[Trac] [OpenMath] #38: Incorporate Dublin Core and Creative Commons instead of idiosyncratic metadata

[OpenMath]

#38: Incorporate Dublin Core and Creative Commons instead of idiosyncratic
metadata
------------------------------+---------------------------------------------
     Reporter:  clange        |          Owner:  kohlhase  
         Type:  proposal      |         Status:  new       
     Priority:  major         |      Milestone:            
    Component:  OM3 Standard  |        Version:            
   Resolution:                |       Keywords:            
Include_gantt:  0             |   Dependencies:            
   Due_assign:  YYYY/MM/DD    |      Due_close:  YYYY/MM/DD
------------------------------+---------------------------------------------
Changes (by clange):

  * priority:  blocker => major

Old description:

> This is a rephrased version of proposal #12. I think many of these
> problems can be solved by getting rid of some (not all) of the
> idiosyncratic OpenMath CD metadata elements and replacing them by Dublin
> Core, Creative Commons, and other well-known metadata ontologies. One
> could either do this by statically importing certain XML schemas into the
> OpenMath XML schema, as e.g. OMDoc 1.2 does, too. That would give us
> elements like
> {{{
> <dc:author>Name</dc:author>
> }}}
> Several existing elements of the CD language could be abolished or
> deprecated in favour of this:
> * CDGroupName maps to dc:identifier
> * CDGroupDescription maps to dc:description
> * CDName maps to dc:identifier (see comment below for a discussion)
> * Description maps, quite naturally, to dc:description
> * CDDate maps, naturally, to dc:date
> * For CDComment, there is nothing in Dublin Core, but rdfs:comment from
> the widely used RDF Vocabulary Description Language (RDFS) could be used.
> * CDDefinition/Name maps to dc:identifier (see comment)
> * CDSignatures/Name maps to dc:identifier
> See http://dublincore.org/documents/dcmi-terms/ for the specification of
> Dublin Core.

New description:

 This is a rephrased version of proposal #12. I think many of these
 problems can be solved by getting rid of some (not all) of the
 idiosyncratic OpenMath CD metadata elements and replacing them by Dublin
 Core, Creative Commons, and other well-known metadata ontologies. One
 could either do this by statically importing certain XML schemas into the
 OpenMath XML schema, as e.g. OMDoc 1.2 does, too. That would give us
 elements like
 {{{
 <dc:author>Name</dc:author>
 }}}
 Several existing elements of the CD language could be abolished or
 deprecated in favour of this:
  * CDGroupName maps to dc:identifier
  * CDGroupDescription maps to dc:description
  * CDName maps to dc:identifier (see comment below for a discussion)
  * Description maps, quite naturally, to dc:description
  * CDDate maps, naturally, to dc:date
  * For CDComment, there is nothing in Dublin Core, but rdfs:comment from
 the widely used RDF Vocabulary Description Language (RDFS) could be used.
 * CDDefinition/Name maps to dc:identifier (see comment)
 * CDSignatures/Name maps to dc:identifier
 See http://dublincore.org/documents/dcmi-terms/ for the specification of
 Dublin Core.

Comment:

 To shortly recapitulate the discussion about CDName or Name vs.
 dc:identifier (2008/02/26 on the OM3 list):

  * I (Christoph) proposed this.
  * James had general concerns
  * I said that according to the [http://dublincore.org/documents/dcmi-
 terms/#terms-identifier Dublin Core spec], such an identifier is meant to
 be "unambiguous within a given context"
  * Paul said that actually CDBase + CDName is the identifier
  * Michael agreed with Paul

 … and then I didn't reply anymore and forgot the issue. I'm sorry for
 that.

 So let me now reply: I'm still in favour of dc:identifier, because
  1. the CDBase has to be given anyway. (CDBase is, BTW, a metadata field
 for which there is no obvious Dublin Core or other replacement.)
  2. dc:identifier is not meant to be the same as a URI. In fact, a URI for
 some OpenMath CD element with a name would be constructed as CDBase +
 CDName, same as the URI of a symbol is constructed as CDBase + CDName +
 Name. Still, in the local scope of a CD, the unambiguous identifier of a
 symbol is just its Name.

-- 
Ticket URL: <https://trac.mathweb.org/OM3/ticket/38#comment:1>
OpenMath <http://www.openmath.org>
The development of the OpenMath Standard and Content Dictionaries.