Describe the bug
The identifier used to identify references is not linked to an account but is machine specific, meaning that using the same account on a different machine results in the identifiers being different.
To Reproduce
Steps to reproduce the behavior:
Synchronize Zotero on a different machine
Open application
Observe that references are different
Expected behavior
References should work as long as Zotero is linked to the user's account and synchronized.
It should not matter what machine they use.
Additional context
So we should instead pick an identifier in the Zotero database, if there is any, which is unique to the item and machine-independent (not the row index of the item, which is what we appear to be using now).
It would also be nice if instead of using the identifier directly in entries, we have some intermediate mapping, so that we can assign additional properties to references in the future if we so choose (priority, importance, etc.).
Describe the bug The identifier used to identify references is not linked to an account but is machine specific, meaning that using the same account on a different machine results in the identifiers being different.
To Reproduce Steps to reproduce the behavior:
Expected behavior References should work as long as Zotero is linked to the user's account and synchronized. It should not matter what machine they use.
Additional context So we should instead pick an identifier in the Zotero database, if there is any, which is unique to the item and machine-independent (not the row index of the item, which is what we appear to be using now).
It would also be nice if instead of using the identifier directly in entries, we have some intermediate mapping, so that we can assign additional properties to references in the future if we so choose (priority, importance, etc.).