[Review recommendation] Add a .janno colum to document the broad chronological era of a sample

[nevrome]

This recommendation was raised in the review of the Poseidon paper.

[nevrome]

Broad chronological era or even cultural attributions would be useful but are hard to maintain as fixed-choice fields given the heterogeneity of terms on a worldwide scale and the many unresolved controversies around their precise definition.

I think this leads to a foundational decision we have to make for Poseidon: Do we want to include more archaeological information with all it's uncertainties or draw a line where we currently are?

[stschiff]

I share your skepticism. One thing we could potentially do is to add a column for a URI identifying a specific culture or time period, which could resolve to a Wikidata entry, or some other entity. But we could also simply push back on this one.

[nevrome]

Maybe this could indeed be the moment to start integrating with the linked open data world. We could create a column that expects a URI, for example a period.do permalink: e.g. http://n2t.net/ark:/99152/p0zj6g8ks9s

The biggest challenge here is imho to provide a documented workflow on how to find the most relevant entity for the package creator.

And for the validation we would need an (optional, because extremely expensive?) process to resolve the URIs and make sure that the referenced entities have certain minimal properties.

[stschiff]

Yes, you've summarized the challenges. Needs to be discussed further when we deal with the revision.