Eric:
I really don't understand why you ordered the elements as you did on the
ballot (is there a meaning to this?). But I am particularly concerned
about the order of the Resource type values. The working group made a
point of putting the list in alphabetical order so as not to appear like
some values are more important than others. I guess it's because you put
"event" first, which may be the least likely of the values to be used,
that it really bothers me. So, please reorder them into alphabetical
order, as the working group presented it.
Rebecca
On Wed, 1 Mar 2000, Miller,Eric wrote:
> Appended to this message is a link to the pre-ballot document that will be
> used as the basis for the balloting system. Please review this document and
> comment on this ASAP. I apologize in advance (again) if I have interpreted
> any of the previous comments incorrectly. It was not my intention. There
> have been several issues that have been identified in the recent months.
> Some with clear solutions some without. Where the necessary 'fill in these
> blanks' has occurred, I've tried to document these.
>
> A few points however should be mentioned:
>
> - Comments have been included where appropriate. it's still not clear which
> kind of comment (general description or context for dc-usage) should be
> included.
>
> - Tokens have been included for each of the items
>
> - Erik Jul raised the issue of grouping together like-qualifiers for the
> relation
> element. This is not done in this version of the document, but will be done
> when the official ballot is introduced.
>
> - The DCMI Point, Box and Agent encoding schemes have been removed... they
> have been identified by several people as not being endorsed by the various
> working groups (date and coverage). Further investigation (via the WG
> documents) confirmed this. We're there other deliverables from these groups
> that the editors missed that support these encoding schemes?
>
> - And now for the big one... :)
>
> How exactly to compartmentalize the "agent core" vocabulary from the DCES
> qualifier vocabulary was not initially clear. The concepts of 'agentType'
> and the identification of two sets of these 'types' (AAT and MARC Relators)
> I believe are intended to be used as means for refining the CCP elements.
> No other working group focused on the mechanisms for refinement but rather
> focused on the actual refining semantics. As such, it was not clear either
> (a) how to incorporate this, or (b) where it belonged.
>
> Andy Powell identified this and proposed a way forward. His set of
> refinements are aligned with the set identified by Rebecca and a set of
> members on the MARC Relator Code working groups. These refinements I
> believe satisfy the function requirements identified by the Agent working
> groups through the notion of 'agentType' and are inline with the rest of the
> elements notion of refinement.
>
> As such, the refinements identified by the LOC group for the CCP elements
> are listed in the ballot and all have the value of an 'DCMI Agent'. The
> vocabularies for defining tha describing DCMI Agent is in a separate set.
>
> Partitioning this as such seems to make a tremendous amount of sense...
>
>
> I'll send the rest of the modification notes out tomorrow, but I wanted to
> make sure people had a chance to see this...
> http://rdf.dev.oclc.org/dc/dcqballot/DCQBallot-20000301.v2.html
>
> eric
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|