GRID/ROR transition

Hi all,

As you may have heard, After 6 years, Digital Science’s GRID will be retiring its public releases in Q4 of 2021 to make way for ROR to develop further. ROR

We’re excited about this transition, and thankful to Digital Science for providing the GRID seed data to launch ROR and for helping to coordinate registry updates.

ROR and Digital Science are committed to making the transition as smooth as possible for anyone who may wish to make the switch from GRID to ROR.

You can find resources about transitioning from GRID to ROR on the (new!) ROR support site, and we’ll be sharing more resources soon. In the meantime, we’re happy to answer questions here, or over on the ROR API users group. You can also get in touch with the GRID team if you have questions for them.

5 Likes

Wow big news! Thanks for sharing @lizkrz

1 Like

Do I understand correctly that the current GRID identifiers will stay resolvable and that this continues to be the responsibility of Digital Science?

1 Like

Hi @wouter.addink , this is something that’s still somewhat in the works. What we’ve heard from the GRID team is that the registry (as in the website) will eventually be sunset as well, though no firm date at this point. @Suze-DigitalScience can you share any additional details from the GRID side?

Hi Wouter, thanks for your question. All previously released versions of GRID will remain available and users can resolve GRID IDs via those releases. The GRID website will eventually undergo a change where organisation pages are no longer available. GRID will continue as an internal tool for Digital Science, but we want to avoid situations with contradicting information available on static pages on grid.ac (e.g. a different organisation name, a former hierarchy) as this will cause confusion for users. This is in line with the approach previously used for OrgRef (see Retiring OrgRef | The DataSalon Blog) as it retired back in 2018. OrgRef IDs have remained available in the GRID database (e.g. https://grid.ac/institutes/grid.9132.9) to help users resolve any old OrgRef IDs if needed. Whether GRID IDs will continue to be relevant for ROR users, and therefore remain available in ROR, will be up to ROR, and we presume will depend on demand from the ROR community.

2 Likes

Hi - the most useful thing to do would be to redirect GRID URIs to the corresponding ROR URIs. This would ease transition for everyone currently using GRID, as there would be no need for us all to update our own data in time for your transition.

The work to do this should be minimal - domain level URI rewriting on the GRID side, and ID mapping on the ROR side.

@CShillum That does sound like a possible solution. One thing I worry about in that case is systems not being aware that this is happening and carrying on thinking that the public GRID registry still exists, because they are getting 200 responses (eventually, after 301 redirect, but it’s somewhat unlikely that an app would pick up on the fact that they are being redirected to a completely different site before finally getting a 200). I’m curious about whether any other GRID users have thoughts on this.

Hi Chris, thanks so much for your feedback.

Digital Science plans to continue to maintain and develop GRID as an internal organisation database, but hand over the torch to ROR, the community-driven research organisation registry.

As a result, public GRID data and ROR data will diverge over time. This means that, in the future, we may see a change in representation of an organisation in the two different sources, including metadata and hierarchy information. A redirect to the ROR record could mean that the public GRID URIs would link to information that may potentially contradict current GRID information.

One of the reasons behind our decision to stop releasing GRID dataset publicly is to prevent confusion with, or contradiction to ROR when it starts its independent curation. Therefore having the GRID URI point to a ROR page that exposes different information should be avoided. This is something that we need to take into account when we choose how to go forward with legacy GRID resolution.

As Daniel mentioned in the ROR Community Call last week, we will take all of the community’s suggestions on board and settle on a technological solution that causes minimal disruption to research infrastructure.

Thanks again for sharing your thoughts with us, and please do keep engaging with us on this topic. We really appreciate these useful insights from our community.