Re: Defining 'preferred form for making modifications'
Henning Makholm <henning@makholm.net> writes:
> The GPL'ed source contains ugly xpm's that upstream created pixel for
> pixel in Emacs because he knew no better and thought he was only
> making a proof-of-concept implementation anyway. I import the xpm into
> the Gimp, painstakingly separate the raw pixels into reasonable
> layers, then add nifty colors and drop shadows. Finally I merge the
> layers and quantisize the image, then save as xpm again.
>
> Will I be in violation of the GPL if I distribute it withough *also*
> saving it as xcf and distributing that? If no, would that change if it
> took me several editing sessions to get the look right, and I saved my
> intermediate work as xcf?
The format you preferred to modify the work in was as a layered
image. Is this not obvious, especially given the work you did in
creating just that layered image?
If you saved your intermediate work as xcf, and you didn't distribute
it, then you are simply someone who has either refused to distribute
source, or who has perhaps deliberately destroyed it.
If you never saved the xpf, then I am disinclined to think this is
ok. Consider the following scenario:
I write a bunch of Scheme code in a fancy Scheme system, never saving
my work, using only an editing buffer. When my program is as I like
it, I use the system's "standalone executable" feature to writeout a
binary of the program, and then I quit.
I do not believe the resulting binary can be distributed "with
complete source", as the GPL requires. I do not believe this is a
wild interpretation of the GPL; it seems exactly right to me. The
form I actually preferred, and everyone else, is the Scheme code, not
the binary.
Reply to: