Clarification of `character encoding` definition

[jakubklimek]

Problem statement In GeoDCAT-AP 2.0.0, the usage note of character encoding refers to the whole instance of the class on which it is used.

• For Catalogue Record:

  This property SHOULD be used to specify the character encoding of the Catalogue Record

• For Distribution:

  This property SHOULD be used to specify the character encoding of the Distribution

This is confusing, as it actually refers to:

• Catalogue Record: the textual metadata properties used on the Dataset linked using foaf:primaryTopic
• Distribution: the textual content of the downloadable file linked using dcat:downloadURL or findable using dcat:accessURL, or in the output of the data service linked using dcat:accessService

Proposal Follow the pattern similar to the definition of CatalogueRecord.language in DCAT-AP saying

A language used in the textual metadata describing titles, descriptions, etc. of the Catalogued Resource.

and change the definition to:

• Catalogue Record:

  A character encoding used in the textual metadata describing titles, descriptions, etc. of the Catalogued Resource.

• Distribution:

  A character encoding used in the downloadable file or output of the data service represented by the Distribution.

This may have also impact on the Distribution.language definition in DCAT-AP as it currently says simply

A language used in the Distribution.

[jakubklimek]

Based on the discussion in the webinar, Distribution usage note will be changed to

A character encoding used in the downloadable file or output of the data service represented by the Distribution.

There was a counterproposal for CatalogueRecord formulated: Drop character encoding from CatalogueRecord, as in RDF, UTF-8 is always used for all data (which is not the case for XML). It was approved.