[osis-core] Further osisRef stuff

[Troy A. Griffitts]

Patrick,
	Exactly.  I believe that we have all always agreed on this.  But thank 
you for articulating it.  I may be wrong and others may not agree.

		-Troy.



Patrick Durusau wrote:
> Todd,
> 
> I was talking to Steve yesterday and he suggested that:
> 
> <work osisWork="kjv">
>        <identifier type="OSIS">Bible.KJV</identifier>
>        <referenceSystem>Bible.KJV</referenceSystem>
> </work>
> 
> Where there appears later in the document:
> 
> <osisRef="kjv:Ps.43.2"/>
> 
> Resolves to *all* that appears between <work osisWork="kjv"> to </work>, 
> in the sense of being a pointer to that set of information.
> 
> Thats makes a lot of sense since most of our Dublin Core entries, for 
> example, will have ISBN data, which is a much more reliable source for 
> text identification than most others.
> 
> In other words, what use the software makes of this bundle of data that 
> is represented by "kjv" is really an application level problem.
> 
> So, is the semantics of <referenceSystem> a claim that I am making how I 
> am making references to the work? In other words, <referenceSystem> is 
> not a claim about the work but about what referenceSystem I am using to 
> refer to it?
> 
> To illustrate:
> 
> <work osisWork="frAndKJV">
>    <identifier type="OSIS">Bible.Fr.Segumend</identifier>
>    <referenceSystem>Bible.KJV</referenceSystem>
> </work>
> 
> so when I say:  <reference osisRef="frAndKJV:Ps.43.2"/>
> 
> That means I am using the KJV reference system for this reference so 
> please find the proper reference in Bible.Fr.Segumend?
> 
> Changes the semantics of referenceSystem to be the referenceSystem 
> (abstract) that I am using to make references to this work. (Most of the 
> time these are the same but not always, cf. St. Augustine's Confessions.)
> 
> Note that <identifier> is not required in <work>. Should it be?
> 
> Patrick
>