Thanks for flagging dan. This is an interesting question of distinguishing restricted access definition from "under certain use restrictions" to "certain audiences." I tend to think that the definition of "under certain use restrictions" would leave more room for additional use cases. Do others have strong opinions on this?

Use restrictions are covered by the license property, as explained by the open licenses page in POD (http://project-open-data.github.io/open-licenses/).

Looking back at this, I understand the logic for both phrases. I'd recommend "to certain audiences or under certain use restrictions" to cover both angles. Would that work with folks?

@haleyvandyck Can you provide an example of something that could be publicly available, but have restrictions on its use?

@dsmorgan77 While your amendment might be honing in on the original intent, it introduces some confusing language by saying it is both public and limited to certain audiences.

Thanks @gbinal and @philipashlock. Gray, the intent of the change is to clearly separate where we talk about who can access (audience restrictions) and what they can do with it (use restrictions). My thought is that use restrictions are clearly handled under the license. Phil, I'm trying to say exactly that something can be public and limited to certain audiences - I think that's the intent of having an access level that's called "restricted public." The best case I can think of there would be certain research datasets...when the government issues a grant to a researcher, that researcher can set terms on access to those data (it's not the government's data...especially in the case of historical grantee agreements). So if a researcher said a dataset might be available only to other qualified researchers, then it's available to "certain audiences." I think I'd also classify things like restricted access Census data that's only available at a Census Research Data Center (http://www.census.gov/research/data/restricted_use_microdata.php).

In summary, the pull request clarifies the meaning of "restricted public" to account for those kinds of scenarios where something is not available to any audience, without restriction...but that is also not withheld from the public. It doesn't change the meaning of "public" and doesn't try to address use restrictions in any way.

a very simple use restriction to this stellar dataset is: "Scientific accuracy is not guaranteed. Please do not use this visualization for interstellar navigation". this has not much to do with licensing of the data, but more with purpose, quality, and perhaps liability.

So to summarize, the intent here is to separate these so that:

• accessLevel is not used to indicate usage rights (use restrictions) whatsoever, only access rights (audience restrictions).
• Usage rights should instead only be indicated through licensing.

Is that a fair characterization @dsmorgan77?

I still think "restricted public" is a confusing way of saying a subset of the public. That terminology does make a little more sense when used to convey restricted usage rights, eg read more like "restricted; public" but I like the idea of separating these out.

@philipashlock: exactly right. I don't like the term "restricted public," either, but I figured that debate was for another day. :-)

@mhogeweg I see where you're coming from there. At the risk of splitting hairs, I think that's more of a usage note (a non-binding suggestion) than a restriction. I have a dataset at my agency where I have to show a disclaimer from 23 USC 409 (http://www.gpo.gov/fdsys/pkg/USCODE-2011-title23/html/USCODE-2011-title23-chap4-sec409.htm).

The schema doesn't allow me to really link to that using any explicit field, but my plan is that I'll included that as one of my references.

My agency has both restricted access (a public data set versus one that like the Census research labs, requires restricted or controlled access subject to security provisions of an agreement between the data requestor and the Director of my institute. We also have restricted uses, insofar as any data collected under a pledge of statistical confidentiality can only be used for research/statistical purposes and are prohibited by law from "... legal process and shall not, without the consent of the individual concerned, be admitted as evidence or used for any purpose in any action, suit, or other judicial or administrative proceeding."

I believe this has been addressed in the v 1.1 updates with the discussion of the rights field. See: #353 Feel free to reopen this if prose can reinforce those conclusions.