CreatorDID

[Doug Tody]

Hi Gerard -

If you have any connection to the project which produces the data,
encourage them to assign a unique CreatorDID to datasets as they
are produced.  For older data, or replicated data which one does not
control, this may not be possible, and the CreatorDID should be omitted.
Note though, that it is always possible for the publisher of the data
to assign a PublisherDID, which is unique within the local address space
of the publisher.

While it is recognized that at present it will often not be possible to
assign a CreatorDID, part of the motivation for defining CreatorDID is to
encourage data producers to assign these values so that in the future we
can more easily manage capabilities such as replication and description
of provenance.  To define a CreatorDID, the data creator needs to register
an authority ID, then they can assign unique identifiers.

	- Doug


On Wed, 1 Aug 2007, Gerard wrote:

>
>
> Dear Doug
>
> We have problems understanding how to assign values to the CreatorDID in the
> SSA protocol.
>
> >From the text we understand that this is an identifier assigned to a dataset
> (here a spectrum) by the creator of the data.
> In our case, where we provide access to archived spectra, we understand this
> to mean an id assigned by the astronomer(s) who created these spectra.
>
> We also understand that the CreatorDID should be an "IVOA dataset
> identifier", ie have a form following the Resource Identifier syntax, i.e. a
> URI starting with ivo:// .
>
> Our creators have not assigned such an id to their spectra. So what should
> we do ?
>
> Thanks
>
> Gerard Lemson
>