In the attached document model (BuchmachersteuerWithWorkflowsMetadata) we have the field “Status” and “Erledigungsart” for which we have set up a dependency in the associated form model, (Abschluss, see json screenshot: pink is Status, blue is Erledigungsart).
However, in the SME, I do not see this dependency (despite the icon indicating that there is a dependency)
How is this possible?
BuchmachersteuerWithWorkflowsMetadata_4.json (841.5 KB)
Buchmachersteuer-Abschluss_4.json (34.5 KB)
You haven’t defined what values are allowed when your status is “stornierung”. I assume that makes the dependent enumeration invalid (to the SME) and the SME refuses to display it correctly. I manually added it to the json and that fixed it.
Buchmachersteuer-Abschluss_4_fixed.json (33.4 KB)
I compared the two jsons. Nothing was added. Instead, what you seemingly did was to remove a few lines that referenced the same field, because it had a double mention in the field configuration: one with the dependency and one without. No idea how that got in there, but I am surprised that the SME didn’t flag the json as invalid.
So it doesn’t seem to have anything to do with whether the json lacked constraint values for “stornierung”. The SME seems to interpret a lack of constraint values as simply allowing all Erledigungsarten. Had the dependent enum actually been invalid, our application would crash at the startup and the SME would probably not have considered this json to be a valid form model and rejected it outright.
But you in the end did give me a working Json. 
Thanks!
Huh, weird. I did add the line for “stornierung” to the json and that’s how I got it to display in the SME initially and then saved once more. Maybe I also unintentionally removed the double reference when doing that and then the SME “shortened” the line for “stornierung” away again because it didn’t restrict any values. But if it’s fixed now, that’s great 
Hi @jasper-silent-root, @lea-clear-clover
Could you confirm how you were editing the json?
The SME should be expected to catch errors that you created when editing models in the SME and show the modeled settings correctly. However, I wouldn’t expect the SME to catch all possible errors in a json file or display elements added to the json that were added using a text editor if they do not exactly match the json format produced by the SME.
Adding the dependent enumeration in the SME appears to add the dependency correctly.
Yeah, I don’t think that my case is relevant necessarily. I just had a mystery that needed solving to alleviate my curiosity,