Open points for CSH Data Model design

Open points for CSH Data Model design

Q: Should Rights Metadata be attached at the pcdm:File level, or at pcdm:Object level?

A: (Unmil) Great question. This is something we should discuss with the project team. I think attaching the rights at the file level will give us more flexibility. For example, we could allow different rights for high-res and low-res versions or non-OCR-ed files vs. textual transcriptions.

Q: Should the low-res derivatives for multi-page documents like Board Meetings, and Patient Registers also have PDF files for entire meetings/registers?

A: (Unmil) Yes, I think so. The digitizing company already created these for us. I don't think that they are particularly usable, given their file sizes but we should think about either using the files that we have or generated something at a more usable file size (not sure if this is possible).

Q: Should the technical metadata info about the capture device assume labels such as "scanner model"? Or since cameras were (also?) used for captures, should the relevant metadata fields be named just "capture device model"? Or should we keep both sets of fields (as also in the MIX specifications)?

A: (Carson) In the interests of interoperability, keep both sets of fields ("elements") and use whichever is appropriate, rather than creating a new element. Or bring in another schema if its elements more closely match project needs. (24-Oct-2016)

Q: Should detailed photometric information (if available) also be embedded as metadata? That includes information like 'color space' (CMYK, sRGB, etc.), 'color profile', YCbCr sampling info etc.?

A: Not at this stage. (09-Aug-2016)

Q: Is there something called "too much metadata"? (tongue)

A: Yes, depending upon the balance between design complexity and time to implement.