W3C home > Mailing lists > Public > www-rdf-interest@w3.org > September 2001

Re: Bitzi File Metadata RDF Dump

From: Gordon Mohr <gojomo@bitzi.com>
Date: Tue, 25 Sep 2001 15:01:45 -0700
Message-ID: <007b01c1460d$b0cc9040$6601a8c0@gojovaio>
To: "Aaron Swartz" <aswartz@upclink.com>
Cc: <www-rdf-interest@w3.org>, "Mike Linksvayer" <ml@bitzi.com>
Hi, Aaron.

Thanks for the comments!

You write:
> Some comments on the RDF you used:
> 
> You seem to have created a new URI scheme for bitprints:
>    
> bitprint:3KIZIJB64XP3NCXAE4ISQZT3QNCTF7VDNK5UNR8ZPQ5MFASNGVB5MISV7ESUSB2MN5R3IY2
> 
> AFAIK, this URI scheme is unregistered. Why go to all this 
> trouble when these bitprints already have URIs? I think it'd be 
> much better to just use:
> 
> http://bitzi.com/lookup/3KIZIJB64XP3NCXAE4ISQZT3QNCTF7VDNK5UNR8ZPQ5MFASNGVB5MISV7ESUSB2MN5R3IY2
> 
> instead.

Well, that HTTP URL doesn't actually name or return the entity
that is being described -- that URL returns human-readable 
information about the entity. 

My current thinking is that the right approach will be something
like:

   urn:bitprint:3KIZIJB64XP3NCXAE4ISQZT3QNCTF7VDNK5UNR8ZPQ5MFASNGVB5MISV7ESUSB2MN5R3IY2

...with "bitprint" as a (yet-to-be) registered Namespace ID.

> A schema would be good, since many of the properties aren't 
> intuitively obvious (first20?). Let me know if you need help 
> with a schema.

Formal schemas and documentaton are planned for when the choice 
of what to include, and how to name it, settles a bit. 

Your assistance in suggesting and crafting schemas would be 
appreciated! Given the large number of files we'll be tracking, 
and wide variance in how much info is available per file, we 
may wind up with a number of independent schemas which can be 
"mixed-in" to each file's record, as appropriate.

> <bz:url>http://example.org</bz:url> would be better as 
> <rdfs:seeAlso rdf:resource="http://example.org/" />
> 
> (The RDFS namespace is at: http://www.w3.org/2000/01/rdf-schema#)
> 
> To avoid confusion, properties like bz:album should be called 
> bz:albumName (unless you start storing records for albums, like 
> MusicBrainz does, in which case you can give them URIs.)
> 
> <bz:society>Joe_Frank</bz:society> would be better as 
> <bz:society rdf:resource="http://bitzi.com/society/Joe_Frank" />

Good points, all.

> Otherwise, it looks pretty good. Pretty simple. I assume there 
> aren't any more properties in the large dataset than there are 
> in http://preview.openbits.org/data/openbits-20010917-sample.rdf

Actually, no. That's merely a representative audio file. I've
added a note about the other fields present in this first dump
at the http://preview.openbits.org page.

> Ouch, 404 on the namespace[1] And your 404 message isn't even 
> returned with a 404 status :-(
> 
> [1] http://bitzi.com/xmlns/2001/09/10/experimental#

As our formats and namespace names stabilize, they will settle
into being valid URLs, supplying schema or other documentation.
In the meantime, though, isn't it still true that there is no
formal requirement for namespace-names to have retrievable 
content?

(As far as not-founds returning 404, I know we've encountered
some oddities with our ACS installation in this matter in the
past. I'll check into what's up.)

- Gordon
____________________
Gordon Mohr, gojomo@
bitzi.com, Bitzi CTO
_ http://bitzi.com _
Received on Tuesday, 25 September 2001 17:56:11 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:51 GMT