Samuel Colvin

Humm, I'm not sure about this anymore. There are two fairly intuitive representations of `b'123'` as an int: ```py b = b'123' assert int(b) == 123 assert int.from_bytes(b, 'big') ==...

no because we create the schema once and use it many times, perhaps days later.

Well `FutureDate`, `Datetime`, `PastDate` and `Datetime` can all be bear types, no further configuration is required. We'll also want `condatetime` etc., but we're also going to support the [annotated-types](https://github.com/annotated-types/annotated-types) allowing...

I've thought about this a lot, not just for rust validation but for an entirely theoretical "tydantic" package, see #10. The problem is that in rust dyn means heap and...

The other thing to say is that you can't really separate the two things - python types are used throughout the library doing otherwise would hammer performance.

Sounds good. Let me know when something is ready to review, best to fail early if we're going to fail.

I'm amazed by what I wrote. I almost never want to allow str for something like this. Up to you, but I would suggestion default false.

I've thought about this and made a start on it in #182, but I don't actually think a `Sequence` type is very useful. To match python's definition of Sequence (which...

@caniko can you give an example of what you mean?

We could do, that but really that would effectively be a union of `list | tool | set` or whatever. In which situation, it would be better to be explicit...