For Identifiers (just the complete URI and local part, not about mechanisms)
...
#permanent
...
#permanently resolve to a digital object
...
#unique within our PID system without the DNS name portion of the
...
#URL
...
#works with existing systems
...
#resolvable through a web browser
...
#Simplicity
...
#support for a local namespace prefix and an identifier part
...
#support for opaque identifiers
...
#Local part should not be Cornell branded
...
#may be surrogate for physical object
...
#should be short
...
#should be easy to copy by hand
For Resolver and System
...
#Supports billions of identifiers
...
#Robust architecture
...
#Robust implementation
...
#ability to request metadata about the identifier
...
#lightweight understanding of identifier equivalence
...
#should be easily discoverable by Google
...
#integrate well with the web architecture
...
#vitality checking
...
#Need to avoid unbounded generation of surrogate persistent identifiers
...
#PID corresponding to every Cornell NetID
...
#useful error when check digit is wrong
Governance Issues
...
#Requires payment to external organization
...