Laura Huis in 't Veld (Jul 10 2026 at 09:12): Laura Huis in 't Veld (Jul 10 2026 at 09:12):Hi Everyone,
For one of our projects that is dataverse related, the wish to select several datasets at once and to do a 'Bulk' access request came up. Similar to putting them in a 'basket' and perform an action on this selection (in this case an access request). Is there someone in the community with a similar wish? Or someone who has already been working on this?
It would be nice to know how the rest of the Dataverse Community thinks about this idea.
Kind regards,
Laura
Philip Durbin 🚀 (Jul 10 2026 at 10:00):Believe it or not there's already a (shopping) cart object in the Dataverse code that was 9 nine years ago!
The context for this was the "Select multiple datasets for compute" issue at #4246: "As a user of secure multi party computation, I would like to be able to select multiple datasets at one time so I can compute on them together in my computing environment."
Philip Durbin 🚀 (Jul 10 2026 at 10:01):Access request as a use case makes complete sense as well.
The "Request access to and download multiple data files across datasets at the same time" issue at #4480 by Steve McEachern is related.
Philip Durbin 🚀 (Jul 10 2026 at 10:05):Another use case is exporting metadata.
See the "Feature Request: Bulk Metadata and Citation Export for Multiple Datasets and Datafiles" issue at #11892 opened by @Johannes D.
Philip Durbin 🚀 (Jul 10 2026 at 10:20):Open OnDemand Loop (#community > OnDemand Loop) has a feature that's a bit like a shopping cart. Check out this "add files to project" button at ~8:30 in this video:
Like he says, you can select files across multiple datasets. Then you download them from Dataverse into a computational environment.
Laura Huis in 't Veld (Jul 10 2026 at 11:57):Wow, 9 years ago!
I was combing to the Github issues, but apparently I was using the wrong keywords . Thanks a lot for the information. Really helpful to know what is already there. The #4480 issue from Steve is very similar to our wish.
Laura Huis in 't Veld (Jul 10 2026 at 11:59):I guess the 'cart' functionality is not heavily in use, as it was experimental and only to be combined with Swift storage?
Philip Durbin 🚀 (Jul 10 2026 at 12:14):That sounds right, that the "Compute" button is tied to our Swift storage driver, which I believe we consider deprecated at this point. (S3 emulation from Swift is not deprecated and works fine, I hear.)
Yes, this functionality was always experimental. Here's how it's described, even today, in the guides:
"The “Compute” button on dataset and file pages will allow you to compute on a single dataset, multiple datasets, or a single file. You can use it to build a compute batch and go directly to the cloud computing environment that is integrated with a Dataverse installation." -- https://guides.dataverse.org/en/6.11/user/dataset-management.html#cloud-storage-computing
"Compute on multiple datasets" -- https://guides.dataverse.org/en/6.11/installation/config.html#setting-up-compute-with-swift
Philip Durbin 🚀 (Jul 10 2026 at 12:17):We should probably just take it out of the guides. I'd be shocked if anyone is using it. Maybe we should rewrite it and say to try Open OnDemand Loop instead, which is somewhat equivalent, being a computational use case.
Last updated: Aug 18 2026 at 08:27 UTC