Time
S
Nick
Message
00:56 jri joined #dataverse
06:55 jri joined #dataverse
07:36 poikilotherm joined #dataverse
08:18 stefankasberger joined #dataverse
09:50 jri joined #dataverse
10:02 jri_ joined #dataverse
10:18 stefankasberger joined #dataverse
11:13 jri joined #dataverse
11:28 jri joined #dataverse
12:14 jri joined #dataverse
12:32 jri joined #dataverse
12:51 pmauduit joined #dataverse
13:19 donsizemore joined #dataverse
13:35 jri joined #dataverse
13:51 jonas63 joined #dataverse
13:51 jonas63
whaddup homies
13:52 poikilotherm
Hi :-)
13:53 jonas63
just a quick one: i would like to keep visibility and access to some dataverses restricted to users from the local network. do you know of any way to achieve this in the current state?
13:53 poikilotherm
I remember some options to create groups from IP ranges
13:54 poikilotherm
http://guides.dataverse.org/en/latest/admin/ip-groups.html
13:54 poikilotherm
Maybe this can help you?
13:55 jonas63
that sounds perfect! in a second step, that IP Group would need the right "see dataverse" which currently does not exist, or does it?
13:56 poikilotherm
You will need to go to the dataverse you want to have restrictions on it and edit permissions. Then you should be able to choose a group and the ip group should show up in that list.
13:57 poikilotherm
You can assign different rights to that group, e.g. "see" as you asked above
13:59 jonas63
thank you, you are very helpful! i have never heard of (or seen) a "ViewPublishedDataset/verse" permission though....
14:00 jonas63
and allowing the IP Group to ViewUnpublishedDataset/verse would go too far
14:02 jonas63
if you don't have an idea, never mind, Oli!
14:03 jonas63
you have been very helpful already :D
14:03 poikilotherm
Could you maybe elaborate a bit more on your idea? I am not sure what you are trying to achieve...
14:04 poikilotherm
Or maybe raise an issue if you think sth. is missing from the permission system. pdurbin is not around, but I am sure someone else will pick up and respond on Github
14:05 jonas63
three levels of access: a) external users: see published datasets/verses in the two public dataverses, b) internal users: see published datasets/verses in all dataverses, c) owners (can be people or units): see unpublished datasets/verses they own
14:06 jonas63
yes, i intended to open an issue, but found that it's a good idea to talk about it a bit first.....
14:07 poikilotherm
I don't think this level of granularity does exist yet...
14:08 poikilotherm
When you are using DOIs, this will not fit at all
14:08 poikilotherm
Because the metadata is published at DataCite/EZID anyway
14:08 jonas63
well, i think the way of thinking is "public is public", which kind of makes sense, but does not facilitate internal sharing
14:08 poikilotherm
So you would like to see some kind of "internally published"?
14:09 poikilotherm
Maybe this is sth. solvable once DataTags are implemented?
14:09 poikilotherm
That permission level definitly does not exist yet
14:09 poikilotherm
Maybe you can use a workaround
14:09 poikilotherm
Permissions are not inherited by design
14:10 jonas63
we are using HDL --- "real" publications happen in other repositories and we just use the DOI from there
14:10 poikilotherm
You can inherit on group creation when you enable a certainly new option
14:10 poikilotherm
But you could create a Dataverse which has your group added and leave things unpublished there
14:11 poikilotherm
Currently, you can move Datasets via API , UI support is in the works
14:13 poikilotherm
http://guides.dataverse.org/en/latest/admin/dataverses-datasets.html#move-a-dataset
14:15 poikilotherm
The UI support progress is tracked here https://github.com/IQSS/dataverse/issues/2073
14:16 poikilotherm
But as pdurbin likes to say: please open an issue if you feel like it ;-)
14:29 jonas63
thank you so much for the input! i am planning to base it all on scripts calling the API anyways...
14:30 pameyer joined #dataverse
14:30 jonas63
https://github.com/IQSS/dataverse/issues/5764
14:31 pameyer
@jonas63 your levels of access sound close to some ideas that have been kicked around; but not anything that's directly supported
14:31 pameyer
"internal publication" is a tricky-ish thing
14:32 jonas63
yeah, let's not call it "publication" - that's just confusing
14:33 poikilotherm
Maybe call it staging :-D
14:33 poikilotherm
Like in the Linux kernel ;-)
14:33 pameyer
unfortunately, I don't have a better work for "change this thing that only I can see to things everybody at my institution can see"
14:34 pameyer
;)
14:34 jonas63
i used "internal sharing" in the issue
14:34 poikilotherm
Well you could use a special dataverse with group permission to see unpublished
14:34 jonas63
maybe "instititutional sharing"?
14:35 poikilotherm
Everything you want to share internally / institution wide could be moved over there
14:35 poikilotherm
s/Everything/Every dataset/g
14:35 pameyer
"sharing" seems to me to be more confusing - the "share" button in dataverse does social media rather than permissions
14:36 jonas63
right......
14:36 pameyer
@poikilotherm we'd been thinking along the lines of "not-publication" as moving between private dataverse, less private dataverse, and public dataverse
14:37 pameyer
I thought "publication funnel" was a cool buzzword; but I'm bad at picking buzzwords
14:37 poikilotherm
It could also be an extension to "submit for review"
14:37 jonas63
the example i am thinking about:
14:37 poikilotherm
Because sharing interally might be relevant for such use cases where reviews are not just done by curators/admins
14:44 jonas63
how about "InstitutionalAccess"
14:45 jonas63
or "RestrictedVisibility"
14:47 pameyer
InstitutionalAccess sounds good to me (for what that's worth ;) )
14:47 pameyer
@poikilotherm good idea about mixing the review process into this; I don't think that was an angle that had been considered
14:48 poikilotherm
Thx pameyer :-)
14:48 poikilotherm
I try to leave a comment tomorrow on the issue
14:48 poikilotherm
Have to go now, pick up kiddies
14:48 poikilotherm
Cu guys
14:48 jonas63
cu oli, and thank you so much!
15:07 donsizemore
@jonas63 could you do this with an institutional group, say read-only or member role?
15:08 jonas63
the thing is: there is still stuff that should not be visible to anyone but the owner...
15:09 donsizemore
@jonas63 i believe at present you'd need shibboleth for that. i think the piece you're missing is file restriction once you assign roles to the dataset
15:10 donsizemore
@jonas63 you could probably do what you want with a combination of shib group and http://guides.dataverse.org/en/latest/user/dataset-management.html?highlight=roles but more robust permission sets are always a good thing
15:11 jonas63
yeah, we're not gonna have shib for the foreseeable future
15:11 jonas63
but also, i don't understand how that would help me to restrict visibility of a dataset
15:13 donsizemore
"If specific files in your dataset are restricted access, then you can grant specific users or groups access to those files while still keeping them restricted to the general public. "
15:15 jonas63
yes, for files... but that doesn't apply to the dataset's metadata or (even more importantly) the places where this dataset will be listed
15:15 jonas63
right?
15:15 donsizemore
correct
15:22 jonas63
don, i would love to see your input on https://github.com/IQSS/dataverse/issues/5764 if you have the time
15:24 donsizemore
we've run into a similar issue with harvesting but it's our metadata librarian who didn't like DV's current behavior
15:24 donsizemore
this definitely seems like something that should go into 5. about to head out for lunch here, but i'll look at it after?
15:25 jonas63
thank you!! enjoy your meal!
15:27 donsizemore
hope so. today is our first above 26°
15:52 jonas63
thanks for using °C :)
16:56 donsizemore joined #dataverse
17:10 jri joined #dataverse
18:34 Amber59 joined #dataverse
