[osis-core] <hi> types

[Chris Little]

Todd Tillinghast wrote:

> Chris,
> 
> Can you explain what you mean by 'illuminated letters'?  (I have a
> mental image of electronically back lit letters, which I suspect is not
> what you mean.)
> 
> Can you scan an example?

See http://prodigi.bl.uk/GutBible/midgl2/191.jpg .  The big L is an 
illuminated letter, albeit a fairly boring one.

> It would seem that we should encode some non-presentation related
> semantics.  'Emphasis' with a variety of types (like 'strong', 'weak',
> etc...) might be appropriate.

I don't think anyone ever writes in a way that they would wish to imply 
weak emphasis that could be tagged.  <hi> was designed for emphasis with 
indeterminate semantics.  If we wish to encode this, we might as well 
supply a few standard types so that everyone speaks the same language.

> Aren't drop caps are usually a purely presentation related thing and
> don't need to be encoded.

See the above image for an example drop cap (the C in the right column). 
  In encoding manuscripts, I think we should have the ability to 
identify these as special.  But maybe this is considered too far afield 
from our current mission.

--Chris