[tei-council] facsimile diagram

Conal Tuohy Conal.Tuohy at vuw.ac.nz
Thu Aug 2 06:33:26 EDT 2007


G'day Lou!

I don't think we need recursion in order to have graphics that represent particularly interesting bits of the page, and I think that the extra complexity involved in processing recursive markup should therefore rule it out, to be frank.

To represent a detail image using the markup I'm proposing, you would simply add a <graphic> (as a child of a particular <surface>) whose coordinates would indicate that it covered just a small region.

In short, each graphic can be declared to cover any area of the surface it depicts, just as each zone can be declared to represent a region any particular region on that surface.

I also don't think that <zone> elements should contain <graphic> elements. I think the value of the original proposal is precisely in that it defines the graphics independently of the zones: the graphics and zones relate only via the surfaces (i.e. the physical pages) which those graphics and zones depict, and analyse, respectively. Hence both zones and graphics would be empty elements.

I think this makes sense if you consider that the function of the graphic elements is to depict the pages, while the function of the zone elements is to analyse the pages. A particular area of the page is of interest to a particular analysis independently of the graphics which depict that area of the page, and a particular graphic is a depiction of the page (or part of the page), independently of whether any part of that graphic is of analytical interest.

As far as linking textual elements (such as tei:p) to the stand-off analytical elements (tei:zone), I can see a few ways to do this, and I'll post something about this shortly. But to my mind it's important that we can reach a consensus on surfaces, graphics, and zones, too, so we are clear on what is being linked.

Con


-----Original Message-----
From: Lou Burnard [mailto:lou.burnard at computing-services.oxford.ac.uk]
Sent: Thu 02/08/07 21:18
To: Conal Tuohy
Cc: TEI Council
Subject: Re: [tei-council] facsimile diagram
 
I still prefer the idea of recursive surfaces

TEI -> header, facsimile? , text?

facsimile -> front?, (surface+ | model.graphicLike+), back?)

surface -> (model.graphicLike*, surface*)

This allows for the possibility that you might have additional images 
for particularly interesting bits of the page ("zone"s in your terminologuy)

However if you're really unhappy with the recursion, I can live with

surface -> (model.graphicLike*, zone*)
zone -> (model.graphicLike*)

The big headache for me is the attributes. The draft I am working on 
proposes a new global attribute @facs to point from any element to a 
corresponding surface (or zone) but we need more than that.

James reminded me that there is work in the MPEG/ISO/W3C communities on 
defining new Xpointer schemes to point into assorted kinds of multimedia 
which we should at least acknowledge somewhere.



Conal Tuohy wrote:
> I have uploaded a simple diagram showing the (graphical) relationships between the <surface>, <graphic>, and <zone> elements as used in the latest "facsimile" draft.
> 
> http://tei.oucs.ox.ac.uk/trac/TEIP5/attachment/ticket/291/fax-example.jpg
> 
> Please take a look at it to make sure we are all on the same page (pun intended) regarding the meaning of the proposed terms, since the proposed element names have changed as a result of last week's discussion.
> 
> The diagram shows 1 page with 1 associated graphic, and 4 zones:
> 
> A <surface> represents a physical page itself, and may contain <graphic> and <zone> elements. The example <surface> represents a recto page from a manuscript.
> 
> A <graphic> represents a digital image of that page. In the example the graphic covers an area slightly larger than the page itself (represented by the <surface> element). 
> 
> A <zone> represents a rectangular area on a page. In the example the zones show 4 areas of interest, which would be linked to pieces of transcript.
> 
> I will upload some other updated material after dinner (shortly)
> 
> Conal
> 
> 
> _______________________________________________
> tei-council mailing list
> tei-council at lists.village.Virginia.EDU
> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
> 






More information about the tei-council mailing list