Comment by TimonKnigge

I see -- thank you for clarifying. So for a formally uniform distribution, is rejection sampling your only option ?