Move embargo, co_authors, and shared_with from entry to upload
We want to provide uploads with a standard owner, read, write permission set. Those overlap with our previous co_authors
and share_with
metadata key. We should introduces the former on an upload level and remove the latter from the entry level.
What about the current data?
- no* mismatching embargo that is newer than 3 years, OK
- only 3 uploads with mismatching
shared_with
that are newer than 3 years (for lifted embargo shared_with is meaningless), OK - ~1000 uploads with mismatching co_authors, NOT OK
We can move with_embargo
and shared_with
. We should introduce an additional entry-based additional_authors
key to capture co authors that have not directly participated in the upload.
Maybe we should have a metadata
field on uploads to distinguish the metadata from processing, id, etc. keys?
What should the keys be?
-
uploaders
|owners
,authors
|co_authors
,reviewers
,additional_authors
(on entries) -
owners
,writers
|editors
,readers
- do we need a dedicated single creator or is owners enough?
- (
with_embargo
,embargo_period
) |embargo
,publish_time
Actions:
- create (authors)
- delete (authors)
- publish (authors)
- upload stuff, delete stuff, move stuff (authors)
- change metadata (upload + entries) (authors)
- see stuff (authors, shared_with, everybody if published without embargo)
Transformation
- *authors = [uploader]
- *shared_with = [Calc.shared_with for Calc in calcs]
Example:
upload: {
upload_id: ...,
process_status: ...,
errors: [],
...
metadata: {
upload_name: ...
authors: [],
shared_with: [],
embargo: 12,
create_time: 114344324234,
publish_time: 114344324234
}
},
entry: {
entry_id: ...,
parser: ...,
process_status: ...,
errors,
...,
metadata: {
comment: ...,
references: [],
datasets: [],
co_authors: []
}
}
Edited by Markus Scheidgen