Make recursive normalizer default to be false in library option

Author: phoebusm

python/arcticdb/options.py

@@ -33,6 +33,8 @@ class LibraryOptions:
         See `__init__` for details.
     columns_per_segment: int
         See `__init__` for details.
+    recursive_normalizers: bool
+        See `__init__` for details.
     """
 
     def __init__(
@@ -43,7 +45,7 @@ def __init__(
         rows_per_segment: int = 100_000,
         columns_per_segment: int = 127,
         encoding_version: Optional[EncodingVersion] = None,
-        recursive_normalizers: bool = True,
+        recursive_normalizers: bool = False,
     ):
         """
         Parameters
@@ -126,7 +128,7 @@ def __init__(
             The encoding version to use when writing data to storage.
             v2 is faster, but still experimental, so use with caution.
 
-        recursive_normalizers: bool, default True
+        recursive_normalizers: bool, default False
             Whether to recursively normalize nested data structures when writing sequence-like or dict-like data.
             The data structure can be nested or a mix of lists and dictionaries.
             Note: If the leaf nodes cannot be natively normalized and must be written using write_pickle, those leaf nodes

python/arcticdb/version_store/library.py

@@ -995,7 +995,10 @@ def write(
             table.
         recursive_normalizers: bool, default None
             Whether to recursively normalize nested data structures when writing sequence-like or dict-like data.
-            If None, falls back to the corresponding setting in the library configuration.
+            If None, falls back to the corresponding setting in the library configuration. For libraries created with < v6.4.0,
+            the default library configuration is True, otherwise it is False.
+            The library configuration can be modified via Arctic.modify_library_option(). Please refer to
+            https://docs.arcticdb.io/latest/api/arctic/#arcticdb.Arctic.modify_library_option for more details.
             The data structure can be nested or a mix of lists and dictionaries.
 
         Returns