[rfc-i] Metadata

Dave Crocker dhc at dcrocker.net
Tue Sep 25 12:57:59 PDT 2012

On 9/25/2012 12:31 PM, Paul Hoffman wrote:
> On Sep 25, 2012, at 12:10 PM, Dave Crocker <dcrocker at bbiw.net>
> wrote:
>> This excludes metadata that is added to the document from other
>> sources.
> Just to be clear, metadata is no *added to* a document, it is data
> that is *about* a document.

You think information that structures a document or directs its
presentation is not "part of" the document?

I think I understand the theoretical basis for such a view, but
typically find that practical uses of the term work far better with a 
simpler model in which all of it together is 'the' document.

Especially for that 99% you cite...

> * Metadata: information about a document that is often derived from,
> or extracted from, the document. Metadata can be collected manually,
> with tools that make guesses based on document structure, by markings
> in the document, or from the history of the production or publication
> of a document.
> Having said that, in line with the other threads this morning about
> getting rid of the "Currently" stuff, I'd be happy to drop the second
> sentence above.

Unfortunately that leaves the definition as wholly conceptual.  Examples 
anchor it.


    Metadata:  Information associated with a document, such as defining 
its structure, presentation, topic or author.

> Regardless, I think using the Wikipedia definition ("The term
> metadata is ambiguous, as it is used for two fundamentally different
> concepts (types)...") is not a good thing for this document.
> --Paul Hoffman

  Dave Crocker
  Brandenburg InternetWorking

  Dave Crocker
  Brandenburg InternetWorking

More information about the rfc-interest mailing list