Skip to Content.
Sympa Menu

cc-licenses - Re: Can't find the right license

cc-licenses AT lists.ibiblio.org

Subject: Development of Creative Commons licenses

List archive

Chronological Thread  
  • From: pitrp AT wg78.de (Peter Prohaska)
  • To: cc-licenses AT lists.ibiblio.org
  • Subject: Re: Can't find the right license
  • Date: Fri, 29 Oct 2004 15:57:26 +0200

On Thu, Oct 28, 2004 at 04:26:44PM -0400, Evan Prodromou wrote:
> On Thu, 2004-28-10 at 21:15 +0100, Rob Myers wrote:
>
> > > But the fact is that to be useful works should really be
> > > self-contained, including all source and all license information. It
> > > makes life in 5, 10, or 50 years much easier (and actually, probably
> > > in 6 months...).
> >
> > I agree absolutely, and so do the FSF. They recommend including the CC
> > license with the work.
>
> Do they provide examples of how to do that?

Interesting point. I was wondering what self-contained is, though.

1. The case of software source code:
Commen practice is to package a set of files that can be compiled
into an executable or library of some sort. The license can easily be
included in the toplevel directory. It seems that self-contained
means that the toplevel directory has to be relocatable without
loosing information in this case. T.i. it may not depend/link to
anything in a parent directory of the toplevel one.
If i wanted to distribute only the documentation inculded in a
subdirectory f.i., i would either have to copy the license to that
directory and create a new package or i have to keep the directory
structure in place and remove unwanted content.

2. The case of a wiki:
It seems reasonable to include a page with the license in the wiki and
reference this one when needed. If our wiki is located at
http://example.com/wiki, then it might be self-contained in
"path component space" if i can relocate the "/wiki" part of
the address to some other domain. From the point of the server's
directory structure, it might be self-contained only if i copy
files from various directories into a virtual root directory.
When i want to create a derivative of one page in the wiki and
publish it on a diffenert website, i have to copy the license page as
well. If I repeat this process, I would duplicate the license. It
seems more apropriate to think of creating a derivative of the wiki
and to copy the license only once.

3. A collection of samples:
Could be like the source code example if i provide them for download
in a package format but it could also be like the wiki example if I
provide them for download on a per sample basis.
Probably there would be a license file/page and an identifyer for the
license in every sample file. I'd probably prefer to put a URI to a
license page into the sample instead of a reference to some "toplevel
COPYING" file because it is more likely that the COPYING file
accidentially gets lost during reuse than CC going down. Including
the complete license in every 2 second sample is complete overkill of
course.

The interesting question is when a license file counts as included in the
distribution. The fsf seems to think in terms of package-wise
distribution but i think a set of webpages (like a blog) is a
distribution on it's own. The problem is to decide where the root of a
work is (DNS root, unix/windows/mac os FS root, some other FS dir,...)
or more precicely how to identify the location of the license text.
Not to use a URI as id seems counter intuitive. It might be useful to
include the date/version of the referenced URI in a comment and maybe
to include a file with a name similar to the URI if you distribute
your work in a packaged format (in addition to some COPYING file).
The FSF favours a similar safetynet approach with the GPL, i think.
The copyright statement in a source file normally contains a note that
you should have received a copy of the GPL and that you can get one
under snail mail address XYZ if this is not the case.
Using a URI instead of a postal address might be or might not be more
relieable; that remains to be seen for CC. The problem is that addresses
of all kinds can change but organisations like CC, the FSF or OSI try to
provide stable addresses of various types.
Addresses change because humans change them and the more people depend
on a certain address the less likely someone will dare to change it.
Using a resonable URI reference instead of a COPYING file might
therefore be more shock prove against all sorts of errors.

Maybe it boils down to whom you trust most. Yourself making no packaging
error, someone reusing your work making no packaging error or some
organisation keeping a URI stable.

regards, peter.

Attachment: pgpatpngPEHse.pgp
Description: PGP signature




Archive powered by MHonArc 2.6.24.

Top of Page