App Issues
(list updates daily)
Unified unit search field
To be able to search across records with different (but compatible) units, one of two things should be done.
- Add A field with a converted (unified unit) value to the search index.
- Add A field with a converted (unified unit) value to the record search index.
It should be noted that not all commonly used units can be directly converted to each other.
Render a full-view of the record
null
Auto save functionality
We might consider auto saving the record every 5-10 minutes to prevent data loss
Make validation errors links
It should be made easier for the user to navigate to where particular validation error has occurred. This could be done by turning the errors into links
Undo functionality
It would be useful with undo functionality in case of accidentally deleting element or similar
Extract name from ORCID
If an ORCID is given for a person, the given name and family name can be extracted:
REQUEST: GET
HEADER: "Accept: application/json"
URL: "https://pub.orcid.org/v3.0/<ORCID>"
The required information can be obtained from response to the request in the fields:
person.name.given-names.value
person.name.family-name.value
Edit: This will either require a change in the app’s permission policy, or creating a backend endpoint the fronted can use
Add state as a tag to the record listing
It would be useful display the status of the records that are not publicly visible (draft, submitted, accepted, retracting) in the search result record listing.
The state
field is at the top level in the record
copy_number is not clear nomenclature
considering using a different word or longer description.
SPR Data analysis, tooltip
In SPR Data Analysis The first drop-down Data analysis 1 Measurement 1 has tooltip “List of measurement links that was analyzed together”
Should this read rather “Measurements that were analyzed together” ?
NOT Critical
SPR Meas protocol Flow Path - graphics misbehaviour
When adding under SPR method Measurement protocol Flow Path Picking a path from the list, OK, then click into the next field and the title of the Path field goes back into the main field instead of staying minimized on the top frame.
NOT Critical
Record view - width of the graphical representation
The width is now dynamically adjusted according to the content. This is generally fine. However, I would suggest that by default it should not shrink below a certain width.
NOT CRITICAL - JUST DISTURBING when I compare two or more records.
Measurements are tedious to input
It is annoying to always repeat the same values in the case of the measurements. For example, the name is the same, the concentration of the target, the chemical environment and the ‘container’ …
MST datanalysis define Fhot and Fcold in single line
In ‘Data analysis’, we believe that using Cold and Hot region a single line for each value would be more appropriate than F cold or F hot: Cold region: start (field to fill) end (field to fill) Hot region: start (field to fill) end (field to fill)
Include option reuse person information in more sections
It could be great to include an option to indicate that the ‘principal contact’ is the same as the ‘depositor’ in order to not repeat information and reduce the number of fields the user has to complete manually.
Vocabulary API performance
Performance of the vocabulary API appears to be slow, particularly when a larger number of items is requested.
On the testing setup that I have available there are 1003 items present in the “organisms” vocabulary. mbdb-app is running in development mode in a VM form a mechanical disk. The VM is accessed via a guest-host network bridge. A request to fetch all 1003 items from the vocabulary often fails with a timeout error. The timeout is set 15 (fifteen) seconds, which should be more than enough to retrieve around ~1000 entries. If the timeout is reduced even further, (say to 1500 ms) even queries with the ?suggest
parameter begin to time out.
The “affiliations” vocabulary (with 1000 entries) fares a little better but a fetch of the entire vocabulary still takes over 5 secs.
These results are very well reproducible.
The questions that should probably be asked:
- Why is there such a performance disparity between two vocabularies that have the same number of entries?
- Is this performance expected or is this some defect of the development setup?
Deposition form - autofilling owner infomation and usage information
Tester notes that:
Depositor and Principal Contact
Is it possible to have a button to autofill the information from the owner information? It would save time!
I wonder if people will really fill out all of these (owner, depositor, principal contact) with different information… how do you see this information being used?
Once someone is registered in the database, will they be searchable to include automatically, or are you relying on them entering orcid to be able to link all of the records associated with that person?