Hi,
I have good and bad news :-)
The bad news: Yes this is a small bug, this should be handled more apropriate.
The good news: We can work around it!
The Decision Tree Learner node only uses column which have possilbe values in their domain saved. And only columns with little numerical columns have those assigned on default. To add the poss values please use the Domain Calculator and deactivate the option "Restrict number of possible values:" (or just set it very high)
However, this should be used with care. If you have a unique identifier the DT algorithm will use it as the first split, based on the information gain methodology.
I will let you know when the bug is fixed.
Best, Iris