Yes, there has been some recent discussion about that issue. There should actually be a language attribute for every element in response messages that are intended to be read by people. This includes data records that might be returned in a search response.
When provided, the lang attribute shall apply to all child elements of the element in which it is defined.
To provide backward compatibility with the existing code and interfaces, when a "lang" attribute is not specified, then the default should be assumed to be "en"
shouldn't there be a dataelement "language" for the "abstract" dataelement or for resource metadata in general?
Yes, there has been some recent discussion about that issue. There should actually be a language attribute for every element in response messages that are intended to be read by people. This includes data records that might be returned in a search response.
Might I suggest that we consider using the "lang" attribute as defined in the W3C document http://www.w3.org/TR/REC-html40/struct/dirlang.html?
When provided, the lang attribute shall apply to all child elements of the element in which it is defined.
To provide backward compatibility with the existing code and interfaces, when a "lang" attribute is not specified, then the default should be assumed to be "en"
see also
--------
Tags for the Identification of Languages: http://www.ietf.org/rfc/rfc1766.txt
"XPath lang() function": http://www.w3.org/TR/xpath#function-lang
Technical contents of ISO 639:1988 (E/F)
"Code for the representation of names of languages":
http://ftp.ics.uci.edu/pub/ietf/http/related/iso639.txt
I am cross posting this message to digir-dev so that broader comment is available.
To use the lang function, I think XPath requires lang to be qualified xml:lang. If I am right, I suggest that this be the digIR spec also.