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

["Matthew A. Miller" <[e-mail]>]

Any other comments from the Working Group?  All things being on the
table, is mandating UTF-8 an acceptable outcome?

Is Pete's suggestion adequate?  I suspect not as it's very similar to
earlier text that was rejected, but confirmation would be helpful.


- m&m

Matthew A. Miller

On 4/28/17 8:23 AM, Pete Cordell wrote:
> Thanks.  I agree with your summary.
> 
> If you removed the sentence which mentions "UTF-16, or UTF-32" from the
> text that you proposed on 17 April, I'd be happy with that text.
> 
> Thanks again,
> 
> Pete.
> 
> On 27/04/2017 17:21, Matthew A. Miller wrote:
>> I see consensus for text in Section 8.1 pending an Appendix on encoding
>> detection, but nothing for the Appendix itself.
>>
>> Looking deeper in the threads again, it appears to me that:
>>
>> * There is consensus to say "just use UTF8" in many (most) scenarios
>> * There is rough consensus to say "always use UTF8", but concern this
>> change goes beyond the charter
>> * There is no consensus on a detection algorithm
>>
>> While the charter calls for "absolute minimal changes", it calls out RFC
>> 7159 (and its -bis) as documenting "interoperability concerns when
>> exchanging JSON over a network".  Documenting this interoperability is
>> one of its primary goals.
>>
>> Therefore, the argument is made that a change to the effect of "JSON
>> text MUST be encoded as UTF-8" is considered an "absolute minimum
>> change" in order to achieve the goal of network interoperability.
>>
>> To the working group:
>>
>> * Is there strong objection to mandating only UTF-8?
>> * Does anyone have suggested text to that effect?
>>
>>
>> - m&m
>>
>> Matthew A. Miller
>> JSONbis Chair