[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: GNU FDL 1.2 draft comment summary posted, and RFD



On Wed, Jun 12, 2002 at 04:21:50PM -0700, Mark Rafn wrote:
> On Wed, 12 Jun 2002, Branden Robinson wrote:
> 
> > I sympathize with your concerns but I've having difficulty reconciling
> > '"Also, you can print this out and distribute it" would seem to weaken
> > the copyleft' with "Hi, your current license means I can't print out the
> > documentation and give people copies easily, consider this one".
> 
> True.  It's also an objection I didn't list in my 3.  The GPL makes it 
> hard to legally do ad-hoc distribution in non-source form.  Making 
> exceptions here doesn't bother me a lot, as printed text is easier to 
> reverse-engineer than an application. 

RMS has claimed -- though I don't know if this is an official FSF
position -- that he just plain doesn't care about ad-hoc distribution in
non-source form.  The discussion came up on this list when I opined that
it was absurd that I couldn't send a friend by /bin/ls so that he could
discern whether or not his was broken.  (/bin/ls is part of the GNU
fileutils and thus GPL'ed, so I can't distribute a compiled binary
without the source or the promise to deliver source on demand).

> It does bring to my mind an issue WRT the virality (aka "copyleftedness",
> aka protection from non-free forks).  Is there a way to keep this feature
> and not prohibit things like distributing a printed copy that uses a
> non-free font?

The issue of non-free fonts, when they're not part of the DFCL-licensed
work itself but rather applied to the work as part of a transformative
process, is probably too thorny for us to tackle in the DFCL.  In the
U.S., you can't claim copyright protection on a typeface, just on the
font software that generates the typeface.  However, this isn't the law
everywhere.

I do not think we want to DFCL to attempt to restrict people from
applying proprietary "transforms" to DFCL-licensed data, as long as the
DFCL content is recoverable or otherwise available.  Hmm.  This could
get really, really, thorny.  Consider:

1) We don't want to tell Professor Smith that he's in violation because
he printed out a DFCL-licensed document, but the font in his laser
printer is proprietary to, say, Hewlett-Packard (or Adobe).  He then
makes a dozen copies for his grad students and hands them out.  This
should be legitimate.

2) We don't want a music webcaster to take DFCL-licensed piece of music
out of "the commons" because he runs the music stream through some sort
of highly proprietary equalization/compression process before
transmitting it.  Such a transformation might not be reversible.  If
this is the most popular form of dissemination for that piece of music,
we have to be sure that this broadcaster is obligated to make the
original piece of music, as licensed under the DFCL, available.

> I'm with Branden here.  If it's distributed to that many folks, it should
> be made electronically available.  I hope that photocopies are used mostly
> because the original isn't available electronically, which won't be a
> problem for things under this license.  

Well...there's what easy for professors at big western universities, and
there's what's easy for Peace Corps volunteers in Africa (they are often
assigned to schoolteaching duties, especially in the English-speaking
countries).

It's fine to mandate that the professor supply a URL to the source-form
of the document when we're talking about Carnegie-Mellon.  It's not so
fine to mandate that when we're talking about a school that doesn't even
have a name in Ghana.

> I don't much like the 100-copy exemption, anyway.  Free documents should 
> stay free, and that pretty much means electronic.  On the edge of the 
> definition, though, a photocopy can become the preferred editing format.  
> If I use liquid paper and ink to change a document, then photocopy it, 
> aren't the images of the pages rather than the text the form I preferred 
> to edit?

Only if the original author wrote it that way, or if the only content
remaining in the document is maintained that way.

> Kind of.  We could dual-license things to have one license for some and
> another for everyone else that have different terms.  As long as both are
> free, we're technically ok.

Yes.

> But it's ugly, will cause confusion, and will probably not be used by many 
> (I know I'd recommend people just use one of the licenses).

Also true.

> Worse, it violates the spirit of the DSFG.  "No discrimination against
> fields of endeavor" means something.

I disagree here.  You could say that the GNU GPL discriminates against
people who don't distribute in source form only, or don't accompany
binaries with source.  However, because the rules that apply to
binary-only distributors are still Free according to all the *other*
clauses of the DFSG, we do not regard this as "discrimination against a
fiend of endeavor".

> And it's a wierd distinction.  You need to specify why you want educators
> to be able to make 1000 photocopies and NOT allow amazon.com to do so.  
> Why should students not have the ability to modify and distribute the 
> document?

I don't think anyone proposed that the copyleft would not propagate
forward to the students.  The issue is how much work the professor needs
to go to ensure that the students can exercise their rights under
copyleft.

-- 
G. Branden Robinson                |      Never underestimate the power of
Debian GNU/Linux                   |      human stupidity.
branden@debian.org                 |      -- Robert Heinlein
http://people.debian.org/~branden/ |

Attachment: pgpUPQXfQ5WTz.pgp
Description: PGP signature


Reply to: