Stream: community

Topic: Dataverse installations at re3data.org

Knut Wenzig (Sep 04 2025 at 10:53):

I recently compared the list of Dataverse installations (available at https://dataverse.org/installations and in JSON format at https://iqss.github.io/dataverse-installations/data/data.json) with entries in re3data.org, the registry of research data repositories. I found that a significant number of Dataverse installations are not yet listed on re3data.org. (While you have 137 on your list, at re3data.org 222 repositories report to use Dataverse: https://www.re3data.org/search?query=&software%5B%5D=Dataverse)

My findings are documented in this table, which presents the JSON data in tabular form. The results of the comparison are shown in the last column:
https://docs.google.com/spreadsheets/d/1nAyDYDJ0MVP3Ibhtvf337r3qQY2GQZtM1hwR2VbbY_M/edit?usp=sharing

Here’s how to interpret the entries:

• For 86 installations I could suggest an identifier from re3data.org (r3d...)
• "no OAI-PMH": The installation does not provide access via OAI-PMH. This may (but doesn’t necessarily) mean it’s not listed on re3data.org. I only checked installations against re3data.org if I found an endpoint.
• OAI-PMH endpoint links: These installations were not found on re3data.org.
• "???": The installation was not online at the time of access.

Some thoughts based on my findings:

• Since I’m particularly interested in OAI-PMH endpoints, it might be useful to include this information in your data—although it often can be inferred.
• Including the re3data.org identifier in the Dataverse data could encourage institutions with new installations to register their repositories.
• For Dataverse users, some guidance on how to register with re3data.org would be helpful. Ideally, this would include naming the software (Dataverse), mentioning the use of the DDI metadata standard, and specifying the OAI-PMH endpoint (if enabled). These are just the features I focus on, but I imagine there’s a broader set of common characteristics across Dataverse installations.

You might consider discussing this with the team at re3data.org (info@re3data.org). In my experience, they are very responsive. (I also provided them my list, because they might want to add some of the missing instances to the registry.)

Philip Durbin 🚀 (Sep 04 2025 at 11:19):

@Knut Wenzig wow, this is awesome! A few points.

• https://iqss.github.io/dataverse-installations/data/data.json is used as a data source for https://hub.dataverse.org (see here in the code for the hub). It's still completely valid to look at that upstream data.json as the source of truth for the list of installations but it gets massaged into https://hub.dataverse.org/api/installations which we would like people to switch to, eventually. @Julian Gautier is the primary maintainer of data.json via the script at https://github.com/IQSS/dataverse-installations . @Juan Pablo Tosca Villanueva is the primary developer of the hub, and the code is https://github.com/IQSS/dataverse-hub
• There's a suggestion to incorporate re3data at https://github.com/IQSS/dataverse-installations/issues/190 and https://groups.google.com/g/dataverse-community/c/xFRXZ7BAuQA/m/8xH_iPTqBQAJ if you'd like to link to this Zulip topic (or I can).
• I'm glad to hear the re3data team is so responsive. I've been aware of the service for years but haven't interacted with it much apart from looking at the occasional listing.

Knut Wenzig (Sep 04 2025 at 13:48):

Done: https://github.com/IQSS/dataverse-installations/issues/190#issuecomment-3253817142

Last updated: Nov 01 2025 at 14:11 UTC