On 07/04/11 06:00, Piotr Bański wrote:
> a. xml:lang is used for language identification of element content;
> b. it can be (mis?)placed on e.g. <ptr> elements and empty elements in
> general when they perform a stand-off function;
> c. it is (mis)used for element identification.
I would like to argue very strongly that there are times when (c) is
incredibly useful for processing parallel texts of various kinds.
We have six separate editions of the Treaty of Waitangi in our
collection, each of them different (See
http://www.nzetc.org/tm/scholarly/name-122436.html ). Improving our
automatically generated browse structures between such works is on my
todo list, but seems essentially impossible if I'm not allowed to use
xml:lang to separate the English and Maori versions (I believe that
there may also be a French version, but we don't have a copy).
I would separately like to argue that the TEI community has a role to
lead the XML community (i.e. the w3c) on some issues. I notice a number
of places (such as http://www.w3.org/TR/xml11/#sec-lang-tag ) the w3c
seems to be using TEI-derived examples, which suggests that this may
have happened in the past.
Library Technology Services http://www.victoria.ac.nz/library/