Use less extreme values for min and max timedelta in tests

When getting so close to the limits, weird things can happen, so we
should give some extra room to avoid flaky errors.

Signed-off-by: Leandro Lucarella <[email protected]>

Author: llucax

tests/test_timer.py

@@ -31,19 +31,12 @@ def event_loop() -> Iterator[async_solipsism.EventLoop]:
     loop.close()
 
 
-_max_timedelta_microseconds = (
-    int(
-        timedelta.max.total_seconds() * 1_000_000,
-    )
-    - 1
-)
+# We give some extra room (dividing by 10) to the max and min to avoid flaky errors
+# failing when getting too close to the limits, as these are not realistic scenarios and
+# weird things can happen.
+_max_timedelta_microseconds = int(timedelta.max.total_seconds() * 1_000_000 / 10)
 
-_min_timedelta_microseconds = (
-    int(
-        timedelta.min.total_seconds() * 1_000_000,
-    )
-    + 1
-)
+_min_timedelta_microseconds = int(timedelta.min.total_seconds() * 1_000_000 / 10)
 
 _calculate_next_tick_time_args = {
     "now": st.integers(),