WIP: Simulation Nodes: Implement subframe mixing

Add an implementation for the placeholder to subframe mixing.
The purpose of subframe mixing is higher quality motion blur,
without requiring baking the entire geometry output for every
substep. Linear mixing can fill the gaps while maintaining lower
memory usage. All attributes are mixed, besides builtin topology
storage attributes. Mixing is only supported when the domain size
is unchanged, or when an `id` attribute gives a mapping between
items when the topology is changed. Attributes that are ignored
for better performance.

TODO?
- [ ] ID attribute based mixing on domains besides points
- [ ] Mixing geometries from nested instances
- [ ] Testing and demo file

@@ -369,0 +377,4 @@
+  if (a != b) {
+    return false;
+  }
+  if (a->version() != b->version()) {

@@ -369,0 +645,4 @@
+      }
+
+      void *prev = value_or_field_type.get_value_ptr(prev);
+      const void *next = value_or_field_type.get_value_ptr(next);