Improve framing of single point of contact policy

[cthoyt]

RE: https://github.com/biopragmatics/bioregistry/blob/main/docs/CONTRIBUTING.md#contact-and-attribution

Twice recently, I've got feedback from resources that would rather have a group contact email than a single point of contact. This is explicitly against the Bioregistry's policy on contacts and attribution, which has the dual purposes of:

1. Making it transparent who works on a resource. For example, the MeSH resource is historically super opaque. Figuring out who actually works at a place is often the first step towards making a real interpersonal connection, from researcher to researcher.
2. Making it possible to reach out to someone who is actually responsible. In a small number of cases, a group email address is actually responsive, but it feels like most of the time, these group email address simply don't go anywhere.

Without changing our policy, how can we re-frame the way we communicate how we keep contact information (e.g., changing from contact to something else?) to make groups happier with the way we're doing things?

Some possible extensions, all of which can allow for curating more information in addition to, but not instead of the primary contact:

• New schema field - group email #1389
• New schema field - contact page #1392
• New schema field - secondary contact list #1388

[bgyori]

One simple solution is to keep the schema as is, and just allow group emails to be used under contact / email when requested and determined to be the best choice (as I believe it is the case for UniProt). This would still allow always preferring a specific person as contact, and leaving the exceptions up for reasonable arguments and human decision.

If we want to be more structured, creating a separate key for group_contact, and treating this as additional information that is curated when it exists is possible. We would still attempt to always curate a specific person as contact when one exists, but the two fields would be independent otherwise.

[JervenBolleman]

The ideal of connecting people is great. However, as someone working on projects where the help desks are responsive and monitored this works badly. The curation workload for this is also quite high. Case in point is the addition of Rolf Apweiler as contact to UniProt. Rolf, had left the UniProt consortium 8 years before the first commit to the Bioregistry.

For projects that run decades or longer, mesh, uniprot, ensembl, ena, genbank etc. The contact people will change over time. Their e-mail addresses will not be stable, nor are the selected people actually aware of all the different details and are often overwhelmed by very busy e-mail boxes. The helpdesks are monitored by multiple people and issues that are forgotten are monitored and staff chased to get the right answers.

Again a real case: as I am the public face of the UniProt RDF and SPARQL, I do get direct e-mails on my work account. E-mails that at one point where not answered for months, because of medical leave. If those people writing those e-mails had written used the contact form instead they would have had a response by one of my colleagues. This is what my out-of-office reply encouraged people to do, but a few did not. Causing them unnecessary aggravation.

This is why I suggest that the policy is to prefer helpdesks as a contact point, and only if those are not available or shown to be inactive peoples -emails are added. In any case I think that you should ask permission for adding these people as contact points, as they might not wish that. If permission has been asked to Rolf to add him as contact point he would have at least pointed you to the current EBI UniProt PI Alex Bateman, whom would have pointed you to his policy of using the webform/helpdesk.

As a non lawyer note: collecting personal e-mails and names without explicit consent seems iffy to me.