[cc-devel] liblicense: named constant URIs
Nathan R. Yergler
nathan at yergler.net
Thu Sep 13 14:06:25 EDT 2007
On 9/13/07, Peter Miller <millerp at canb.auug.org.au> wrote:
> On Wed, 2007-09-12 at 21:49 -0700, Jon Phillips wrote:
> > Cool...Peter, what are you thinking now?
> I think I'll commit the patch as is. It rounds out the API nicely, and
> doesn't seem to get anybody's dander up.
> At the risk of confusing things with more than one issue per email, here
> are the immediate things I'd like to do:
> (1) in liblicense
> I've been looking at implementing my "license intersection" function.
> It seems straight forward, now that I'm more familiar with the code.
> But this raises a problem: what if the intersection is the empty set?
> I.e. no rights at all. Is there scope for come kind of "on the fly"
> license URI that can be queried via the API, etc, but never actually
> exists as a URL you could view with a web browser, as you can for CC
> attributes URIs and rights URIs. (Probably the same thing: what if the
> intersection of the permitted rights is valid but isn't a CC license
This is to determine how you could re-use two or more works together, right?
If the intersection is the empty set, then the works can't be used
together. I think we just need to notify the user (raise an
Note that even if it's not the empty set, there may still be
problematic semantics. For example, the Share-Alike requirement has
specific semantics beyond just having to exist in both licenses (ie,
you can't remix a BY-SA 1.0 work w/ BY-SA 3.0 -- this is a
braindead-ness of the 1.0 license that was later corrected).
> (2) in icon-compositor
> In trying to be a conscientious web citizen, I want to be able to
> display copyright notices with icons, as well as the license.
> But this raises a problem: there is no place to host copyright
> information within liblicense. This concerns me because I could wind up
> with not one but *two* files alongside each image: the one for the
> license information, and the one for the copyright information.
Why? The sidecar is XMP (which is essentially RDF) and can handle all
sorts of assertions, including copyright holder, source work
(dc:source, IIRC). Is the problem that there's no way to add
arbitrary assertions with the liblicense API?
> It gets worse: when I need to say "(C) me, derived from a work (C) Fred
> Bloggs" how do I? We don't have this problem in source code, we just
> add our notice to the list in the header comment, right next to the
> license comment. They belong together in source code - shouldn't they
> belong together in images?
A dc:source assertion can be used to refer to the URI of the source
work; this is what http://creativecommons.org/license does if you
provide the additional information.
> One more thing: I feel that the CC site (http://creativecommons.org/ns)
> is very unclear on the difference between reproduction and distribution.
> Is distribution == reproduction * N, where n > 0 or n > 1 or n > 42
> or ...? Surely I can make as many copies on my laptop as I like, or is
> there a problem there, too? Is distribution different than
> reproduction? Is distribution different than publication? I.e. giving
> a copy to someone vs making available to world on web site (the
> difference between personally handing out fliers vs leaving them on a
> table for people to pick up as they walk past). I've been wrestling
> with copyright for too many decades to be happy with the nebulous
> wording at the CC web site (and, yes, I get it that the answer varies by
> jurisdiction, I just feel the words could be much much less vague).
> Peter Miller <millerp at canb.auug.org.au>
> /\/\* http://miller.emu.id.au/pmiller/
> PGP public key ID: 1024D/D0EDB64D
> fingerprint = AD0A C5DF C426 4F03 5D53 2BDB 18D8 A4E2 D0ED B64D
> See http://www.keyserver.net or any PGP keyserver for public key.
> "Once you change how you think it makes sense." -- Final Cut Pro easter egg
> cc-devel mailing list
> cc-devel at lists.ibiblio.org
More information about the cc-devel