[osis-core] New Schema!

[Todd Tillinghast]

Troy,

> > Would there be a problem with using <p> rather than <div> to contain
the
> > "loose" text and milestones when encoding the cases you described?
(In
> > this case <p> acting as a "leaf div" and not really taking on the
> > meaning of <p>.  Another alternative being an element identical to
<p>
> > that you could use instead of <div> and not have to overload the
meaning
> > of <p>.)
> 
> I don't like the idea of using <p> when, in fact it is not marking a
> paragraph.  As far as using an alternative tag, I still don't see the
> benefits of forcing all text under, say, a <div osisID="Matt.1">Text
for
> Matthew chapter 1</div> to be under yet another child.

I also don't like the idea of using an element for a purpose other than
the purpose that its name intends.  That is why I suggested introducing
a different named element.  (Of course the other element is only
necessary if you buy the rest of the argument.)

I am trying to separate the behavior of "nesting container that contain
other containers" from the behavior of "nitty gritty text container".

Other than <div> no other element can nest within itself endlessly.  We
made sure the <p> can not contain another <p> and <verse> cannot contain
another <verse> or <p>.  As it is we are assured that the "nitty gritty"
text will only occur in leaf level containers and not in containers that
contain other containers (exception being leaf level elements that
contain <verse>).  

I think that <div> should be a simple container that contains other
elements that contain the nitty gritty text AND should not exist at the
lowest level.  I don't have a problem with having an entire chapter
contained by a single element.  I just think that we should reserve the
meaning of <div>.  So in your example we might have the following:
<div osisID="Matt">
	<newElementNotDiv osisID="Matt.1">text of Matt.1 with milestones
and no elements like p, verse, etc... if desired
	</newElementNotDiv>
</div>

Knowing that <div> only plays a single role simplifies the set of
options users will have to account for when using an OSIS document.
They will be able to use <newElementNotDiv> in the same way that they
use <p>, <verse>, <line>, etc... and will be able to reliably treat
<div> EXCLUSIVELY as a container of containers.  Having one more element
that acts in the same fashion as other "leaf level" containers will not
add any significant burden, but having an element that act like a leaf
level container at any level of the containment tree adds complexity.

I don't think you intend to use an individual <div> element as both a
container of other <div> elements as well as a direct container of text.
This is the case I am trying to eliminate.  If you do intend to use
<div> in this way then that would explain our different perspectives.
Let me know if that is the case.

Have I made more sense this time around?

Todd
<snip>