[cc-community] How Does Google Index CC-friendly MP3s?

David Delbridge creativecommons at rigel7.com
Sat Feb 28 14:21:38 EST 2009


Thank you, Mike.  This is very helpful.

So, what is the point of SHA1 hashes (described in both 
http://www.w3.org/Submission/2008/SUBM-ccREL-20080501/ and 
http://wiki.creativecommons.org/Nonweb_Tagging)?  I now understand how 
MAGNET links utilize the hash to locate a file by name on the Gnutella 
network, for example.  But then, as I understand it, CC suggests that an 
SHA1 hash is useful for verifying an audio file's (or any file's) 
authenticity.  [The Embedded Metadata example 
(http://wiki.creativecommons.org/Embedded_Metadata) also speaks of a 
"verification" page, but (for simplicity, I presume) breezes over the 
RDF and presents no screenshot of the verification page.  Would a hash 
appear on this page in human-readable form?  In the RDF?]

In other words, am I correct in presuming that there is a reason to 
display, in human-readable form, the MP3 audio file's SHA1 hash on its 
official verification page?  If not, and the hash is useful only in 
machine-readable form, then how might this hash be used by 
third-parties?  What is the intended application?  Indexing?

Again, thank you very much for your kind assistance.

Mike Linksvayer wrote:
> On Fri, Feb 27, 2009 at 6:54 PM, David Delbridge
> <creativecommons at rigel7.com> wrote:
>   
>> I'm publishing MP3s on my website for indexing by CC-friendly search
>> engines.  How does this work?  Neither Google nor Yahoo! describes
>> clearly what must be done to get an MP3 audio file CC-indexed.  Does
>> Google, for example, read the MP3 audio file's ID3 frames?  Or, does
>> Google index the URN:SHA1 tag (RDFa) embedded in your MP3 audio file's
>> official download page?
>>     
>
> No search engine that I know of does CC aware indexing of audio. I
> suspect the first one that does will look at descriptions of http URLs
> returning audio files rather than URNs with hashes of the files. The
> latter is primarily intended for use by P2P clients to verify that the
> URL they have a reference to is talking about the file with the
> reference ... turned out to not be a big use case.
>
>   
>> I've studied the example RDFa code fragment on
>> http://wiki.creativecommons.org/Nonweb_Tagging but am confused by the
>> clarifying statement, "This fragment includes the statement
>> <urn:sha1:MSMBC5VE...."  Huh?!  No it doesn't!  Is there a typo here or
>> am I misunderstanding something?
>>     
>
> It could be explained much better.  Both have the same statement
> (subject, predicate, object), first as RDFa, second as something
> N-Triples like.  The page needs to be rewritten for clarity.
>
> See http://www.w3.org/Submission/2008/SUBM-ccREL-20080501/ for a
> longer and hopefully clearer exposition.  For a music site, I'd to
> start with recommend simply something like
>
> <div about="http://example.com/foo.mp3">
> <a href="http://example.com/foo.mp3">foo.mp3</a>
> <a href="http://creativecommons.org/licenses/by/3.0/" rel="license">CC BY</a>
> </div>
>
>   

-- 

David M. Delbridge
Circa 3000
ColdFusion Hosting
http://www.circa3k.com
866-CIRCA3K (247-2235)
Outside U.S: +1.775-832-2445



More information about the cc-community mailing list