Fix dependent parameter division expressions triggering infinite recursion (develop)

src/easyscience/variable/descriptor_number.py

@@ -423,16 +423,13 @@ def __rmul__(self, other: numbers.Number) -> DescriptorNumber:
 
     def __truediv__(self, other: Union[DescriptorNumber, numbers.Number]) -> DescriptorNumber:
         if isinstance(other, numbers.Number):
-            original_other = other
             if other == 0:
                 raise ZeroDivisionError('Cannot divide by zero')
             new_value = self.full_value / other
         elif type(other) is DescriptorNumber:
-            original_other = other.value
-            if original_other == 0:
+            if other.value == 0:
                 raise ZeroDivisionError('Cannot divide by zero')
             new_value = self.full_value / other.full_value
-            other.value = original_other
         else:
             return NotImplemented
         descriptor_number = DescriptorNumber.from_scipp(name=self.name, full_value=new_value)

src/easyscience/variable/parameter.py

@@ -877,7 +877,6 @@ def __truediv__(self, other: Union[DescriptorNumber, Parameter, numbers.Number])
                     combinations = [self.min / other.min, self.max / other.max, self.min / other.max, self.max / other.min]
             else:
                 combinations = [self.min / other.value, self.max / other.value]
-            other.value = other_value
         else:
             return NotImplemented
         min_value = min(combinations)
@@ -926,7 +925,6 @@ def __rtruediv__(self, other: Union[DescriptorNumber, numbers.Number]) -> Parame
         parameter = Parameter.from_scipp(name=self.name, full_value=new_full_value, min=min_value, max=max_value)
         parameter._convert_unit(parameter._base_unit())
         parameter.name = parameter.unique_name
-        self.value = original_self
         return parameter
 
     def __pow__(self, other: Union[DescriptorNumber, numbers.Number]) -> Parameter:

tests/unit_tests/variable/test_parameter.py

@@ -388,6 +388,51 @@ def test_dependent_parameter_return_is_descriptor_number(self):
         assert dependent_parameter.min == 2*descriptor_number.value
         assert dependent_parameter.max == 2*descriptor_number.value
 
+    def test_dependent_parameter_division_expression_order(self):
+        """Test that division expressions work regardless of operand order.
+
+        This is a regression test for https://github.com/easyscience/corelib/issues/190
+        where 'a / b' would fail with an observer notification error while '1/b * a'
+        would work correctly.
+        """
+        # When
+        angstrom = DescriptorNumber("angstrom", 1e-10, unit="m")
+        jump_length = Parameter(
+            name="jump_length",
+            value=float(1.0),
+            fixed=False,
+            unit="angstrom",
+        )
+
+        expression = "jump_length / angstrom"
+        dependency_map = {
+            "jump_length": jump_length,
+            "angstrom": angstrom,
+        }
+
+        # Calculate the expected result from direct division
+        expected_result = jump_length / angstrom
+        expected_value = expected_result.value
+
+        # Then - This should not raise an error
+        dependent_param = Parameter(name='a', value=1.0)
+        dependent_param.make_dependent_on(
+            dependency_expression=expression,
+            dependency_map=dependency_map,
+        )
+
+        # Expect - The dependent parameter should have the same value as the direct division
+        assert dependent_param.value == pytest.approx(expected_value)
+
+        # Also test the alternative expression that previously worked
+        expression_alt = "1/angstrom * jump_length"
+        dependent_param_alt = Parameter(name='b', value=1.0)
+        dependent_param_alt.make_dependent_on(
+            dependency_expression=expression_alt,
+            dependency_map=dependency_map,
+        )
+        assert dependent_param_alt.value == pytest.approx(expected_value)
+
     def test_dependent_parameter_overwrite_dependency(self, normal_parameter: Parameter):
         # When
         dependent_parameter = Parameter.from_dependency(