EDIT

This generally works. During review I only wonder where to put it and how to handle it best.

My suggestion is to move it directly below "Reference" as one should either select a reference OR a specimen for the source. Also we should label "or Specimen" to make clear that it is an alternative and should not be used both (except for the rare case where a specimen is only a figure in a reference but this should be handled differently, there is a ticket for this).

Is it possible to make this "either .. or" mandatory by deleting reference if specimen is selected and vice versa? I could also add a strict validation rule that allows only either reference or specimen (I will do this in #10194). We should implement this now to avoid that we must handle incorrect data in future.

Andreas Müller wrote in #note-9:

Katja Luther wrote in #note-8:

Andreas Müller wrote in #note-7:

This generally works. During review I only wonder where to put it and how to handle it best.

My suggestion is to move it directly below "Reference" as one should either select a reference OR a specimen for the source. Also we should label "or Specimen" to make clear that it is an alternative and should not be used both (except for the rare case where a specimen is only a figure in a reference but this should be handled differently, there is a ticket for this).

Is it possible to make this "either .. or" mandatory by deleting reference if specimen is selected and vice versa? I could also add a strict validation rule that allows only either reference or specimen (I will do this in #10194). We should implement this now to avoid that we must handle incorrect data in future.

Remove and add elements are always a little bit tricky in rcp, I would implement it with enabling and disabling the fields and adding "or " to the label for this release and for next release I already started implementing the removing and adding of the fields.

Sorry, I was not exact. I didn't mean to remove the field but the value. But enabling and disabling is the much better solution I realize. If this works we can leave it like this.

This is implemented now. Please review.

When recreating details view the state is not yet evaluated. So if a reference exists already it is still possible to enter a specimen.

It works fine if you enter a new reference/specimen and do not set the focus elsewhere.

Andreas Müller wrote in #note-11:

When recreating details view the state is not yet evaluated. So if a reference exists already it is still possible to enter a specimen.

This is fixed.

works as expected now