed pkgadd as dpkg.

> Claiming that this makes the front end a legal deriviate of any random
> program it can control is suspiciously close to claiming an interface
> copyright. In a wierd, backwards way, but an interface copyright
> nevertheless.

More like a copyright for where there is no interface.

For example, if Corel supplies its own implementation of dpkg to run
under the front end then there would obviously be no problem.  Here,
the commonality between two distinct implementations create an interface
and there's no interface copyright on that interface.

To give an example of the other extreme, consider a program which used
the command line interface plus the ptrace interface to interact with a
verbatim copy of gcc.  With a little thought you can do anything in this
fashion that you can do with relinking (and more).  Would you argue that
that's legal?

If not, and if the part of the program which implemented ptrace was
itself released under the GPL, would that make the derived compiler
legal under the GPL?

Allow me to repeat the definition of a computer program under us
copyright law:

    A ''computer program'' is a set of statements or instructions to be
    used directly or indirectly in a computer in order to bring about
    a certain result.

Do you see anything in there which says that if execve() is used then
the instructions it executes aren't part of the computer program?

So yeah, for the purposes of copyright law, it's reasonable to consider
that all programs called by a shell script are a part of that computer
program.  And, yeah, this can have interesting implications if you're
trying to distribute this computer program.  

Then again, if you don't have the right to distribute all the pieces of
a program why would you be distributing it?  Mostly this is a non-issue...

-- 
Raul


Reply to: