I found your post to be somewhat confusing. Since we are a publisher rather than a funder, I am only interested in how this changes the schema for registering DOIs for articles. If I understand what you are proposing, you should probably update the schema 5.3.1 to indicate that the fr:assertion/@name attribute can now take the valueror. The current specification says that it can be fundgroup, funder_identifier, funder_name, and award_number.
<xsd:attribute name="name" use="required">
<xsd:simpleType>
<xsd:restriction base="xsd:NMTOKEN">
<xsd:enumeration value="fundgroup"/>
<!-- fundgroup: used to group funding info for items with multiple funding sources.
Required for items with multiple award_number assertions, optional for items with a single award_number -->
<xsd:enumeration value="funder_identifier"/>
<!--funder_identifier: funding agency identifier, must be nested within the funder_name assertion-->
<xsd:enumeration value="funder_name"/>
<!--funder_name: name of the funding agency-->
<xsd:enumeration value="award_number"/>
<!---award_number: grant number or other fund identifier-->
</xsd:restriction>
</xsd:simpleType>
</xsd:attribute>
I checked this against the latest release 0.3.2. The fundref.xsd that is included with this bundle does not have the name value of ror in it.
We have tests on our code to generate the crossref XML for the crossref 5.3.1 schema, but if we start putting a name value of ror in the fr:assertion element, then it will fail validation.
It would also be helpful to provide complete examples of how to report a <journal_article> that has the following:
- one funder, one grant id
- one funder with two grant ids
- three funders, one of which is identified by ROR and one of which is identified by fundreg, and one of which has neither identifier so presumably should be identified by name.
The samples under “best practice examples” reference schema 5.3.0, so they appear to be out of date.
I also never got a suitable response to my previous question #1 about whether funding can appear under <journal_article> instead of <crossmark>. It’s peculiar to include it under <crossmark> since it has nothing to do with update policy. Some DOIs don’t participate in crossmark.