[tei-council] updated facsimile odd
James Cummings
James.Cummings at computing-services.oxford.ac.uk
Thu Aug 2 17:03:22 EDT 2007
Sebastian Rahtz wrote:
> What I do NOT see is how one <l> can link to
> coords 20,20,100,100 in the _surface_. Can it?
> It can only relate to a particular graphic?
I'm not entirely sure of the relationship in coordinates between the graphic and
the surface. Unless the unit of measurement is determined on the surface and
then all the coords are defaultly in that unit of measurement (and in that case
it must be a real-world physical unit, i.e. not pixels).
> If I have a hi-res image and a lo-res image,
> and describe the same are within each,
> the coords of each will be different. Correct?
I'm not sure. If the coords is on graphic, in units of measurement referring to
the surface, then they both cover the same distance. (One simply does it with
greater resolution in its reproduction of that surface.) If coords is on a zone,
then it makes sense that it is in relation to whatever unit of measurement is
use for the graphic (presumably pixels).
> When I have one <surface>, with three <graphic>,
> each with a <zone> pointing at the original of my <l>,
> then the <l> points at all three?
I think then the l/@corresp (or specialised attribute) could point to all three
zones, and that these zones would have different coords in relationship to the
graphic which contains them. However, it is equally possible for there just to
be a l/@xml:id which then each of the zones point to. (And in fact, that way
round of zones pointing to lines stand-offishly rather than the reverse would
naturally be my preference, I believe.)
I'm still a bit fuzzy on the details of the relationship between surface and
graphic and would like more clarification from Conal and others how they see
that working. Maybe the examples being reworked as Conal suggested would help.
-James
--
Dr James Cummings, Oxford Text Archive, University of Oxford
James dot Cummings at oucs dot ox dot ac dot uk
More information about the tei-council
mailing list