Regarding optional/mandatory discrepancies between Data Dictionary and MIX
1.0- there were two reasons for this:
- the DD specifies many mandatory elements. The experience of the
implementors of the earlier version of MIX was that they could not in
practice comply with so many mandatory elements but still wanted to use
MIX to provide the data they could capture. The assumption here is that
still holds true.
- the DD specifies that some elements are conditionally mandatory
(e.g. element foo is mandatory if the value of element bar is x. XML
Schema does not accomodate this very well and so it seemed another reason
to not attempt make the schema follow the data dictionary in that regard.
Morgan Cundiff
On Sun, 18 Mar 2007, Gary McGath wrote:
> [log in to unmask] wrote:
> > Hi Gary,
> >
> > I'm looking at the ANSI/NISO Z39.87-2006 document. There (on pp.
> > 76-78), it states that these two data elements (bitsPerSampleValue and
> > bitsPerSampleUnit are mandatory, not optional.
>
> Then there's a disagreement between the data dictionary and the schema.
> All of these elements in the schema have minOccurs="0", which makes them
> optional.
>
> > What may be confusing about bitsPerSampleUnit is that it really
> > specifies how the bits are to be interpreted, as "integer" or as
> > "floating point."
> >
> > To answer your question, based on the NISO document's examples, for
> > 24-bit RGB:
> > bitsPerSampleValue = 8,8,8
> > bitsPerSampleUnit = integer
>
> Well, OK, but that's counterintuitive at least. "8,8,8" is not what
> anyone would consider an "integer" value. I hope documentation will be
> forthcoming to explain this. The encoding of multiple values as a
> comma-separated string isn't great XML practice, but I guess we'll have
> to live with it.
>
>
> > Hope this was helpful,
>
> Thanks.
>
> --
> Gary McGath
> Digital Library Software Engineer
> Harvard University Libraries, Office for Information Systems
>
|