Closed ways closed 1 month ago
It would be quite confusing for the user if they have to check every response for a license link in the case that some response superseeds the license in collection as a whole.
I would much prefer a rule like: "A collection SHALL have a license (public domain or CC by 4). All data delivered as part of this collection is assumed to be under this license". @mrauhala @Teddy-1000 Is it possible to setup RODEO so that its one license pr. collection? Which would probably mean more than one collection for e-soh? If not, how should we advice the user how to handle this?
I think we can advise the implementer to split data in to several collection if the data in a collection have incompatible licenses, if all licenses are a form of free use, having the most restrictive one should suffice. We could also specify a data query dimension where you can select a list of licenses. For this profile, it is for RODEO, I suggest we assume most, if not all, data will be free to use.
I have another question related to this. Say we set a license of CC by 4 for the rodeo open data collections . This will then in some cases (like E-SOH) include data from many countries. How do we practically fulfill the requirements of attribution? Asking people to attribute "Rodeo" does not make a lot of sense I think? Or do we need to list all institutes who supply data to the collection?
This would be a lot easier if we release the data to the public domain (CC0), but I assume that is not really an option here.
In E-SOH all observations are required to provide the naming_authority
, and can provide creator_name
, institution
. This metadata is attached to every observation. We could require this property to be included in all data replies.
Sounds doable. And in the end, it might be enough with an attribution to EUMETNET members or something like that. I will close this.
Should all data within a collection be required to have the same license?