Re: New ways to evade copyright law (was Re: Vicarious liblity (was: KDE not in Debian?))
Raul Miller wrote:
> On Fri, Feb 04, 2000 at 01:30:50PM -0500, Brian Ristuccia wrote:
> > Distributing two separate works for which you have authorization, on
> > the other hand, is perfectly ok even if you don't have permission
> > to distribute the combination of the two. Any combination of those
> > works would be done by the end user. Only if the end user chooses to
> > distribute the result would they (not you) be in violation of the
> > copyright.
>
> But the only reason you have permission to distribute a GPLed executable
> is if you're distributing the source under appropriate terms (terms that
> let you modify and redistribute that source code with no restrictions
> beyond those imposed by the GPL). And that includes the source for any
> needed modules.
^^^^^^^^^^^^^^
No, as I recently observed in another post, that is not what the GPL says. It
says: "all the source code for all modules [the executabe work being
distributed] contains". To contain is something different from to need. While
statically and dynamically linked executables eventually will need the same
set of modules, they don't contain the same set. I agree that what the GPL
seems to want to say is the modules needed, and that the FSF gives it this
interpretation, but in fact the GPL doesn't say it. Now I am not advocating to
use this weakness of the GPL to ignore the problem that QPL does not permit
what GPL requires; however other distributors will (and do) do just that, and
if any GPL author would ever try to press charges against them, those
distributors will have no difficulty pointing out this subtlety in court.
(Alternatively they could argue that Qt is normally distributed with the
kernel of the operating system, and not with the GPL-ed appliction, which
could well be true in their case.)
What I do not understand is that when I make an attempt to suggest a stronger
definition, specifically
> For object code or other kinds of executable work, complete source code
> means the full source text for all executable code that will be executed as
> a direct consequence of executing the work, and whose presence on the same
> computer as the work is a prerequisite for proper execution of the work;
you just dismiss that by saying I've just defined the executable code as the
source code. (Where did I define executable code? I'll gladly admit my
formulation is not perfect and maybe I've goofed up completely by overlooking
something, but I did try to define "complete source code" as all sources
needed for a working program.) So what is your position? On the one hand your
analysis of the situation hinges on an interpretation of one clause of the
GPL, but you object to changing that clause to more unambiguously support that
interpretation. This leaves me a bit puzzled.
Marc van Leeuwen
Reply to: