On Tue, Dec 20, 2016 at 5:20 PM, Christian Chiarcos <
chiarcos@informatik.uni-frankfurt.de> wrote:
Deviating from this is literally tag abuse. Bad practice, though sometimes,
> it does happen -- and normally for a reason.
> However, if existing language tags really aren't sufficient, it is better
> *not* to use a (potentially misleading or ill-defined) language tag, but to
> represent language variety information of a resource abc:xyz with an
> explicit link to a repository such as Glottolog:
>
As I just wrote, using invalid tags is no longer permitted in RDF 1.1, so
it's worse practice than it used to be.
Since RDF refers to "language tag"s, it is possible to use a private use
tag ("x-dothraki"), it is known.
However, since the PlainLiteral datatype refers to the langtag production,
OWL implementations can't. However since the mapping from RDF to OWL says
you can, if it will work, it is known that it is not known.
Simon
Received on Tuesday, 20 December 2016 23:10:36 UTC