An organisation can commit to persisting the identifiers used to identify items and entities within its collections. By persisting an identifier, they must commit to continue to use it as the primary means of identifying a collection object or entity, including migrating it when systems are upgraded and using it across different systems.
Persistence of the identifier depends on your own organisation's ability to manage your local web pages, so if you change the web site all existing identifiers will still resolve.
Given the expectation that any identifiers will persist indefinitely, once you have minted identifiers it is difficult to reverse that step. The things you have chosen to identify, and the format of the identifier are decisions that your organisation will potentially have to live with for a long time.
If you want the identifiers to be used (e.g., cited) then you will need to make sure those identifiers are easily visible, and ideally provide a way for them to be cited.
Identifiers as links with no commitment to persistence
We value your feedback on this resource. Any suggestions on the content or additional examples would be welcome.
To submit feedback, you can raise an issue via Github (requires an account) or you can use the Feedback button in the bottom right of each screen.
The British Library has committed to maintaining its ARK identifiers for the long term; this will be officially described in a forthcoming persistent identifier policy. It has also committed to maintaining the namespaces which are used to access the resources associated with the ARK identifiers.
The Natural History Museum's data portal uses URIs, compatible with the CETAF Stable Identifiers initiative which they have committed to persisting.
Consult the different aspects, from here you may want to look at Persistent identifiers externally resolvable in a human readable format.