Re: for those who want "free only" buttons

On Wednesday 14 February 2007 06:32 pm, Mike Linksvayer
wrote:
> On Wed, 2007-02-14 at 07:20 -0500, drew Roberts wrote:
> > On Wednesday 14 February 2007 12:20 am, Mike
Linksvayer wrote:
> > > So the current best practice (and usable by
Google, Yahoo!, MozCC) is
> > > to annotate the visible link with
rel="license" (denoting a licensing
> > > relationship; we don't just happen to be
linking to a URL that happens
> > > to be a license), which you can think of as
the rel-license microformat
> > > or producing a "<> :license
<license uri> ." triple if you prefer RDF.
> >
> > Mike, could you give a little clear write up on
what current best
> > practice is? I have read the recent relevant posts
but think I must not
> > have understood.
> >
> > How dod the search engines identify licensed
content? By links back to
> > the CC license pages?
>
> More or less, qualified with rel="license"
>
> > So if a site is licensed BY-SA and points to BY-NC
songs, and a person
> > searches for BY-SA music, wil the engines return
this page as a result? I
> > run into issues like that often.
>
> Yes, this is a problem with inexact metadata and using
search engines
> that index at the level of web pages only.

Yep, what a pain.
>
> On the metadata front, metadata for an object (e.g.,
mp3) that is not a
> web page can be specified with RDFa -- see
> http://wiki.crea
tivecommons.org/RDFa and for an example, view source on
> htt
p://ccmixter.org/media/view/media/remix/latest or view
the page with
> Songbird with MozCC installed.  There's also a
possibility that someday
> a "media microformat" will be defined which
includes support for
> licensing, see http://microfo
rmats.org/wiki/licensing for some high
> level discussion, most of which is not actually
microformat-specific.
> License info for specific objects may also be conveyed
in feeds, see
> http://wi
ki.creativecommons.org/Syndication
>
> Buf for search to work in the way that you want there
needs to be a
> consumer for any of these methods of annotation.  AFAIK
http://owlmm.com
> is the only one so far (importing music to be searched
via feeds).  I'm
> reasonably confident things will get much better in
this vein this year.
> The other searches available via http://search.creat
ivecommons.org and
> documented at http://wiki.creativecommons.org/CcSearch_implementations

> are doing web page or internal database searches.

So, giventhis, as things stand now, it might be best for
people to list media 
of various licenses all on seperate pages with the page
using the same 
license. Then, if needed, license the page itself with a
dual license but on 
a seperate page. Would that help today with how the search
engines work?

all the best,

drew
-- 
(da idea man)
_______________________________________________
cc-community mailing list
cc-communitylists.ibiblio.org
http://lists.ibiblio.org/mailman/listinfo/cc-community

On Wed, 2007-02-14 at 18:49 -0500, drew Roberts wrote:
> So, giventhis, as things stand now, it might be best
for people to list media 
> of various licenses all on seperate pages with the page
using the same 
> license. Then, if needed, license the page itself with
a dual license but on 
> a seperate page. Would that help today with how the
search engines work?

It would help given the way Google and Yahoo! web search for
CC work
currently, but I think the need to get most sites publishing
CC licensed
music to reorganize their sites around this scheme for it to
be
effective makes it a nonstarter.

Deployment of more precise metadata in pages and feeds and
implementation of media-specific CC-enabled search seems
more promising
to me.

That said, feel free to reorganize any sites you control or
influence
*and* add precise metadata to their pages and feeds.

-- 
  
http://wiki.creativecommons.org/User:Mike_Linksvayer

_______________________________________________
cc-community mailing list
cc-communitylists.ibiblio.org
http://lists.ibiblio.org/mailman/listinfo/cc-community