LINDAT/CLARIN DSA

• In 2013, by 2010 rules, in the middle switch to 2014/2015 rules
• It took us more than 6 months. 2 personal attendances of workshops an several attempts (and refusals)
  • Once we were refused for a grammatical error in one sentence (missing argument of a verb).
• Still, not really hard, and
• quite useful. Good opportunity to:
  • clarify processes,
  • find out things you haven't thought about yet.

Main problems

1. Vagueness of the questions and absolutely no explanation, "how do we mean it". (gone by now, I think). E.g. #13:

13. The technical infrastructure explicitly supports the tasks and functions
described in internationally accepted archival standards like OAIS.

Minimum Required Statement of Compliance:
3. We are in the implementation phase.

- Sometimes it is *controversiality*, not vagueness. E.g. *#2: For DSA data
formats must be required and data producers must provide data (only) in
these formats.* **What if a centre doesn't think this is a good idea? No
DSA?**

2. Almost equal (and given the questions absolutely understandable) vagueness of the answers of the centres who have already been successfully assessed.
3. Subjectivity of assessment (probably unavoidable, given the above). Answering just like somebody before you might suddenly not be enough.

Main benefits

1. Helped us clarify and formalise lot of our policies
   • e.g. write a code of conduct
2. Made us write a bit more documentation
3. Made us even make some arrangements. E.g. #7: Plan for long-term preservation.

Other points

• Related to the above. E.g. some subjectivity is needed in the assessment, at least with the current rules. It just needs to be done right :) Otherwise we would never get DSA: #2 We don't really require a closed list of formats. #3 We don't always require data to be submitted either. All of this has excelent reasons, of course. Assessors did accept our reasoning.
• Some requirements surprisingly weak. E.g.

The data repository enables the users to discover and use the data and refer to
them in a persistent way.

Minimum Required Statement of Compliance:
2. We have a theoretical concept.

So it is not really needed to be able to search the repository, use the data, and reliably refer to them? To me this is the single most important thing!

Finally

• With enough explanatory information (e.g. Paul's slides) DSA is not too hard
  • Given you have a good repository in the first place
• Mostly writing explicit documentation and clearing up processes
  • Good things!
• Some questions could be made clearer
  • for no explanations to be needed
  • not to sound so scary (How DO you handle international law requirements?)
• Some other criteria could be stronger (only theoretical concept of persistence?)