Comments by
Nick Lothian
Almost every dataset is going to need custom interpretation, and usually the time saved from using standardized metadata is quite low compared to the time spent analyzing it.
The Semantic Web is a great idea, but the costs associated with transforming data from whatever form it is in into RDF triples should never be used as an excuse to avoid making it available.
I believe that data should be made available as quickly as possible in whatever form in convenient, provided is is readable by freely available tools.
RDF triples, Custom XML dialects, CSV files and raw database dumps are all acceptable.
The value is in the data, and other people can easily convert it to other formats if needed.
One set of data which is becoming increasingly important is geo-spatial data.
Cost recovery should not be allowed as an excuse. Government data is paid for by taxes and so should be freely available.