Originally reported by: Marcel Bollmann (Bitbucket: mbollmann, GitHub: mbollmann)
Right now, tagsets can be (and mostly are) referenced by their class. This is definitely true for the client-side, but possibly the server side as well.
Referencing tagsets by ID would have the advantage that multiple tagsets of the same type could be linked to a text, and generally seems like the saner approach.
However, this would probably involve a lot of work:
Identify all places in the code with class-based references to tagsets, and figure out how to convert them to ID-based refs. Tagsets would never be allowed to be referenced by their class anymore.
The EditorTable would have to be generated dynamically since we cannot know in advance how many occurences of a class we need. This could be more difficult than it sounds, since code might refer to the static IDs/classes that are currently defined for EditorTable components.
CorA-XML annotation tags would have to be modified to include tagset ID, since tagset classes alone are no longer guaranteed to be unique.
It's unclear if the benefits justify the huge amount of work this would require.
Originally reported by: Marcel Bollmann (Bitbucket: mbollmann, GitHub: mbollmann)
Right now, tagsets can be (and mostly are) referenced by their class. This is definitely true for the client-side, but possibly the server side as well.
Referencing tagsets by ID would have the advantage that multiple tagsets of the same type could be linked to a text, and generally seems like the saner approach.
However, this would probably involve a lot of work:
It's unclear if the benefits justify the huge amount of work this would require.