diff --git a/docs/tutorials/8-fireworks.md b/docs/tutorials/8-fireworks.md
index 981f8c5d..f1ae2618 100644
--- a/docs/tutorials/8-fireworks.md
+++ b/docs/tutorials/8-fireworks.md
@@ -94,6 +94,16 @@ flow.update_config({"manager_config": {"_fworker": "fworker1"}}, name_filter="jo
 flow.update_config({"manager_config": {"_fworker": "fworker2"}}, name_filter="job2")
 ```
 
+NB: There are two ways to iterate over a `Flow`. The `iterflow` method iterates through a flow such that root nodes of the graph are always returned first. This has the benefit that the `job.output` references can always be resolved.
+`Flow` also has an `__iter__` method, meaning you can write
+
+```py
+for job_or_subflow in flow:
+    ...
+```
+
+to simply iterate through the `Flow.jobs` array. Note that `jobs` can also contain other flows.
+
 ### Launching the Jobs
 
 As described above, convert the flow to a workflow via {obj}`flow_to_workflow` and add it to your launch pad.
diff --git a/tests/core/test_flow.py b/tests/core/test_flow.py
index bbba6611..2216e753 100644
--- a/tests/core/test_flow.py
+++ b/tests/core/test_flow.py
@@ -456,7 +456,11 @@ def test_dag_validation():
     job2 = Job(add, function_args=(job1.output, 2))
     job1.function_args = (job2.output, 2)
     flow = Flow(jobs=[job1, job2])
-    with pytest.raises(ValueError):
+    with pytest.raises(
+        ValueError,
+        match="Job connectivity contains cycles therefore job execution order "
+        "cannot be determined",
+    ):
         next(flow.iterflow())