Catalyst now uses capture.enable() as well, removing experimental_capture=True (#7298)

Related PR in Catalyst:
PennyLaneAI/catalyst#1657 (cc @rauletorresc)

Author: isaacdevlugt

doc/news/program_capture_sharp_bits.rst

@@ -404,16 +404,17 @@ Using program capture with Catalyst
 -----------------------------------
 
 To use the program capture feature with Catalyst, the ``qml.capture.enable()`` toggle
-is not required. Instead, when decorating a workflow with :func:`~.pennylane.qjit`, 
-add the ``experimental_capture=True`` flag:
+is also required.
 
 .. code-block:: python
 
     import pennylane as qml
 
+    qml.capture.enable()
+
     dev = qml.device('lightning.qubit', wires=1)
 
-    @qml.qjit(experimental_capture=True)
+    @qml.qjit
     @qml.qnode(dev)
     def circuit():
         qml.RX(0.1, wires=0)
@@ -907,4 +908,4 @@ TransformError: Input is not an Operator, tape, QNode, or quantum function
 
 >>> circuit()
 ...
-NotImplementedError: 
+NotImplementedError: 

doc/releases/changelog-dev.md

@@ -46,6 +46,12 @@
 
 <h3>Documentation 📝</h3>
 
+* The entry in the :doc:`/news/program_capture_sharp_bits` page for using program capture with Catalyst
+  has been updated. Instead of using ``qjit(experimental_capture=True)``, Catalyst is now compatible 
+  with the global toggles ``qml.capture.enable()`` and ``qml.capture.disable()`` for enabling and 
+  disabling program capture.
+  [(#7298)](https://github.com/PennyLaneAI/pennylane/pull/7298)
+
 <h3>Bug fixes 🐛</h3>
 
 * Adds an informative error if `qml.cond` is used with an abstract condition with