I have a related concern, and I am afraid that I have not made it clear.
Let me try again.
In another mail, you wrote:
> It is trivial (using any sort of search tool) to identify those parts of
> the document which are CML information components. The prefix/ns mechanism
> does this even if the prefix is not 'CML'.
Suppose an article consists of hypertext and metadata. The hypertext
portion contains the metadata as a subordinate structure. If we allow
xml:lang to be inherited from the hypertext to the metadata, programs (e.g.,
search enginges) that are interested in only metadata have to examine
the hypertext portion as well. Yes, this is not a big deal, but
this is undesirable. Since the metadata is very likely to be
created by some software that concentrates on metadata, inheritance of
xml:lang from the hypertext portion is unlikely to be exploited.
Makoto
Fuji Xerox Information Systems
Tel: +81-44-812-7230 Fax: +81-44-812-7231
E-mail: murata@apsdc.ksp.fujixerox.co.jp