diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
index 9a16ccc6..b1aa3e59 100644
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/tests.yml
@@ -120,7 +120,7 @@ jobs:
       docker-services: |
         zoo
         minio
-      pytest-arguments: --ibis --pdtransform --no-postgres
+      pytest-arguments: --ibis --pdtransform --no-postgres --no-lock_tests
 
   s3_test:
     name: S3 Tests
diff --git a/docs/source/changelog.md b/docs/source/changelog.md
index 6113284a..e5519f85 100644
--- a/docs/source/changelog.md
+++ b/docs/source/changelog.md
@@ -1,6 +1,9 @@
 # Changelog
 
-## 0.12.1 (2025-10-DD)
+## 0.12.2 (2025-10-DD)
+- Support datetime.time and timedelta as task input and output (or anywhere where JSON serialization is needed).
+
+## 0.12.1 (2025-10-10)
 - Create all metadata tables even if some metadata tables already exist.
   This fixes problems with conditional need for sync_views table.
 - Make table hooks work even without ConfigContext (see example_mssql/download_parquet_files.py)
diff --git a/docs/source/reference/api.rst b/docs/source/reference/api.rst
index 15445799..f6221cd3 100644
--- a/docs/source/reference/api.rst
+++ b/docs/source/reference/api.rst
@@ -97,7 +97,7 @@ IBM DB2
 
 .. autoclass:: pydiverse.pipedag.backend.table.sql.dialects.IBMDB2TableStore
 .. autoclass:: pydiverse.pipedag.backend.table.sql.dialects.ibm_db2.IBMDB2MaterializationDetails
-.. autoclass:: pydiverse.pipedag.backend.table.sql.dialects.ibm_db2::IBMDB2CompressionTypes
+.. autoclass:: pydiverse.pipedag.backend.table.sql.dialects.ibm_db2.IBMDB2CompressionTypes
     :members:
     :undoc-members:
 
diff --git a/src/pydiverse/pipedag/util/json.py b/src/pydiverse/pipedag/util/json.py
index 50175913..f8f6a312 100644
--- a/src/pydiverse/pipedag/util/json.py
+++ b/src/pydiverse/pipedag/util/json.py
@@ -44,6 +44,8 @@ class Type(str, Enum):
     PATHLIB_PATH = "pathlib:path"
     DT_DATE = "dt:date"
     DT_DATETIME = "dt:datetime"
+    DT_TIME = "dt:time"
+    DT_TIMEDELTA = "dt:timedelta"
 
     def __str__(self):
         return self.value
@@ -142,6 +144,19 @@ def json_default(o):
             TYPE_KEY: Type.DT_DATE,
             "date": o.isoformat(),
         }
+    if isinstance(o, dt.time):
+        return {
+            TYPE_KEY: Type.DT_TIME,
+            "time": o.isoformat(),
+        }
+    if isinstance(o, dt.timedelta):
+        return {
+            TYPE_KEY: Type.DT_TIMEDELTA,
+            "days": o.days,
+            "seconds": o.seconds,
+            "microseconds": o.microseconds,
+            # docstring of timedelta says: Representation: (days, seconds, microseconds).
+        }
     if get_origin(o) is not None:
         # must be GenericAlias
         # somehow isinstance(o, GenericAlias) did not work reliably
@@ -270,6 +285,10 @@ def get_stage(name: str | None):
         return Path(d["path"])
     if type_ == Type.DT_DATE:
         return dt.date.fromisoformat(d["date"])
+    if type_ == Type.DT_TIME:
+        return dt.time.fromisoformat(d["time"])
+    if type_ == Type.DT_TIMEDELTA:
+        return dt.timedelta(d["days"], d["seconds"], d["microseconds"])
     if type_ == Type.DT_DATETIME:
         return dt.datetime.fromisoformat(d["datetime"])
     if type_ == Type.DATA_CLASS:
diff --git a/tests/parallelize/session.py b/tests/parallelize/session.py
index 6d35acbf..e3afed12 100644
--- a/tests/parallelize/session.py
+++ b/tests/parallelize/session.py
@@ -2,10 +2,10 @@
 # SPDX-License-Identifier: BSD-3-Clause
 
 import itertools
+import multiprocessing as mp
 import os
 import signal
 import time
-from multiprocessing import Process, Queue
 from queue import Empty
 from threading import Thread
 
@@ -19,8 +19,9 @@
 class Session:
     def __init__(self, config: Config):
         self.config = config
-        self.msg_queue = Queue()
-        self.work_queue = Queue()
+        self.ctx = mp.get_context("spawn")  # or "forkserver"
+        self.msg_queue = self.ctx.Queue()
+        self.work_queue = self.ctx.Queue()
 
         self.workers = []
         self.worker_id_counter = itertools.count()
@@ -148,7 +149,7 @@ def start_workers(self, num_workers):
 
         for _ in range(num_workers):
             worker_id = next(self.worker_id_counter)
-            worker = Process(
+            worker = self.ctx.Process(
                 target=start_worker,
                 name=f"pytest-worker-{worker_id:03}",
                 args=(
diff --git a/tests/test_annotation_integrations/test_colspec_polars.py b/tests/test_annotation_integrations/test_colspec_polars.py
index 9692aec5..57dad31b 100644
--- a/tests/test_annotation_integrations/test_colspec_polars.py
+++ b/tests/test_annotation_integrations/test_colspec_polars.py
@@ -512,7 +512,9 @@ def consumer_collection(coll: CollectionType):
 
         if with_violation:
             with pytest.raises(cs.exc.MemberValidationError, match="2 members failed validation"):
-                coll.validate_polars(cast=True)
+                with structlog.testing.capture_logs() as logs:
+                    coll.validate_polars(cast=True)
+            assert logs == [{"exc_info": True, "event": "Dataframely validation failed", "log_level": "error"}]
         else:
             if with_filter:
                 # it is not really without violation