Contributer & curator role in root database

41 views
Skip to first unread message

Erin MacPherson

unread,
Jun 27, 2017, 1:55:47 PM6/27/17
to Dataverse Users Community


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

Philip Durbin

unread,
Jun 27, 2017, 2:13:27 PM6/27/17
to dataverse...@googlegroups.com
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil



--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/f81f49b4-a431-487a-bea0-00a702772e90%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

Erin MacPherson

unread,
Jun 27, 2017, 2:21:48 PM6/27/17
to Dataverse Users Community, philip...@harvard.edu
Hi Phil,
I think it's just me :) I'm wondering if someone just wants to put a dataset in the root dataverse and publish it. 
Erin


On Tuesday, June 27, 2017 at 3:13:27 PM UTC-3, Philip Durbin wrote:
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil


On Tue, Jun 27, 2017 at 1:55 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Philip Durbin

unread,
Jun 27, 2017, 2:25:27 PM6/27/17
to dataverse...@googlegroups.com
Yeah, dunno. :)

Recently I added a "Going Live: Launching Your Production Deployment" and have sort of been thinking that maybe we should include a bit about how one might set up permissions before launch:

http://guides.dataverse.org/en/4.7/installation/config.html#going-live-launching-your-production-deployment

Just a thought. Pull requests welcome! :)

Phil

On Tue, Jun 27, 2017 at 2:21 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:
Hi Phil,
I think it's just me :) I'm wondering if someone just wants to put a dataset in the root dataverse and publish it. 
Erin

On Tuesday, June 27, 2017 at 3:13:27 PM UTC-3, Philip Durbin wrote:
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil


On Tue, Jun 27, 2017 at 1:55 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Durand, Gustavo

unread,
Jun 27, 2017, 2:27:08 PM6/27/17
to dataverse...@googlegroups.com, Durbin, Philip
In that scenario you would want them to be a Curator on that Dataset only. This can be accomplished manually for an individual dataset OR you can have it be true for any new Dataset created at that Dataverse, by changing the setting on the permissions page:

Curator - Edit metadata, upload files, and edit files, edit Terms, Guestbook, File Restrictions (Files Access + Use), Edit Permissions/Assign Roles + Publish


Note, this is not the same as assigning the Curator role to a user at the Dataverse itself. If you do that, then the user would be a Curator for ALL datasets in that dataverse.

Does that make sense?

Gustavo

On Tue, Jun 27, 2017 at 2:21 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:
Hi Phil,
I think it's just me :) I'm wondering if someone just wants to put a dataset in the root dataverse and publish it. 
Erin

On Tuesday, June 27, 2017 at 3:13:27 PM UTC-3, Philip Durbin wrote:
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil


On Tue, Jun 27, 2017 at 1:55 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

Erin MacPherson

unread,
Jun 27, 2017, 2:27:51 PM6/27/17
to Dataverse Users Community, philip...@harvard.edu
Hi again - to hopefully clarify, would assigning a curator role to someone at the root dataverse level allow them to publish another persons dataset in the root dataverse. I think that is where I am not clear.
Erin

Durand, Gustavo

unread,
Jun 27, 2017, 2:31:59 PM6/27/17
to dataverse...@googlegroups.com, Durbin, Philip
I think our e-mails crossed paths. :) Hopefully what I sent answered your question, but to be clear, yes it would allow them to publish another persons dataset in that dataverse.

What you want to do is change the setting I mentioned, which is not the same as assigning Curator role at the Dataverse level. That setting defines the role assigned by default to creators of new Datasets at the Dataset they created.


On Tue, Jun 27, 2017 at 2:27 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:
Hi again - to hopefully clarify, would assigning a curator role to someone at the root dataverse level allow them to publish another persons dataset in the root dataverse. I think that is where I am not clear.
Erin

On Tuesday, June 27, 2017 at 3:21:48 PM UTC-3, Erin MacPherson wrote:
Hi Phil,
I think it's just me :) I'm wondering if someone just wants to put a dataset in the root dataverse and publish it. 
Erin

On Tuesday, June 27, 2017 at 3:13:27 PM UTC-3, Philip Durbin wrote:
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil


On Tue, Jun 27, 2017 at 1:55 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

Erin MacPherson

unread,
Jun 27, 2017, 2:55:00 PM6/27/17
to Dataverse Users Community, philip...@harvard.edu
Hi Phil and Gustavo,
I think you have both answered my questions, thank you :) I'll keep it at the Contributer level at the root Dataverse, and change it to curator at the Dataset level. I think I just got confused with the different levels.
Thanks again, and also thank you all for the awesome community workshop!
Erin


On Tuesday, June 27, 2017 at 3:31:59 PM UTC-3, Gustavo Durand wrote:
I think our e-mails crossed paths. :) Hopefully what I sent answered your question, but to be clear, yes it would allow them to publish another persons dataset in that dataverse.

What you want to do is change the setting I mentioned, which is not the same as assigning Curator role at the Dataverse level. That setting defines the role assigned by default to creators of new Datasets at the Dataset they created.
On Tue, Jun 27, 2017 at 2:27 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:
Hi again - to hopefully clarify, would assigning a curator role to someone at the root dataverse level allow them to publish another persons dataset in the root dataverse. I think that is where I am not clear.
Erin

On Tuesday, June 27, 2017 at 3:21:48 PM UTC-3, Erin MacPherson wrote:
Hi Phil,
I think it's just me :) I'm wondering if someone just wants to put a dataset in the root dataverse and publish it. 
Erin

On Tuesday, June 27, 2017 at 3:13:27 PM UTC-3, Philip Durbin wrote:
Hi Erin,

My mental model for permissions in Dataverse is that each dataverse is an "island of permissions." By this I mean that there is no inheritance of permissions. The permissions don't leak out from the root dataverse to sub-dataverses. At the Harvard Dataverse, we let people publish datasets (and dataverses they create) in the root dataverse.

I'm curious about where the documentation is falling down for you. I know it needs a lot of work! https://github.com/IQSS/dataverse/issues/2653 is related but if you want to create a fresh GitHub issue in your own words, that would probably be better.

I hope this helps!

Phil


On Tue, Jun 27, 2017 at 1:55 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:


Hi everyone,

I'm looking for some feedback on how others manage their workflows within the root database.

The default role for anyone assigned is contributer. However, we want it to be a self-curated tool, so I would like researchers to be able to publish their own data. However, if I gave them a curator role, does this mean that they would be able to publish any other dataset within the root dataverse? We will be setting up multiple dataverses for the different groups, but I was thinking if someone just wanted to add data for one project, they could just add it to the main dataverse and publish it, but I only want them to be able to publish their own. I've read a lot of the documentation surrounding this but I'm wondering how do others handle this? 

Many thanks,
Erin

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/f81f49b4-a431-487a-bea0-00a702772e90%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Durand, Gustavo

unread,
Jun 27, 2017, 4:13:59 PM6/27/17
to dataverse...@googlegroups.com, Durbin, Philip
One thing about what you wrote: "I'll keep it at the Contributor level at the root Dataverse", I'm concerned that this may be not what you want.

If you mean assigning contributor role to a user at the Dataverse level, then that user would be able to edit (though not publish) all datasets. Again, any role assigned at the Dataverse level gets inherited by its Datasets. 

But again, if you mean that setting about default role on new datasets, then having that be contributor or curator is fine. But it seems like for your case, you'd want that to be Curator.



On Tue, Jun 27, 2017 at 2:54 PM, Erin MacPherson <Erin.Ma...@dal.ca> wrote:
Hi Phil and Gustavo,
I think you have both answered my questions, thank you :) I'll keep it at the Contributer level at the root Dataverse, and change it to curator at the Dataset level. I think I just got confused with the different levels.
Thanks again, and also thank you all for the awesome community workshop!
Erin

On Tuesday, June 27, 2017 at 3:31:59 PM UTC-3, Gustavo Durand wrote:

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

Sebastian Karcher

unread,
Jun 27, 2017, 4:22:46 PM6/27/17
to dataverse...@googlegroups.com, Durbin, Philip
The confusing thing is that "Contributor" appears in the Dataverse-level permissions:
Inline image 1
But it actually refers to the role the user has for datasets they added, not for the dataverse as a whole. It took me quite some time to figure that out. I do wonder if there's some better way to explain/label/conceptualize that, but I don't have anything constructive at the moment.

To post to this group, send email to dataverse-community@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Sebastian Karcher, PhD
www.sebastiankarcher.com

Durand, Gustavo

unread,
Jun 27, 2017, 4:27:14 PM6/27/17
to dataverse...@googlegroups.com, Durbin, Philip
Thanks, Sebastian. Yes, that is what I was trying to explain. I think it would be great if we could fine a way to express this. Maybe we just need to be extra explicit and say something like what you said: this role does not apply to the Dataverse level, but only to new Datasets created. (or something like that)

On Tue, Jun 27, 2017 at 4:22 PM, Sebastian Karcher <kar...@u.northwestern.edu> wrote:
The confusing thing is that "Contributor" appears in the Dataverse-level permissions:
Inline image 1
But it actually refers to the role the user has for datasets they added, not for the dataverse as a whole. It took me quite some time to figure that out. I do wonder if there's some better way to explain/label/conceptualize that, but I don't have anything constructive at the moment.
On Tue, Jun 27, 2017 at 4:13 PM, Durand, Gustavo <gdu...@iq.harvard.edu> wrote:
To post to this group, send email to dataverse-community@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Sebastian Karcher, PhD
www.sebastiankarcher.com

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

Erin MacPherson

unread,
Jun 27, 2017, 6:50:31 PM6/27/17
to Dataverse Users Community
Ah yes I think this is it - I want them to have a curator role. I agree with Sebastian - that's where my confusion set in - at the Dataverse level.
Thanks,
Erin
Reply all
Reply to author
Forward
0 new messages