On Thu, 7 Jun 2001, "Derek D. Martin" <[EMAIL PROTECTED]> wrote:
> Karl J. Runge said:
>
> No... you are using a utility that exists on the system. If the
> sysadmin hasn't installed it seperately (often through OS install),
> your exec will fail. Unless you distribute gzip with your
> software... then you might be getting yourself into some issues.
I knew using gzip was a bad example... replace with "arbitrary GPL'd program".
I guess you are saying if I tell the user to go get "arbitrary GPL'd
program" and install it so my proprietary program will work (my program
can't work without it) that is OK . I imagine it is, since that is a
user doing these acts "the private". Still, it seems there is some room
for fuzziness... (e.g. how "helpful" can I be their getting of "arbitray
GPL'd program"). Oh well, IANAL.
> AH! But that's not the same thing, is it? In the first case, you're
> exec-ing a seperate program. In this case, you'd be LINKING against a
> GPL'd library. That code runs in the same address space.
My point was zlib is apparently released under much less restriction,
evidently not even LGPL.
> > I am worried that the GPL does not spell out this fork+exec is explicitly
> > permitted, this is a crack that lawyers can insert a wedge and start
> > pounding...
>
> It spells out "in a shared address space." Once you exec(2) a
> different program, it gets its own, seperate, address space. Yes?
Where in the GPL does it say "shared address space"? The FAQ does, but
the GPL doesn't. The LGPL works toward a general definition of
"linking", but certainly is not that explicit about address space, etc. (I
guess there really can't be, because paradigms change over time...)
Karl
**********************************************************
To unsubscribe from this list, send mail to
[EMAIL PROTECTED] with the following text in the
*body* (*not* the subject line) of the letter:
unsubscribe gnhlug
**********************************************************
