Apart from not making it clear that allow_nan=False is about RFC compliance whether than for some other purpose, it also doesn't mean what it says. The argument really means "allow_nan_and_infs."

On Wed, Jun 17, 2020, 8:18 AM Greg Ewing <greg.ewing@canterbury.ac.nz> wrote:

On 17/06/20 5:42 pm, David Mertz wrote:
> I think the argument 'allow_nan' is poorly spelled. Spelling it 'strict'
> would have been much better. Maybe 'conformant'.

It should be called 'json_me_harder'.