Mr Hickson, I understand now, things are to be tested in the real world. I hope you judge WebDevData a reliable source. I ran some tests on latest WDD dataset (January 2015)
All list items below have been searched as value for meta@name dc.date.issued (along with possible variants dct.date.issued and dcterms.date.issued) 0 occurrances (dc.date.issued does not exist in DC spec - probably a copypaste error in the table) dcterms.collection (along with possible variants dc.collection and dct.collection) 0 occurrances (DC spec also defines some details for its properties, e.g. classes, and "collection" belongs to this category; it isn't a real property to be used in <meta> elements) dc.(date(Copyrighted|Submitted)|license|m(edi(ator|um)|odified)|provenance|references|temporal|valid) 0 occurrances (DC spec defines the properties above as dcterms.xxxx, not dc.xxxx) dcterms.(has(Format|Part|Version)|is((Format|Part|Version)Of|Re(ferenc|placed|quir)edBy)|re((ferenc|plac|quir)es|lation)|s(ource|ubject)) 0 occurrances (DC spec states that the above properties are to be specified as resources, rather than literals) There are no incorrect uses in the wild of these properties, at least in the group of pages analysed by WebDevData. Probably some authors wanted to reflect the whole DC spec in order to standardise his/her favourite subset of DC metadata, and in doing so s/he made some mistake in copypasting it. Can the table be fixed accordingly? Yours respectfully, Andrea
