[WF-General] License stuff. Fun!

David Turner novalis at novalis.org
Wed Apr 10 16:45:02 PDT 2002

On Wed, 2002-04-10 at 07:59, Karsten-O. Laux wrote:
> you wrote:
> >
> > I talked to Eben Moglen, the FSF's lawyer about this.  PNGs count as
> > transparent formats for images.  This is a different standard than the
> > GPL.  We (the FSF) are considering whether and how to change this for
> > future versons of the GFDL.
> >
> > Moglen says that one can take a layered image distributed under the GFDL
> > and release a PNG version without also distributing the layered
> > version.  As I see it, this is incentive to require GFDL 1.1 *or
> > greater*, since we may change this at some point.
> quoting the GFDL text:
> " A copy made in an otherwise Transparent file format whose markup has been
>   designed to thwart or discourage subsequent modification by readers is not
>   Transparent. A copy that is not "Transparent" is called "Opaque". "
> According to this phrase a png derived from a layered image file is an 
> "opaque copy" in the sense of the GFDL.

Here's a what the png spec says about its design goals (see "designed",

Notice that the proposed goal "m. Thwarting: To thwart subsequent
modification by readers" was rightly rejected by the png committee ;)

There was brief discussion of thwarting modifications to mtimes (but not
even really that), but that was decided against for technical and
philosophical reasons. 

> Reason: The process of combining the different original layers to a single 
> plane is definitely a way to 'thwart' any subsequent modifications of the 
> image content. Despite the fact that the visual information is the same, the 
> possibilities to edit the image are dramatically limited in comparison to the 
> original layered image. 
> The situation is comparable to TeX v. Postscript format. You may be able to 
> change a postscript file to fit your needs, but this is very hard, if not 
> impossible. So in my eyes the layered image is comparable to the LaTeX 
> source, while the combined image as png is comparable to the rendered 
> postscript file.

But postscript is clearly (raytracer hacks asside) an output-only
format, while people edit pngs all the time and don't complain about it.

What Moglen thought was really important was Uta's clear intent to
release under the GPL/GFDL.  To go back on this intent now is not only
in bad faith, but on bad legal grounds.  Whether taking layered xcf in
and giving only nonlayered png out would violate the GFDL or not (and it
would not), Uta's pngs were, from the perspective of the GFDL
distribution, the originals.  And they can be and have been modified,
edited, etc. 

-Dave Turner                                Stalk Me: 617 441 0668
"Somebody would have done it, if we hadn't done it." 
    -Laurence Canter, of Canter and Siegel, spammer.
"And where have I heard that before?" 
    -Mr. Meany (a demon), _Meadow Of The Damned_, by Scott McCloud

More information about the General mailing list