I think we should not go down this route. When I want a brief record or
a full record I do  not want something that is narrower or broader than
something for which I do not know how broad or narrow it is.
If I am the only one for which this was an issue I would say let's
leave it as it was.


>>> [log in to unmask] 16-12-2004 0:59 >>>
> Date: Wed, 15 Dec 2004 23:28:37 +0000
> From: Dr Robert Sanderson <[log in to unmask]>
> > What I want is some way in Explain to say that two schemas are
> > related and the nature of that relationship.  The relations are:
> > broader, narrower and variant.
> Ahh. Well, that's somewhat harder, and I should pay more attention
> to the discussion I guess =)


Sounds I mention RDF?  No probably best not ...

> We'd need something like:
> <schema name="dc" identifier="..."> ... </schema>
> <schema name="dc-b" narrowerThan="dc" identifer="..."> </schema>

If we want to go down this kind of route at all (relationships stated
explicity rather than implied by nesting) I would prefer something
more like:

  <schema name="dc" identifier="..."> ... </schema>
  <schema name="dc-b" identifer="...">
    <relation type="narrower" name="dc"/>

Not that I am overly fond of this approach.

> Or preferably to make schema profiles nest.
> <schema name="dc">
>    <title>
>    <profiles>
>      <schema name="dc-b">...</schema>
>    </profiles>
> </schema>

This is cute, but it's going to make life awkward if the nesting
changes (e.g. by the introduction of "dc-fairlybrief").

/o ) \/  Mike Taylor  <[log in to unmask]>
)_v__/\  "rm() { /bin/rm -f $*; echo 'Were you sure?'; }" -- Tee-shirt
         at a Unix conference.

Listen to free demos of soundtrack music for film, TV and radio