Re: [Json] FYI ECMA, W3C, IETF coordination on JSON

["Peter F. Patel-Schneider" <[e-mail]>]

I think that everyone agrees that JSON text is a sequence of Unicode code 
points, but that's not the issue.  The issue is whether JSON strings represent 
sequences of Unicode characters or something else.

The JSON syntax in ECMA-404 appears to allow "\uDEAD" as a string. What 
Unicode character does this represent?

A similar problem infects JSON numbers.  There is no debate as to the syntax 
of JSON numbers.   However, there is very little guidance on what a JSON 
number represents.  Is -0 the same as 0?  Is -0.0 the same as 0?  Is 
1.10000000000000000000000000000000000000000000001 the same as 1.1?  Is 1E1 the 
same as 10?

peter

On 10/08/2013 12:37 PM, Carsten Bormann wrote:
> On Oct 8, 2013, at 20:26, "Peter F. Patel-Schneider" <[e-mail]> wrote:
>
>> The paragraph on strings, also below, ignores all the problems with escaped code units that do not represent a Unicode code point.
> Actually, no, this point is not ignored at all.
>
> It makes it very clear that JSON is composed of Unicode code points.
> UTF-16 only occurs once, where the twelve-character-escape for non-BMP code points is explained.
> So that may be one point where following ECMA-404*) might make the work of this WG simpler.
>
> Grüße, Carsten
>
> *) Great numbering :-)
> We should try to get RFC 7404 :-)
>