Call for review of new metadata documents

[Jeff Good]

Dear OLAC-Implementers,

First, let me thank Gary and Steven for pulling together all these  
comments and making the usage guidelines revisions. It's great to see  
these things moving forward.

It will take me some time to put together all of my comments on the  
revised guidelines, but I have one technical question now that I'm  
hoping those better informed about Dublin Core can answer.

It's agreed that the isTranscriptOf and hasTranscriptOf relations are  
needed, but the conclusion is that we can't do anything about this in  
revision 1.1, but this has to be held off for revision 2. What I don't  
understand is why we can't use the existing model of OLAC controlled  
vocabulary refinements for this in the meantime.

For example, why can't we use something like the second element, which  
looks to me to be mostly parallel to the first element, which comes  
out of the guidelines, using the prescribed method for encoding  
subject language:

<dc:subject xsi:type="olac:language" olac:code="sky"/>

<dc:relation xsi:type="olac:lingrelations"  
olac:code="isTranscriptOf">some-unique-identifier</dc:relation>

I guess the problem here is that "olac:code" in the second case would  
not be encoding a "thing" but a "relation". But is that so bad? Could  
we just call this "olac:predicates" to deal with this? The OLAC->DC  
mapping would still be straightforward (just strip out the "olac:"  
attributes), right?

If this isn't possible for some technical reason, however, perhaps we  
can get a jumpstart on what will be an important feature of OLAC 2.0,  
by having someone draft the relevant document that will be needed to  
describe these refinements at some point anyway?

I also wonder if, as a stopgap measure, why there can't be a  
recommendation about how to encode this in the meantime even if it is  
not officially part of the standard. For example, can't we informally  
agree to at least do something like this:

<dc:relation>IsTranscriptOf: some-unique-identifier</dc:relation>

At the very least, this should help people prepare for the fact that  
in OLAC 2.0, there will be an official way to code this.

Jeff