Keyref matching text

Hi,

Sending this out as promised, to get it on the TC agenda for next week.

It's been widely recognized that the rules for "keyref matching text" --
that is, using keyref in a way that pulls in text when processed -- needs
to be clarified in the specification. Many core rules for this process are
clearly stated, others are present but are easy to overlook, and others
really need to be updated. An example of this last group is what "matching
text" means for an image, which is currently grouped with keyword and term
but can't work quite the same way. This subject was discussed on the list
and at the TC several months ago, and I took on a proposal item to clarify
the spec.

I've started laying out the rules based on earlier discussion, and I'd like
to get early feedback before I complete the Stage 2 proposal document. The
early work is available on the wiki here:
https://wiki.oasis-open.org/dita/KeyrefMatchingText

I'd appreciate any feedback here on the list, or next week at the TC call.

Thanks -

Robert D Anderson
IBM Authoring Tools Development
Chief Architect, DITA Open Toolkit (http://dita-ot.sourceforge.net/)