[osis-core] references and self-ids Part 3 - Proposal

[Harry Plantinga]

Todd,

See comments below.

> Consider the values of the osisID, prev, and next attributes of both the
> segmented verses and quotes.
> 
> <p cite="Matt.13.11-Matt.13.13">
> 	<verse osisID="Matt.13.11" next="Matt.13.11">Jesus answered,
> </verse>
> 		<q type="blockQuote" osisID="Matt.13.11"
> next="Matt.13:14">
> 			<verse osisID="Matt.13.11" prev="Matt.13.11">The
> knowledge about the secrets of the Kingdom of heaven has been given to
> you, but not to them.</verse>
> 			<verse osisID="Matt.13.12">For the person who
> has something will be given more, so that he will have more than enough;
> but the person who has nothing will have taken away from him even the
> little he has.</verse>
> 		<verse osisID="Matt.13.13">The reason I use parables in
> talking to them is that they look, but do not see, and they listen, but
> do not hear or understand.</verse>
> 		<verse osisID="Matt.13.14" next="Matt.13.14">So the
> prophecy of Isaiah applies to them:</verse>
> 	</q>
> </p>
> <p>
> 	<q osisID="Matt.13.14" type="blockQuote" prev="Matt.13.11"
> next="Matt.13.16">
> 		<q type="blockQuote">
> 			<lineGroup>
> 				<line><verse osisID="Matt.13.14"
> prev="Matt.13.14">This people will listen and listen, but not
> understand;</verse></line>
> 				<line><verse osisID="Matt.13.14"
> prev="Matt.13.14">they will look and look, but not see,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> next="Matt.13.15">because their minds are dull,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">and they have stopped up their
> ears</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">and have closed their
> eyes.</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">Otherwise, their eyes would
> see,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">their ears would
> hear,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">their minds would
> understand,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15" next="Matt.13.15">and they would turn to me, says
> God,</verse></line>
> 				<line><verse osisID="Matt.13.15"
> prev="Matt.13.15">and I would heal them.</verse></line>
> 			</lineGroup>
> 		</q>
> 	</q>
> </p>
> 
> The trouble is that without some sort of grain all of the identifiers
> for prev, next are the same for a single verse.  Even though software
> can determine which goes first and which goes next there would be
> multiple matches for a search for "Matt.13.15".  It would seem to me
> that you would want to find a single "Matt.13.15" verse element and
> check to see if it has been segmented by checking for a next.  If the
> desire is that all pieces of a verse be found when looking for
> "Matt.13.15" then if they are encoded with a grain then you could simply
> find all verses with an osisID that begins with "Matt.13.15@" or equal
> to "Matt.13.15".  
> 
>  Conceptually, it does not make sense for more than one verse to an
> identical osisID.
> 
> There is further benefit in that the only processing of segmented verses
> will not be for the purpose of stepping from the first to the next in
> sequential order.  Having an truly unique osisID for each verse element
> allows for reliable pointing into the document.  Both pointing
> unambiguously to the "head" of a segemented verse and pointing
> unambiguously at a subsequent fragment.
> 
> The style of grain used should be left up to the encoder.  I prefer
> @char due to its precision but others may prefer to number the pieces
> sequentially (@enum or @seq) in order to simply make unique ids. 
> 
> Todd


It seems to me that there are two different semantic operations going
on here. One: identify a (possibly segmented) container as "I am Matt.1.3".
Two: connecting the segments of a segmented container.

I know there was a previous discussion on how to connect the segments, but
I don't recall the outcome. But I propose that we don't conflate IDs
from a reference scheme and a means of connecting segments of a segmented
container.

-Harry