[tei-council] facsimile markup

Lou's Laptop lou.burnard at oucs.ox.ac.uk
Wed Sep 12 13:24:54 EDT 2007


Unless it's gone astray, I haven't heard anything from Conal about the 
definition of <surface>. This is a bit worrying, because the more I 
think about it, the more I think the current definition is wrong (and 
he's right!)

The present "surface" identifies some graphic on which we place a grid. 
With reference *to that grid* we then identify zones. The <surface> 
element is however also used to identify groups of graphics which are, 
in some sense, alternatives.

Now as Sebastian has repeatedly pointed out, there are very many cases 
in which an "alternative" graphic cannot  reasonably use the same grid.

So, it looks as if we do need two elements: one to say "these graphics 
relate to the same thing in the real world"; one to say "this graphic 
(representing that thing) has this grid".

So here is yet another last minute straw person proposal, with apologies 
for late arrival:

1. Redefine <surface> to mean "container for one or more images  of an 
inscribed surface", with a content model of surfaceImage+

2. Define <surfaceImage> with attributes @grid and @url. The latter 
points to the graphic, the former defines the grid. It contains any 
<zone> elements defined wrt this image.

If you want to show that a zone of one surfaceImage relates to a zone of 
another one... hm, you;re stuck. You probably need @corresp for that.

Must dash...









More information about the tei-council mailing list