Possible fix for "Floating islands inside of support"

src/TreeSupport.cpp

@@ -230,7 +230,7 @@ std::vector<TreeSupport::LineInformation> TreeSupport::convertLinesToInternal(Po
     // NOTE: The volumes below (on which '.inside(p, true)' is called each time below) are the same each time. The values being calculated here are strictly local as well.
     //       So they could in theory be pre-calculated here (outside of the loop). However, when I refatored it to be that way, it seemed to cause deadlocks each time for some settings.
     // NOTE2: When refactoring ensure that avoidance to buildplate is only requested when support_rest_preference == RestPreference::BUILDPLATE as otherwise it has not been precalculated (causing long delays while it is calculated when requested here).
-    
+
     std::vector<LineInformation> result;
     // Also checks if the position is valid, if it is NOT, it deletes that point
     for (const auto& line : polylines)
@@ -2679,7 +2679,7 @@ void TreeSupport::finalizeInterfaceAndSupportAreas(std::vector<Polygons>& suppor
         support_layer_storage.size(),
         [&](const LayerIndex layer_idx)
         {
-            support_layer_storage[layer_idx] = config.simplifier.polygon(support_layer_storage[layer_idx].unionPolygons().smooth(FUDGE_LENGTH));
+            support_layer_storage[layer_idx] = config.simplifier.polygon(support_layer_storage[layer_idx].unionPolygons().smooth(FUDGE_LENGTH)).getOutsidePolygons();
             // ^^^ Most of the time in this function is this union call. It can take a relatively long time when a lot of areas are to be unioned.
             //     Also simplify a bit, to ensure the output does not contain outrageous amounts of vertices. Should not be necessary, just a precaution.