search

pkp / pkp-lib

The library used by PKP's applications OJS, OMP and OPS, open source software for scholarly publishing.
https://pkp.sfu.ca
GNU General Public License v3.0
297 stars 443 forks source link

Consider how to handle items with no DOIs when depositing all #7524

Closed NateWr closed 1 year ago

NateWr commented 2 years ago

Describe the problem you would like to solve When I clicked Deposit All in the new DOI management page (#6062), I expected all Unregistered items to be switched to Submitted. However, one of them did not because it did not have a DOI assigned. This makes sense in hindsight, but it was confusing in the UI and at first I thought something was broken.

Describe the solution you'd like I think we need a way to indicate whether a submission has a DOI assigned or not. I suggest the following reorganisation of the filters in the DOI list. This separates the filters to track two statuses: 1) the item's status and 2) the DOI's deposit status.

doi-filters

The first set of statuses would indicate how ready a submission is to be registered:

The second set of statuses specifically track the registration status. They are the same as before, but I've changed a couple of the names of filters. Error is now Has Error and Stale is now Needs Sync. These changes only need to occur in the UI, not in the names of the statuses in the code base.

Who is asking for this feature? This is part of #6062.

Additional information cc @ewhanson

ewhanson commented 1 year ago

Hey @bozana, would you be able to take a look at these? It makes some of the filters more specific as well as groups them differently as described above. Thanks!

PRs:

bozana commented 1 year ago

Hi @ewhanson, just two comments/questions :-) Else, ready for merging...

ewhanson commented 1 year ago

Thanks, @bozana. Merged!

NateWr commented 1 year ago

@ewhanson I'm just looking at this now to prepare for the change orientation presentation, and I notice that the first three filters are in a different order:

Did you and Bozana discuss this? I was scratching my head a bit to figure out what each one meant. The filters seem to overlap -- so one submission might appear in multiple filters. I thought our original intent was to have distinct filters so a submission could only be in one "state" at a time. When they appear in both, the rules become more complex to understand.

What do you think about dropping "Unpublished"? In my view, Needs DOI and DOI Assigned are easy to understand and mutually exclusive. It doesn't really matter if they are unpublished or not, does it? Once something is "published", the editor's interest moves on to the registration filters.

I'd suggest this for the descriptions of the filters:

Filter Description
DOI Missing All items missing a DOI.
DOI Assigned All items assigned a DOI.
Unregistered All items with a DOI that have been published but not yet deposited with a registration agency.
Submitted All items that have been submitted to a registration agency.
Registered All items that have been registered with a registration agency or manually marked as registered.
Error All items that have encountered an error in the registration process.
Stale All items that have been republished since they were last deposited with a registration agency. They need to be resubmitted to the registration agency to update their metadata records.
ewhanson commented 1 year ago

Hey @NateWr, thanks for catching the different ordering. That one slipped through! By different ordering, did you mean "Unpublished" and "Needs DOI" being flipped?

I know we'd previously talked about the need to identify which items had DOIs missing, but I don't recall if we went into detail about the filters specifically.

I agree that we can probably drop "Unpublished." I think the more important thing that was missing previously was identifying which items did not have DOIs assigned, but that's covered now. I also didn't like how "Unpublished" and "Unregistered" had a conflicting filter, so that would solve that neatly.

The text changes look good to me as well. I'll get those updated and post the PRs here.

ewhanson commented 1 year ago

Hey @NateWr, here are the PRs removing the "unpublished" filter and updating the working in the help text. Thanks!

PRs:

NateWr commented 1 year ago

Looks good, I didn't do a manual test but the code changes all look clean enough. When the tests pass feel free to merge.

(This is a one line change you might want to chuck into this PR: https://github.com/pkp/pkp-lib/issues/8677).

ewhanson commented 1 year ago

Merged!