This is more tricky than I thought.
Accepting both comma and period as decimal seperators only works if they are no longer accepted as thousands seperators. Otherwise 1,000 and 1.000 could both mean 1 or 1000.
I tried to make it two different modes by adding a workspace option *useDecimalComma*. In comma-mode, the number field should display and accept as input numbers like 1,2 and reject 1.2. Now I had a problem with the field's text/value distinction. The field's value should be used for code generation and block serialization, while the field's text should be displayed to to user. So, in comma-mode, a text of "1,2" should correspond to a value of "1.2".
But the number field inherits from text field, where value and text are always the same. It sets the field value directly from the user's input. I tried to add a method *setValueFromUserInput*, where the number field can do comma replacement. That works, but turned into more changes than I'm comfortable with.
For now I'll go with the simple solution of accepting both comma and period input and displaying only periods.