FIGI — Clarification regarding process(es) by which the controlled vocabularies get updated.

Raised by Doug Tolbert:

The handling of Associated Data is way underspecified. Where are these items stored? How are they retrieved? Who can update them? Can I create a new one? Who decides what the legal values are? etc.

Clarification regarding process(es) by which the controlled vocabularies get updated.

This affects the specification in a number of places as follows:
1. Section 7.1:
remove ", primarily" from the sentence "While the vocabularies described below and included as a part of the specification are current as of the writing of this document, they are subject to change, primarily through growth, over time."
2. Add a second paragraph to 7.1 reading:
"The process by which the controlled vocabularies changes is directly related to the activities of both the Registration Authority and the Certified Providers (see Appendix B )."
3. Add a new subsection to Appendix B (B.4) as follows:
B.4 New Security Types and new Pricing Sources
In addition to new identifiers, new Security Types and new Pricing Sources can, and will, be created from time to time. The mechanism by which these can be created are entirely parallel to the mechanisms by which new identifiers are created. In particular, as the two conditions necessary for a new identifier: 1. Uniqueness, and 2. Novelty, apply also the Security Types and Pricing Sources, the exact same process will be followed:
• The Registration Authority, as the holder of the definitive repository, is responsible for confirming the uniqueness and novelty of each new Security Type and Pricing Source it creates.
• The Certified Provider(s) are responsible for:
o Confirming the uniqueness and novelty of each new Security Type and Pricing Source it creates against their own inventory.
o Making a call to the RA’s API to confirm the global uniqueness and novelty of the new vocabulary item.
o It is expected that this process can either stand alone as the registration of a vocabulary item or can be packaged as part of the request process for a new identifier which is such that a new vocabulary item is required to support it.