Frank Sachsenheim

with some reasonable exceptions per labels, sure. if you say that there are good experiences. i just wonder how that's good with comments that essentially say '+1' or 'any progress?'.

okay, it's a tool. so it helps with but doesn't conclude decisions. ;-) (btw, i meant the 'old' plain text +1s, but that hasn't been much of an issue here...

please provide a unit test that is as sparse as possible to replicate the problem? it would also be helpful if you provided the full traceback of the exception.

that's an interesting edge case: ```python @mark.parametrize( ("outer_type", "inner_type"), product(("dict", "list"), ("dict", "list")) ) def test_normalization_of_dict_with_a_field_named_type( outer_type, inner_type ): # https://github.com/pyeve/cerberus/issues/545 validator = Validator( {"document": {"type": outer_type, "schema": {"type": inner_type,},},}...

i haven't gone my step debugging routes through this, but it's quiet certain that it stems from the ambiguity of the `schema` rule with regards to lists and dicts. if...

thanks for bringing this up. i think this would be well covered as the test suite runs on all supported platforms. could you please check whether the `importlib_metadata` object could...

oh, iirc this test is covering some edge case. please try to do some research in the commit / pr / issues history.

it's close and it's not. it is generally related as it opens up the question how 'smart' Cerberus should be to complete a schema (in the other question nullable implies...

i'd say the document validation process starts upon calling `Validator.validate`. changing the behaviour could be an opt-in. either by adding a further validator-option (meh) or a subclass (would be useful...

also adding the `2.0` milestone here b/c this needs decision before its release.