SPOCC and other specimen APIs alignment

[mjcollin]

Integrate iDigBio APIs into SPOCC or evaluate another common API for all the various specimen record aggregators.

[sckott]

I assume you mean https://github.com/ropensci/spocc ? If so, :+1: and I can def. help here

[mjcollin]

Indeed. I noticed you wrote it and I'm glad you're coming :) Since you've worked on everyone else's APIs in some way, I'm looking forward to discussing some kind of end-game for them all. Is there enough commonality to what is stored and how it's accessed for occurrence aggregators & providers to settle on some standard behaviors like we've settled (kind of) on DwC for representing information?

[sckott]

There are some similarities, in addition to many differences. wrt DwC, some of them have adopted DwC, including GBIF, VertNet, etc., but I don't think any prefix dwc terms as in idigbio dwc:specificEpithet: "ilia"

Some pain points:

• Variation in spatial search - some only support WKT, some only bounding boxes, some points with radii, etc.
• Taxonomic - each provider can have data for the same taxa, but unless they are following a particular taxonomy, we can run into problems in trying to use data from different providers for the same taxa.
• Duplicates - alot of data eventually ends up in GBIF - so if using X provider in addition to GBIF, be careful
• the I want all the data use case - rate limited REST APIs are fine for most people I think but some want e.g., all the records available for 10K species. This isn't a good fit for the average REST API, but better for a download API. Some providers provide programmatic access to download APIs (idigbio, vertnet, GBIF will soon) - the downloaded data though requires different handling on the client side, but libraries should take care of this.

[sckott]

started adding in idigbio to spocc, see https://github.com/ropensci/spocc/issues/121

[sckott]

Got a few more things to work on, but essenitally working now