TEI Stand-Off Markup WG 2003-06-27 Conference Call Notes

Agreed that minutes were correct — we had agreed to use 3rd description mechanism from last week.

DD has read FV's latest (posted < 30 min ago), thinks it is in general OK. Needs more discussion of why XInclude over other standards or a TEI-grown solution. FV points out that being a W3C standard is the main reason, DD concurs.

On Entities

CC kept all of the NDATA prose from original P4 chapter in his draft. While it all should still work in XML 1.0, it is very foreign to people and requires at least some DTD-syntax declarations if not an entire DTD and a validating parser. DD recommends the <figure> element have an href or xptr attribute whose value is an XPointer instead.

While there are advantages to the SGML method (being able to declare the notation), there are also advantages to the more modern web methods (like content negotiation), including that it is far more popular.

CC asks about other uses of ENTITY in TEI. SB reports there are only 4 uses: <figure> , <xptr> & <xref> , WSDs, and FSDs. We are now pretty settled on recommending the elimination of the use of ENTITY attributes for the first two purposes, it therefore probably does not make sense to use them for the other two, either. Some brief discussion, without resolution, as to whether that question would come under the purview of CE and FS work-groups or under this work-group.

[?...?]

Question of including some of the SVG meta-data as TEI elements, rather than always relying on pointing to an SVG document (or including an SVG namespace element). CC believes that just a few simple examples would be sufficient, no need to suck in elements.

FV had some thoughts about this but his connection was intermittent enough that we never heard what he had to say. FV will post his thoughts shortly.

Question of whether to continue practice of graphic always being external, or to permit in-line graphics (when parseable, e.g., SVG), and what would in-line look like. DD thinks that if in-line it (the <svg> ) should be wrapped by a TEI-element (e.g. <figure> ).

Question of how to handle the simple case of merely wishing to encode, say, the captions; and of what is currently encoded as <figDesc> . CC will either work up (or find what he's already done :-) some simple examples of these in SVG so we can compare.

The lack of attendance at the last few calls, particularly this one, is disappointing to say the least.