@@ -0,0 +1 @@
+:func:`pytest.approx` no longer treats finite :class:`~decimal.Decimal` values outside the :class:`float` range (such as ``Decimal("1e400")``) as infinite, which made in-tolerance comparisons fail.
@@ -0,0 +1 @@
+The failure message for :func:`pytest.approx` comparisons of :class:`~decimal.Decimal` collections no longer reports ``Max absolute difference: -inf``, or crashes for mappings, when the :class:`decimal.FloatOperation` trap is set.
@@ -0,0 +1 @@
+:func:`pytest.approx` no longer raises ``AttributeError`` when comparing :class:`numpy.ndarray` instances with ``object`` dtype, such as arrays of :class:`~decimal.Decimal`.
@@ -0,0 +1 @@
+Restored the tolerance shown by :func:`pytest.approx` for :class:`~decimal.Decimal` values, which since 9.0 displayed ``???`` for the default tolerance and reported the raw ``rel`` ratio instead of the absolute tolerance actually compared against.
@@ -0,0 +1 @@
+:func:`pytest.approx` now emits :class:`pytest.PytestApproxDecimalToleranceWarning` when a ``rel`` or ``abs`` tolerance is given as a :class:`float` that cannot be represented exactly for a :class:`~decimal.Decimal` comparison, since such a tolerance is widened to the float's exact binary value.
@@ -1290,6 +1290,9 @@ Custom warnings generated in some situations such as improper usage or deprecate
 .. autoclass:: pytest.PytestWarning
    :show-inheritance:
 
+.. autoclass:: pytest.PytestApproxDecimalToleranceWarning
+   :show-inheritance:
+
 .. autoclass:: pytest.PytestAssertRewriteWarning
    :show-inheritance:
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+:func:`pytest.approx` no longer treats finite :class:`~decimal.Decimal` values outside the :class:`float` range (such as ``Decimal("1e400")``) as infinite, which made in-tolerance comparisons fail.
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+The failure message for :func:`pytest.approx` comparisons of :class:`~decimal.Decimal` collections no longer reports ``Max absolute difference: -inf``, or crashes for mappings, when the :class:`decimal.FloatOperation` trap is set.
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+:func:`pytest.approx` no longer raises ``AttributeError`` when comparing :class:`numpy.ndarray` instances with ``object`` dtype, such as arrays of :class:`~decimal.Decimal`.
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+Restored the tolerance shown by :func:`pytest.approx` for :class:`~decimal.Decimal` values, which since 9.0 displayed ``???`` for the default tolerance and reported the raw ``rel`` ratio instead of the absolute tolerance actually compared against.
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+:func:`pytest.approx` now emits :class:`pytest.PytestApproxDecimalToleranceWarning` when a ``rel`` or ``abs`` tolerance is given as a :class:`float` that cannot be represented exactly for a :class:`~decimal.Decimal` comparison, since such a tolerance is widened to the float's exact binary value.