Fix Terms of Use message for some dataset versions that have no license or terms metadata

[jggautier]

When the multiple license update was applied to the Harvard Dataverse Repository, in the Terms tab of pages of dataset versions that have no license or terms metadata, text was added to indicate that the version "is made available without information on how it can be used. You should communicate with the Contact(s) specified before use":

But there are datasets with one or more versions that have no license or terms metadata and whose Terms of Use fields do not have that text about how the dataset has no information about how it can be used. As of October 2022, there were at least 4,392, including the dataset version at https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/007GT&version=2.0:

The text on the page reads that "the following Custom Dataset Terms have been defined for this dataset", but there are no custom dataset terms.

Most of these datasets are in one collection, were created and published before the multiple license update and were created using the Dataverse API. The remaining datasets are in two other collections and may have been created using the API, even after the multiple license update, such as the dataset at https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/MWIIHG&version=1.0.

Maybe the database commands to add text to the Terms of Use fields didn't account for these datasets for some reason (e.g. null versus blank values)?

Can these dataset versions be updated to include in their Terms of Use fields that the datasets are "made available without information on how it can be used. You should communicate with the Contact(s) specified before use"?

[jggautier]

IQSS/dataverse#8796 is related and was addressed, which might address this issue. I'm following up with the team to understand what was done. See IQSS/dataverse#8796 (comment)

[jggautier]

Dataverse v6.4 included an update that ensures that for datasets that have no license and no custom terms of use, users are told this instead of being told that the dataset has custom terms of use. See the screenshot at IQSS/dataverse#8796 (comment), taken before v6.4 was applied to Harvard Dataverse, and see that dataset page now, post v6.4.

But in IQSS/dataverse#8796 (comment) I wrote that there are still datasets that tells users that they have "Custom Terms", when they have no custom terms and no licenses. Some of these are in the collection at https://dataverse.harvard.edu/dataverse/MIT-PSFC, including the dataset at https://doi.org/10.7910/DVN/5YY6PE. Some other datasets are at https://doi.org/10.7910/DVN/RFT3ME, https://doi.org/10.7910/DVN/8ACDTT, https://doi.org/10.7910/DVN/FMBUZB, and https://doi.org/10.7910/DVN/IW4V4Y.

@qqmyers wrote that these, and maybe other datasets in Harvard Dataverse, may have been missed when v6.4 was applied to the repository, specifically changes cause by the code at IQSS/dataverse#10614, and shared a query for finding more of these datasets, and updating them so that users are told that they have no license and no custom terms.

I wrote that I'd open a GitHub issue in this Harvard Dataverse repo for these tasks but then I re-found this GitHub issue, so I'm using this one instead.

I'm proposing that the definition of done for this issue to:

• Find all datasets that were missed by the fix in v6.4 (8796 fix no license display and indexing dataverse#10614)
• Use that info to help us see if there are still ways of creating datasets without a license or custom terms of use
• Apply the 8796 fix to those datasets using the query that Jim shared

And of course if we find that there are still holes that let users create datasets without licenses or custom terms of use, we can open one or more GitHub issues about figuring out if and how we should close those holes.

[sbarbosadataverse]

Remaining action items:

1. Do the dbase query that Jim suggested and update the dataset terms page: Feature Request/Idea: Indicate on dataset page when dataset version has no license or custom terms dataverse#8796 (comment)
   to update the datasets that were missed
2. Outreach to these non license datasets to ask them to add a license or we will automatically default to CC0-BY or other more restrictive license (International non commercial BY?)
3. How many datasets have this scenario?

[cmbz]

2025/02/06:

• Please size #1 above.

[cmbz]

2025/02/12

• Running the query is a size: 3
• Note that addressing the other points will take longer.