Secondary categorical variable for box / violin plots

**What's your use case?**
We would like a way to visualize distributions across categories between subsets.

**What's your proposed solution?**
Essentially add a third dimension to produce graphs like:


https://seaborn.pydata.org/tutorial/categorical.html#violinplots

where the new variable would be equivalent to the `hue=sex` parameter.

**Are there any alternative solutions?**
Custom script into a seaborn visualization :)

I may be able to implement this, but I wanted to ask if it would be accepted first.

Read the linked seaborn categorical plotting example, especially the hue=sex behavior, and compare it with Orange's existing box and violin plot workflow. The issue names no repository files or tests; completion would support a secondary categorical variable that produces grouped distributions comparable to the example.

Written by the indexing model from the issue text.