The collection comparer can only be added to _selectedValues state once #8661 is merged.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 90.14%. Comparing base (28bc599) to head (6c6bd42).
Report is 75 commits behind head on dev.

@@            Coverage Diff             @@
##              dev    #8722      +/-   ##
==========================================
+ Coverage   89.82%   90.14%   +0.31%     
==========================================
  Files         412      419       +7     
  Lines       11878    12181     +303     
  Branches     2364     2396      +32     
==========================================
+ Hits        10670    10981     +311     
+ Misses        681      664      -17     
- Partials      527      536       +9     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Added to v7.0.0 Migration Guide #8447

This one is done

Tests failed, I wonder if it's related

Tests failed, I wonder if it's related

Yes. My implementation completely broke down, not sure why, but seeing how unstable it was I had to overhaul it again. I had to do it like in TreeView, that in single selection mode SelectedValues is ignored and in multi selection SelectedValue is ignored.

Yes. My implementation completely broke down, not sure why, but seeing how unstable it was I had to overhaul it again. I had to do it like in TreeView, that in single selection mode SelectedValues is ignored and in multi selection SelectedValue is ignored.

Yea, it seems just adding comparer to SelectedValues break everything and the implementation requires UpdateSelectionAsync invoke when it shouldn't, but I see that my code and the CollectionComparer works correctly when you compare the x, y

Btw, not related question, but why such complicated GetHasCode?
Wouldn't this just work?

public int GetHashCode(IReadOnlyCollection<T> obj)
{
	var hasher = new HashCode();
	foreach (var item in obj)
	{
		if (item is not null)
		{
			hasher.Add(_comparer.GetHashCode(item));
		}
	}

	return hasher.ToHashCode();
}

Why does it require Distinct and an OrderBy (don't we care only about the sum of the items)? Shouldn't the items be unique anyway if the derivering type of IReadOnlyCollection is a HashSet?
Upd: looking at your test I see why you use Distinct

As I understand it hashing algorithms are not commutative. So if you hash [1, 2, 3] and [3,2,1] you'll get different results.

And we don't know if the user supplies a hash set. IReadOnlyCollection can be anything, such as Array, List, etc