Skip to content

Latest commit

 

History

History
17 lines (9 loc) · 3.03 KB

anatomyPersistentId.md

File metadata and controls

17 lines (9 loc) · 3.03 KB

Anatomy of an ARK and Handle request

We received many questions from suppliers on persistent identification. For your convenience we have compiled the following overview that uses ARK and handle.net as an example.

The anatomy of an ARK persistent identifier is comparable to that of Handle.

In ark:/60537/by6i35 the 60537 uniquely represents the institution (in this case "'Gouda Time Machine") and by6i35 is the identifier that Gouda Time Machine allocated to the digital resource. You are free in the way you generate the identifier as long as it is independent of the software system. The identifier is persistently linked to the resource and it must be able to migrate to any other platform in the future.

Handle uses the same pattern. In handle:/10648/b017027e-d0b4-102d-bcf8-003048976d84 the 10648 identifies the institution (in this case the National Archive of the Netherlands) and b017027e-d0b4-102d-bcf8-003048976d84 is the GUID of the resource they created.

You need a resolver to find these resources online. A resolver is a software service that runs at a specific web address, which you ask where an identifier can be found.

The ARK resolver is https://n2t.net/ while Handle identifiers can be resolved through http://hdl.handle.net/. The main difference between ARK and Handle is how resolution works.

ARK is a decentralised system and redirects https://n2t.net/ark:/60537/by6i35 with an HTTP 302 status code to https://www.goudatijdmachine.nl/ark:/60537/by6i35 at that point goudatijdmachine.nl is responsible for further internal resolution and redirection of the request to the proper current URL of the resource. The developers of the website implemented a service that redirects the request with another HTTP 302 status code to: https://www.goudatijdmachine.nl/omeka/s/data/item/31488 The ARK Alliance organisation - who maintains the ARK PID system - only stores identifiers of organisations and not of the individual resources. ARK expects you to handle those yourself.

Handle on the other hand is a centralised system. You not only join as an organisation, you also register every individual PID to each of your resources. Resolving a request on http://hdl.handle.net/10648/b017027e-d0b4-102d-bcf8-003048976d84 will point the user with a single redirect HTTP 302 directly to https://www.nationaalarchief.nl/onderzoeken/fotocollectie/b017027e-d0b4-102d-bcf8-003048976d84. The advantage is that the site nationaalarchief.nl does not need to host a second resolving service. But this comes at a cost. Not only do you pay an annual fee for using handle, you also have to maintain the PIDs in the handle registry. You must file any changes to the redirect URLs to avoid breaking links.