Re: gnuplot not license-compatible with GPL software

[Ethan M. <merritt@u.washington.edu>]

On Thursday 29 July 2004 05:25 am, Roland Stigge wrote:
> Hi,
> 
> while I understand that the name gnuplot is independent from the GNU
> project or the FSF, I find it annoying that through a different license,
> I'm not allowed to link this software with GPL programs, most
> importantly, with GNU readline.

To the best of my knowledge, Debian is the only entity that interprets
the respective licenses in such a way as to avoid using the gnu readline
library with gnuplot.   I have argued in the past that this is a mistaken
interpretation of the GPL; nobody at that end seems to care.
Yeah, I know - I am not a lawyer.  But that doesn't mean I'm wrong :-)

In a nutshell:  
gnuplot is not a derivative work of libreadline.
The gnuplot source code does not contain any part of gnu libreadline.
gnuplot provides its own readline function, and can be used totally
independent of gnu libreadline.

The question would be moot if gnu libreadline were LGPL rather
than GPL, but that is an issue you would have to take up with them,
not with us.

In any event, Debian's refusal to package up a pre-built version
of gnuplot that links to a shared libreadline does not in any way stop you
from building your own copy linked against whatever libaries you like.
The GPL does not limit your own use of the code; it only imposes
restrictions on the conditions under which you may distribute the
result.

So download the source, configure with
	./configure --with-readline=gnu
and be happy

> Please consider issuing gnuplot under the GPL.

It is my personal opinion that the GPL is not a good license for
academic or scientific software.  My opinion matters very little,
because I am only one of many people who have contributed to
gnuplot over the years.  But if you multiply my reservations by
50 or so, to account for the larger group of contributers to the
project, you will begin to see why changing the license would
be difficult.

> Thank you very much!
> 
> Roland Stigge

-- 
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195