#692 Pointing between terms and glosses

[Martin Holmes]

In the main Guidelines example for <term>, we see:

We may define <term xml:id="TDPV" rend="sc">discoursal point of view</term> as
<gloss target="#TDPV">the relationship, expressed
through discourse structure, between the implied author or some other addresser, and the
fiction.</gloss>

Surely it is much more likely that there will be a single central gloss or definition for a term, and that each usage of the term will point to that central definition? In other words, the pointing relationship should be inverted for the most common case:

We may define <term corresp="#TDPV" rend="sc">discoursal point of view</term> as
<gloss xml:id="TDPV">the relationship, expressed
through discourse structure, between the implied author or some other addresser, and the
fiction.</gloss>

[Sebastian Rahtz]

I agree, mildly. I never did really get that example anyway, as I can't imagine using either way in real life. I can imagine linking each occurrence of <term> to a
category in a taxonomy in the header.

I wouldn't call this a bug, just a slight possibility of an odd design pattern.

[Lou Burnard]

It's certainly not a bug. You could argue it either way. It's surely conceivable that you might have a single term with multiple glosses (e.g. in different languages) at least as often as multiple terms with a single gloss. I think that the linking of multiple occurrences of a term with a single taxonomy entry is a different style. In a printed text, it's just the first occurrence of a technical term which gets highlighted, not all of them.

[Martin Holmes]

Not a bug, perhaps, but certainly not a feature request. It throws people who look at it. At the very least, I'd like to add a reverse example to the spec, and a note to explain the two possible scenarios.

[Laurent Romary]

Agree. We need to make a clear case of all the use cases where terms are sequences tagged in a text and pointing to a reference description, whether a definition within the same text (not mainstream to my view) or an external reference. Sebastian mentions thesauri, but it could also be a real terminology (with a multilingual representation). This is typically what we have in the structure of an ISO standard with a distinct "terms and definitions" section. In such cases, this supports the idea that we should have a terminology chapter again somewhere or means to describe TBX like objects as presented in http://hal.inria.fr/hal-00950862 (with a first ODD and examples).

[James Cummings]

At Raleigh F2F Meeting 2014-11-18 Council decided that we should have an additional example -- perhaps the same one the other way around. Using @ref not @corresp of course. Assigning to LB