Persistent Identifiers (PID) for RDMkit

[bedroesb]

What we want to achieve

• We want to make sure that links towards RDMkit do not result in a 404 page not found.
  • This can be prevented with good housekeeping rules when maintaining the website including:
    • Not changing the filename (which also means the url in our case)
    • If the filename would change, a correct redirect is put in place if the same content has moved.

What having a PID registered on w3ID.org does not solve

• Capturing the content of a page at a certain moment in time
• Guaranteeing that content shows up when the link with the PID is followed (this still depends on putting the correct redirects in place)
• Guaranteeing that the content stays the same (similar to the first bullet point). Content can change over time and the PID would stay the same, do people really want to still link out to this same url?

Why could we have the RDMkit domain registered on w3ID.org

• Having a simple https://w3id.org/rdmkit/* to https://rdmkit.elixir-europe.org/* Could be used as a social tool to tell people: look, we do care about longevity of URLs. This wildcard concept where pages with a certain pattern on the w3id.org side are linked to complementary pages on the target website is used for the RO-crate website:

  RewriteRule ^(\d+.\d+)/?$ https://www.researchobject.org/ro-crate/$1/.
  

  This does not solve the problems mentioned earlier.
• If the domain would change, links in the past would not break, but this would also be the case with a simple redirect on the side of elixir-europe.org (no need for w3id).

What can we do at RDMkit?

• State in our about page that we value the preservation of URLs on our website and we will continue to do so.
• In the editorial tasks page state that filenames/urls should not change after merge
• Increase the importance of the filename
  • Currently page_ids are internally in RDMkit used to link pages with each other (related pages). This is most of the time a shorter version of the pagetitle/filename. We could use the same name as the filename for this instead of different one. People can than use the filename/url when tagging tools or linking related pages. No confusion anymore on where the page_id comes from, because it will be the page url/filename, which should not change after all and is always unique because of how RDMkit is set up (all pages are served in the root of the url)

How FAIRCookbook does it

• FAIRCookbook: Using w3id.org to register a redirect for every individual page using a .htaccess file in the w3id.org repository on GitHub
  • one to one redirect where:
    https://w3id.org/faircookbook/FCB049 -> https://faircookbook.elixir-europe.org/content/recipes/assessing-fairness/fair-assessment-fairevaluator.html
  • .htaccess record example:

  RewriteRule ^FCB010$ https://fairplus.github.io/the-fair-cookbook/content/recipes/findability/seo.html [R=303,L,QSA]
  

  • Pro:
    • Easy way of creating a w3id.org/PID url that redirects towards the page on your website of choice
  • Con:
    • You have to keep track of all you redirects in a separate repo, on which you don’t have the rights (somebody has to think about it)
    • You have to keep the .htaccess file up to date with all changes that are made (new pages/url changes)
    • Need to register first the PID on w3id.org before you can use it on your own page/website.
    • Very maintenance heavy