Re: [Json] Call for Consensus: Proposed Text for "8.1 Character Encoding"

[Julian Reschke <[e-mail]>]

On 2017-03-14 07:17, Tim Bray wrote:
> My position is identical to Carsten's. If you want guaranteed
> interoperability, use I-JSON. Let's not weaken this document with
> dubious handwaving that suggests anything but UTF-8 is a sane choice.
> ...

I'm sympathetic with the intent.

However, if the spec allows encodings other than UTF-8, it needs to give 
sufficient information that this needs to be done by inspecting the 
payload, not out-of-band data. Carsten has demonstrated that it's not 
simple, so let's either write down the exact steps, or at least give 
readers something to start with, such as:

"Character encoding detection can be done based on the fact that the 
first character is always US-ASCII, so the UTF-16/32 variants can be 
detected by inspecting the first octets for zeros."

Best regards, Julian