Walter Lewis writes:
> "Peter C. Gorman" wrote:
> > We're producing an e-facsimile edition of a set of surveyors'
> > notebooks, in which individual notebooks will comprise <tei.2>
> > elements and entry sections will be represented as <div1>s. This
> > project is somewhat unusual in that the main access points will be
> > geographic areas rather than words in the text.
> > Has anyone else applied item-level GIS metadata to TEI texts?
> I had rather hoped someone else would take a stab at this (perhaps some
> did off list).
I've got a similar request pending with my geographers. We want to
include locational data in the CELT texts where we identify placenames
in transcriptions of early Irish texts. From the users' point of view,
latitude and longitude are the most obvious candidates, because they
can be looked up manually in any atlas by the user. A GIS server can
convert them to national grid refs, GPRS, or whatever other coordinate
system is needed for the GIS to serve back a map with a dot on it,
which is our prime requirement.
We have looked at implementing it with an extra level of indirection,
using the KEY attribute to point to a database of placenames where the
lat and long would be stored, but it would certainly be easier for the
user (and more portable) if these were dereferenced into LAT and LONG
attributes, or even a LOC attribute as "4.20;56.6" or something
I'm waiting to hear back from the GIS people on this.