You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
After adoption of some code from AgroPortal, there's a new behavior where all the tabs on ontology landing pages are hidden except for the Summary tab if the latest submission failed to parse. In my opinion this causes a serious regression where users can no longer access the ontology data from the last successful processing run.
Prior to adopting AgroPortal's code, all the tabs on ontology landing pages were visible at all times. In the event a new submission was uploaded that failed to parse, BioPortal users could still navigate to the Classes tab and browse the last good version of the ontology data.
For a specific example, see the FOODON ontology. Submission with ID 103 successfully processed in BioPortal. The next submission (with ID 104) failed to parse and is listed in the Submissions table with an "Error Rdf" status. Since the Classes tab is hidden, there's no way for a user to browse the classes from the last good submission. Even worse, if you click on the link in the Submissions table for the last successfully processed version, you end up on a page that looks like this:
This behavior is inconsistent with how our REST API operates. In other words, even though the latest submission of FOODON failed to parse, you can still issue a REST call to successfully retrieve the ontology classes from the last valid submission: https://data.bioontology.org/ontologies/FOODON/classes.
If possible, I think implementing a solution for this should be prioritized. Users shouldn't be prevented from browsing ontology data if the most recent submission failed to parse, particularly since the ontology data from the previous good submission is still accessible via the REST API.
The text was updated successfully, but these errors were encountered:
You can remove or adapt that condition to have the desired behavior.
If you want the previous behavior, you can remove that condition.
If you want the AgroPortal behavior, which is to show the other data sections(classes, properties,...) only if the current or previous submission is ready you need to fix it because the issue I think is that in the AgroPortal alignment work, the variable @old_submission_ready was not assigned. it needs to be added like this https://github.com/ontoportal-lirmm/bioportal_web_ui/blob/master/app/controllers/ontologies_controller.rb#L288-L295 in bioportal ontologies_controller show action.
After adoption of some code from AgroPortal, there's a new behavior where all the tabs on ontology landing pages are hidden except for the Summary tab if the latest submission failed to parse. In my opinion this causes a serious regression where users can no longer access the ontology data from the last successful processing run.
Prior to adopting AgroPortal's code, all the tabs on ontology landing pages were visible at all times. In the event a new submission was uploaded that failed to parse, BioPortal users could still navigate to the Classes tab and browse the last good version of the ontology data.
For a specific example, see the FOODON ontology. Submission with ID 103 successfully processed in BioPortal. The next submission (with ID 104) failed to parse and is listed in the Submissions table with an "Error Rdf" status. Since the Classes tab is hidden, there's no way for a user to browse the classes from the last good submission. Even worse, if you click on the link in the Submissions table for the last successfully processed version, you end up on a page that looks like this:
This behavior is inconsistent with how our REST API operates. In other words, even though the latest submission of FOODON failed to parse, you can still issue a REST call to successfully retrieve the ontology classes from the last valid submission: https://data.bioontology.org/ontologies/FOODON/classes.
If possible, I think implementing a solution for this should be prioritized. Users shouldn't be prevented from browsing ontology data if the most recent submission failed to parse, particularly since the ontology data from the previous good submission is still accessible via the REST API.
The text was updated successfully, but these errors were encountered: