| 12 Jul 2021 |
 nightowlasu | perfect! | 17:39:42 |
 poikilotherm (CET,UTC+1) | You could create a group with a single IP | 17:40:02 |
| poikilotherm (CET,UTC+1) | Not sure this is what you need or want | 17:40:13 |
| poikilotherm (CET,UTC+1) | Beware that you properly setup handing over the client IP when using a proxy and IP groups... | 17:40:53 |
| nightowlasu | i think it is ... ultimately we just want to be able to restrict the files in a particular dataset to an IP range ... or our asu.edu mail domain ... but if we could do it at the dataset level (ie: only certain IP ranges can even view the dataset) that would be even better | 17:41:07 |
| poikilotherm (CET,UTC+1) | Jim and I discussed this a while ago and he made a change to the way this works. | 17:41:19 |
| poikilotherm (CET,UTC+1) | Sure, just create the proper configs and set them up. Remember that IP groups are global, not Dataverse Collection specific. | 17:42:14 |
| poikilotherm (CET,UTC+1) | So you might want to make sure to give those recognisable names | 17:42:31 |
| poikilotherm (CET,UTC+1) | IIRC you cannot see the group details from the UI. | 17:42:52 |
| nightowlasu | for sure .. and we will document it ... this will be very helpful for our research computing department who want a "private" dataset .. that is private to only ASU campus IPs and/or mail domain .. thanks for clarifying it for me ... we will do lots of testing | 17:45:00 |
| nightowlasu | Redacted or Malformed Event | 18:20:47 |
| nightowlasu | one more question, since the groups are per the permissions available, there is no way to actually restrict viewing of a dataset's metadata based on IP address (or any other user/group)? I don't know of a permissions that restricts viewing of a dataset in any way if it is published | 18:22:10 |
| nightowlasu | ^ just talking dataset landing page w/metadata displayed .. that is always public if published, right? there is not way to limit access other than a separate installation with firewall rules limiting the entire dataverse installation (i know we determined above we can limit file access per IP, domain, which is great!) | 19:04:49 |
| poikilotherm (CET,UTC+1) | Yes. Publish or perish 😋 Once published means public metadata. You might create rules restrict access on the reverse proxy. But maybe just is not a good idea. It's not very FAIR. Public metadata is a pretty smart thing, even if you're just using the registry use case of Dataverse. | 19:11:14 |
| poikilotherm (CET,UTC+1) | As soon as the DOI is made available, much metadata is public anyway | 19:11:44 |
| poikilotherm (CET,UTC+1) | I don't think there is a way to hide the metadata of a resolvable DOI. | 19:12:05 |
| poikilotherm (CET,UTC+1) | You could go for unpublished items with reserved DOIs though | 19:12:25 |
| poikilotherm (CET,UTC+1) | But those will not be resolvable | 19:12:33 |
| nightowlasu | i was 99% sure that was the case, but I seem to miss things so just wanted to verify! =D that makes sense, and I know is really the purpose of dataverse to make things findable! | 19:13:01 |
| nightowlasu | yes, the answer to some of these requests may be just to keep unpublished .. but being able to at least restrict files to ASU will be very helpful | 19:13:30 |
| poikilotherm (CET,UTC+1) | I would really love to see a valid use case for a hidden dataset... I don't think it's a good idea for the purpose of RDM, but there might be reasons out there... | 19:13:31 |
| poikilotherm (CET,UTC+1) | In reply to @nightowlasu:matrix.org
yes, the answer to some of these requests may be just to keep unpublished .. but being able to at least restrict files to ASU will be very helpful One of the reasons why we wanted mail domain groups for Jülich DATA (for datasets and collections, though). See and access unpublished ones. | 19:15:16 |
| nightowlasu | so you have unpublished datasets with files restricted to domain groups? | 19:18:34 |
| nightowlasu | which are shared with specific people via the reserved URL? | 19:19:18 |
| poikilotherm (CET,UTC+1) | Most likely not. There are some unpublished datasets, but those seem just to be lying around... | 19:29:04 |
| poikilotherm (CET,UTC+1) | We encourage people not to upload data in our repo, as we want to be a registry on purpose. | 19:29:26 |
| poikilotherm (CET,UTC+1) | Scientists are oblieged to register their datasets they might publish elsewhere in Jülich DATA, so we can one day generate statistics... | 19:30:07 |
| poikilotherm (CET,UTC+1) | And of course we might act as a last resort... | 19:30:26 |
| poikilotherm (CET,UTC+1) | Many of our scientists are having a hard time to actually share their data, but at least show the will to make them FAIR. | 19:30:58 |
| poikilotherm (CET,UTC+1) | (Aside from the immovable state of multi TB datasets) | 19:31:58 |
