Provide a field for orcID

[rly]

Experimenter names are not unique. When creating a database of NWB data, if there are two people with the same name who have generated data in the database, their data will be incorrectly linked to each other. orcID is a popular service that provides a unique ID for each researcher, and is used by many publications to uniquely identify an author.

DANDI shows orcIDs with their dandisets, but I believe these are human-entered. e.g., https://gui.dandiarchive.org/#/dandiset/000003

NWB should allow the inclusion of orcIDs for experimenters so that they can be uniquely identified. Right now the /general/experimenters dataset has shape (None, ) and dtype string. We could add orcIDs in:

1. an attribute on the /general/experimenters dataset with the same size as the dataset
2. a new dataset /general/experimenter_orcids with the same size as /general/experimenters
3. do nothing

[oruebel]

4. An alternative solution for ORCIDs would be to use ExternalResources. I think the mechanism for linking experimenter names to ORCID should essentially be the same as linking any other field to an ontology.

[rly]

That makes sense. Not sure why I didn't think of that!

[oruebel]

Closing this issue for now since this will be addressed by #1 and in #15 in PyNWB and HDMF ExternalResources