Enable question retries on single questions

docs/source/asking_questions.rst

@@ -70,6 +70,13 @@ You can also set the following options when you call :mod:`Child.ask <octue.reso
 - ``asynchronous`` - if ``True``, don't wait for an answer to the question (the result and other events can be :ref:`retrieved from the event store later <retrieving_asynchronous_answers>`)
 - ``timeout`` - how long in seconds to wait for an answer (``None`` by default - i.e. don't time out)
 
+If the question fails:
+- If ``raise_errors=False``, the unraised error is returned
+- If ``raise_errors=False`` and ``max_retries > 0``, the question is retried up to this number of times
+- If ``raise_errors=False``, ``max_retries > 0``, and ``prevent_retries_when`` is a list of exception types, the question is retried unless the error type is in the list
+- If ``raise_errors=False``, ``log_errors=True``, and the question fails after its final retry, the error is logged
+
+
 Exceptions raised by a child
 ----------------------------
 If a child raises an exception while processing your question, the exception will always be forwarded and re-raised in
@@ -185,8 +192,8 @@ access the event store and run:
 
 Asking multiple questions in parallel
 =====================================
-You can also ask multiple questions to a service in parallel. By default, if any of the questions fail, an error is
-raised and no answers are returned.
+You can also ask multiple questions to a service in parallel - just provide questions as dictionaries of `Child.ask`
+arguments:
 
 .. code-block:: python
 
@@ -203,18 +210,11 @@ raised and no answers are returned.
 
 This method uses multithreading, allowing all the questions to be asked at once instead of one after another.
 
-**Options**
+.. hint::
 
-- If ``raise_errors=False`` is provided, answers are returned for all successful questions while unraised errors are
-  logged and returned for unsuccessful ones
-- If ``raise_errors=False`` is provided with ``max_retries > 0``, failed questions are retried up to this number of
-  times
-- If ``raise_errors=False`` is provided with ``max_retries > 0`` and ``prevent_retries_when`` is set to a list of
-  exception types, failed questions are retried except for those whose exception types are in the list
-- ``max_workers``: The maximum number of threads that can be used to ask questions in parallel can be set via this
-  argument. It has no effect on the total number of questions that can be asked via ``Child.ask_multiple``.
-- ``log_errors``: If `True` and ``raise_errors=False``, any errors remaining once retries are exhausted are logged in
-  addition to being returned
+    The maximum number of threads that can be used to ask questions in parallel can be set via the ``max_workers``
+    argument. It has no effect on the total number of questions that can be asked, just how many can be in progress at
+    once.
 
 
 Asking a question within a service

octue/__init__.py

@@ -8,7 +8,6 @@
 logger = logging.getLogger(__name__)
 
 
-PYTHONUNBUFFERED = "PYTHONUNBUFFERED"
 __all__ = ("Runner",)
 
 
@@ -22,9 +21,3 @@
         include_process_name=bool(int(os.environ.get("INCLUDE_PROCESS_NAME_IN_LOGS", 0))),
         include_thread_name=bool(int(os.environ.get("INCLUDE_THREAD_NAME_IN_LOGS", 0))),
     )
-
-    if not os.environ.get(PYTHONUNBUFFERED):
-        logger.warning(
-            f"The {PYTHONUNBUFFERED!r} environment variable isn't set - logs may not appear in real time. Set "
-            f"{PYTHONUNBUFFERED}=1 to fix this."
-        )